Latent Space: The AI Engineer Podcast

• The Agent Network — Dharmesh Shah

  If you’re in SF: Join us for the Claude Plays Pokemon hackathon this Sunday!If you’re not: Fill out the 2025 State of AI Eng survey for $250 in Amazon cards!We are SO excited to share our conversation with Dharmesh Shah, co-founder of HubSpot and creator of Agent.ai.A particularly compelling concept we discussed is the idea of "hybrid teams" - the next evolution in workplace organization where human workers collaborate with AI agents as team members. Just as we previously saw hybrid teams emerge in terms of full-time vs. contract workers, or in-office vs. remote workers, Dharmesh predicts that the next frontier will be teams composed of both human and AI members. This raises interesting questions about team dynamics, trust, and how to effectively delegate tasks between human and AI team members.The discussion of business models in AI reveals an important distinction between Work as a Service (WaaS) and Results as a Service (RaaS), something Dharmesh has written extensively about. While RaaS has gained popularity, particularly in customer support applications where outcomes are easily measurable, Dharmesh argues that this model may be over-indexed. Not all AI applications have clearly definable outcomes or consistent economic value per transaction, making WaaS more appropriate in many cases. This insight is particularly relevant for businesses considering how to monetize AI capabilities.The technical challenges of implementing effective agent systems are also explored, particularly around memory and authentication. Shah emphasizes the importance of cross-agent memory sharing and the need for more granular control over data access. He envisions a future where users can selectively share parts of their data with different agents, similar to how OAuth works but with much finer control. This points to significant opportunities in developing infrastructure for secure and efficient agent-to-agent communication and data sharing.Other highlights from our conversation* The Evolution of AI-Powered Agents – Exploring how AI agents have evolved from simple chatbots to sophisticated multi-agent systems, and the role of MCPs in enabling that.* Hybrid Digital Teams and the Future of Work – How AI agents are becoming teammates rather than just tools, and what this means for business operations and knowledge work.* Memory in AI Agents – The importance of persistent memory in AI systems and how shared memory across agents could enhance collaboration and efficiency.* Business Models for AI Agents – Exploring the shift from software as a service (SaaS) to work as a service (WaaS) and results as a service (RaaS), and what this means for monetization.* The Role of Standards Like MCP – Why MCP has been widely adopted and how it enables agent collaboration, tool use, and discovery.* The Future of AI Code Generation and Software Engineering – How AI-assisted coding is changing the role of software engineers and what skills will matter most in the future.* Domain Investing and Efficient Markets – Dharmesh’s approach to domain investing and how inefficiencies in digital asset markets create business opportunities.* The Philosophy of Saying No – Lessons from "Sorry, You Must Pass" and how prioritization leads to greater productivity and focus.Timestamps* 00:00 Introduction and Guest Welcome* 02:29 Dharmesh Shah's Journey into AI* 05:22 Defining AI Agents* 06:45 The Evolution and Future of AI Agents* 13:53 Graph Theory and Knowledge Representation* 20:02 Engineering Practices and Overengineering* 25:57 The Role of Junior Engineers in the AI Era* 28:20 Multi-Agent Systems and MCP Standards* 35:55 LinkedIn's Legal Battles and Data Scraping* 37:32 The Future of AI and Hybrid Teams* 39:19 Building Agent AI: A Professional Network for Agents* 40:43 Challenges and Innovations in Agent AI* 45:02 The Evolution of UI in AI Systems* 01:00:25 Business Models: Work as a Service vs. Results as a Service* 01:09:17 The Future Value of Engineers* 01:09:51 Exploring the Role of Agents* 01:10:28 The Importance of Memory in AI* 01:11:02 Challenges and Opportunities in AI Memory* 01:12:41 Selective Memory and Privacy Concerns* 01:13:27 The Evolution of AI Tools and Platforms* 01:18:23 Domain Names and AI Projects* 01:32:08 Balancing Work and Personal Life* 01:35:52 Final Thoughts and ReflectionsTranscriptAlessio [00:00:04]: Hey everyone, welcome back to the Latent Space podcast. This is Alessio, partner and CTO at Decibel Partners, and I'm joined by my co-host Swyx, founder of Small AI.swyx [00:00:12]: Hello, and today we're super excited to have Dharmesh Shah to join us. I guess your relevant title here is founder of Agent AI.Dharmesh [00:00:20]: Yeah, that's true for this. Yeah, creator of Agent.ai and co-founder of HubSpot.swyx [00:00:25]: Co-founder of HubSpot, which I followed for many years, I think 18 years now, gonna be 19 soon. And you caught, you know, people can catch up on your HubSpot story elsewhere. I should also thank Sean Puri, who I've chatted with back and forth, who's been, I guess, getting me in touch with your people. But also, I think like, just giving us a lot of context, because obviously, My First Million joined you guys, and they've been chatting with you guys a lot. So for the business side, we can talk about that, but I kind of wanted to engage your CTO, agent, engineer side of things. So how did you get agent religion?Dharmesh [00:01:00]: Let's see. So I've been working, I'll take like a half step back, a decade or so ago, even though actually more than that. So even before HubSpot, the company I was contemplating that I had named for was called Ingenisoft. And the idea behind Ingenisoft was a natural language interface to business software. Now realize this is 20 years ago, so that was a hard thing to do. But the actual use case that I had in mind was, you know, we had data sitting in business systems like a CRM or something like that. And my kind of what I thought clever at the time. Oh, what if we used email as the kind of interface to get to business software? And the motivation for using email is that it automatically works when you're offline. So imagine I'm getting on a plane or I'm on a plane. There was no internet on planes back then. It's like, oh, I'm going through business cards from an event I went to. I can just type things into an email just to have them all in the backlog. When it reconnects, it sends those emails to a processor that basically kind of parses effectively the commands and updates the software, sends you the file, whatever it is. And there was a handful of commands. I was a little bit ahead of the times in terms of what was actually possible. And I reattempted this natural language thing with a product called ChatSpot that I did back 20...swyx [00:02:12]: Yeah, this is your first post-ChatGPT project.Dharmesh [00:02:14]: I saw it come out. Yeah. And so I've always been kind of fascinated by this natural language interface to software. Because, you know, as software developers, myself included, we've always said, oh, we build intuitive, easy-to-use applications. And it's not intuitive at all, right? Because what we're doing is... We're taking the mental model that's in our head of what we're trying to accomplish with said piece of software and translating that into a series of touches and swipes and clicks and things like that. And there's nothing natural or intuitive about it. And so natural language interfaces, for the first time, you know, whatever the thought is you have in your head and expressed in whatever language that you normally use to talk to yourself in your head, you can just sort of emit that and have software do something. And I thought that was kind of a breakthrough, which it has been. And it's gone. So that's where I first started getting into the journey. I started because now it actually works, right? So once we got ChatGPT and you can take, even with a few-shot example, convert something into structured, even back in the ChatGP 3.5 days, it did a decent job in a few-shot example, convert something to structured text if you knew what kinds of intents you were going to have. And so that happened. And that ultimately became a HubSpot project. But then agents intrigued me because I'm like, okay, well, that's the next step here. So chat's great. Love Chat UX. But if we want to do something even more meaningful, it felt like the next kind of advancement is not this kind of, I'm chatting with some software in a kind of a synchronous back and forth model, is that software is going to do things for me in kind of a multi-step way to try and accomplish some goals. So, yeah, that's when I first got started. It's like, okay, what would that look like? Yeah. And I've been obsessed ever since, by the way.Alessio [00:03:55]: Which goes back to your first experience with it, which is like you're offline. Yeah. And you want to do a task. You don't need to do it right now. You just want to queue it up for somebody to do it for you. Yes. As you think about agents, like, let's start at the easy question, which is like, how do you define an agent? Maybe. You mean the hardest question in the universe? Is that what you mean?Dharmesh [00:04:12]: You said you have an irritating take. I do have an irritating take. I think, well, some number of people have been irritated, including within my own team. So I have a very broad definition for agents, which is it's AI-powered software that accomplishes a goal. Period. That's it. And what irritates people about it is like, well, that's so broad as to be completely non-useful. And I understand that. I understand the criticism. But in my mind, if you kind of fast forward months, I guess, in AI years, the implementation of it, and we're already starting to see this, and we'll talk about this, different kinds of agents, right? So I think in addition to having a usable definition, and I like yours, by the way, and we should talk more about that, that you just came out with, the classification of agents actually is also useful, which is, is it autonomous or non-autonomous? Does it have a deterministic workflow? Does it have a non-deterministic workflow? Is it working synchronously? Is it working asynchronously? Then you have the different kind of interaction modes. Is it a chat agent, kind of like a customer support agent would be? You're having this kind of back and forth. Is it a workflow agent that just does a discrete number of steps? So there's all these different flavors of agents. So if I were to draw it in a Venn diagram, I would draw a big circle that says, this is agents, and then I have a bunch of circles, some overlapping, because they're not mutually exclusive. And so I think that's what's interesting, and we're seeing development along a bunch of different paths, right? So if you look at the first implementation of agent frameworks, you look at Baby AGI and AutoGBT, I think it was, not Autogen, that's the Microsoft one. They were way ahead of their time because they assumed this level of reasoning and execution and planning capability that just did not exist, right? So it was an interesting thought experiment, which is what it was. Even the guy that, I'm an investor in Yohei's fund that did Baby AGI. It wasn't ready, but it was a sign of what was to come. And so the question then is, when is it ready? And so lots of people talk about the state of the art when it comes to agents. I'm a pragmatist, so I think of the state of the practical. It's like, okay, well, what can I actually build that has commercial value or solves actually some discrete problem with some baseline of repeatability or verifiability?swyx [00:06:22]: There was a lot, and very, very interesting. I'm not irritated by it at all. Okay. As you know, I take a... There's a lot of anthropological view or linguistics view. And in linguistics, you don't want to be prescriptive. You want to be descriptive. Yeah. So you're a goals guy. That's the key word in your thing. And other people have other definitions that might involve like delegated trust or non-deterministic work, LLM in the loop, all that stuff. The other thing I was thinking about, just the comment on Baby AGI, LGBT. Yeah. In that piece that you just read, I was able to go through our backlog and just kind of track the winter of agents and then the summer now. Yeah. And it's... We can tell the whole story as an oral history, just following that thread. And it's really just like, I think, I tried to explain the why now, right? Like I had, there's better models, of course. There's better tool use with like, they're just more reliable. Yep. Better tools with MCP and all that stuff. And I'm sure you have opinions on that too. Business model shift, which you like a lot. I just heard you talk about RAS with MFM guys. Yep. Cost is dropping a lot. Yep. Inference is getting faster. There's more model diversity. Yep. Yep. I think it's a subtle point. It means that like, you have different models with different perspectives. You don't get stuck in the basin of performance of a single model. Sure. You can just get out of it by just switching models. Yep. Multi-agent research and RL fine tuning. So I just wanted to let you respond to like any of that.Dharmesh [00:07:44]: Yeah. A couple of things. Connecting the dots on the kind of the definition side of it. So we'll get the irritation out of the way completely. I have one more, even more irritating leap on the agent definition thing. So here's the way I think about it. By the way, the kind of word agent, I looked it up, like the English dictionary definition. The old school agent, yeah. Is when you have someone or something that does something on your behalf, like a travel agent or a real estate agent acts on your behalf. It's like proxy, which is a nice kind of general definition. So the other direction I'm sort of headed, and it's going to tie back to tool calling and MCP and things like that, is if you, and I'm not a biologist by any stretch of the imagination, but we have these single-celled organisms, right? Like the simplest possible form of what one would call life. But it's still life. It just happens to be single-celled. And then you can combine cells and then cells become specialized over time. And you have much more sophisticated organisms, you know, kind of further down the spectrum. In my mind, at the most fundamental level, you can almost think of having atomic agents. What is the simplest possible thing that's an agent that can still be called an agent? What is the equivalent of a kind of single-celled organism? And the reason I think that's useful is right now we're headed down the road, which I think is very exciting around tool use, right? That says, okay, the LLMs now can be provided a set of tools that it calls to accomplish whatever it needs to accomplish in the kind of furtherance of whatever goal it's trying to get done. And I'm not overly bothered by it, but if you think about it, if you just squint a little bit and say, well, what if everything was an agent? And what if tools were actually just atomic agents? Because then it's turtles all the way down, right? Then it's like, oh, well, all that's really happening with tool use is that we have a network of agents that know about each other through something like an MMCP and can kind of decompose a particular problem and say, oh, I'm going to delegate this to this set of agents. And why do we need to draw this distinction between tools, which are functions most of the time? And an actual agent. And so I'm going to write this irritating LinkedIn post, you know, proposing this. It's like, okay. And I'm not suggesting we should call even functions, you know, call them agents. But there is a certain amount of elegance that happens when you say, oh, we can just reduce it down to one primitive, which is an agent that you can combine in complicated ways to kind of raise the level of abstraction and accomplish higher order goals. Anyway, that's my answer. I'd say that's a success. Thank you for coming to my TED Talk on agent definitions.Alessio [00:09:54]: How do you define the minimum viable agent? Do you already have a definition for, like, where you draw the line between a cell and an atom? Yeah.Dharmesh [00:10:02]: So in my mind, it has to, at some level, use AI in order for it to—otherwise, it's just software. It's like, you know, we don't need another word for that. And so that's probably where I draw the line. So then the question, you know, the counterargument would be, well, if that's true, then lots of tools themselves are actually not agents because they're just doing a database call or a REST API call or whatever it is they're doing. And that does not necessarily qualify them, which is a fair counterargument. And I accept that. It's like a good argument. I still like to think about—because we'll talk about multi-agent systems, because I think—so we've accepted, which I think is true, lots of people have said it, and you've hopefully combined some of those clips of really smart people saying this is the year of agents, and I completely agree, it is the year of agents. But then shortly after that, it's going to be the year of multi-agent systems or multi-agent networks. I think that's where it's going to be headed next year. Yeah.swyx [00:10:54]: Opening eyes already on that. Yeah. My quick philosophical engagement with you on this. I often think about kind of the other spectrum, the other end of the cell spectrum. So single cell is life, multi-cell is life, and you clump a bunch of cells together in a more complex organism, they become organs, like an eye and a liver or whatever. And then obviously we consider ourselves one life form. There's not like a lot of lives within me. I'm just one life. And now, obviously, I don't think people don't really like to anthropomorphize agents and AI. Yeah. But we are extending our consciousness and our brain and our functionality out into machines. I just saw you were a Bee. Yeah. Which is, you know, it's nice. I have a limitless pendant in my pocket.Dharmesh [00:11:37]: I got one of these boys. Yeah.swyx [00:11:39]: I'm testing it all out. You know, got to be early adopters. But like, we want to extend our personal memory into these things so that we can be good at the things that we're good at. And, you know, machines are good at it. Machines are there. So like, my definition of life is kind of like going outside of my own body now. I don't know if you've ever had like reflections on that. Like how yours. How our self is like actually being distributed outside of you. Yeah.Dharmesh [00:12:01]: I don't fancy myself a philosopher. But you went there. So yeah, I did go there. I'm fascinated by kind of graphs and graph theory and networks and have been for a long, long time. And to me, we're sort of all nodes in this kind of larger thing. It just so happens that we're looking at individual kind of life forms as they exist right now. But so the idea is when you put a podcast out there, there's these little kind of nodes you're putting out there of like, you know, conceptual ideas. Once again, you have varying kind of forms of those little nodes that are up there and are connected in varying and sundry ways. And so I just think of myself as being a node in a massive, massive network. And I'm producing more nodes as I put content or ideas. And, you know, you spend some portion of your life collecting dots, experiences, people, and some portion of your life then connecting dots from the ones that you've collected over time. And I found that really interesting things happen and you really can't know in advance how those dots are necessarily going to connect in the future. And that's, yeah. So that's my philosophical take. That's the, yes, exactly. Coming back.Alessio [00:13:04]: Yep. Do you like graph as an agent? Abstraction? That's been one of the hot topics with LandGraph and Pydantic and all that.Dharmesh [00:13:11]: I do. The thing I'm more interested in terms of use of graphs, and there's lots of work happening on that now, is graph data stores as an alternative in terms of knowledge stores and knowledge graphs. Yeah. Because, you know, so I've been in software now 30 plus years, right? So it's not 10,000 hours. It's like 100,000 hours that I've spent doing this stuff. And so I've grew up with, so back in the day, you know, I started on mainframes. There was a product called IMS from IBM, which is basically an index database, what we'd call like a key value store today. Then we've had relational databases, right? We have tables and columns and foreign key relationships. We all know that. We have document databases like MongoDB, which is sort of a nested structure keyed by a specific index. We have vector stores, vector embedding database. And graphs are interesting for a couple of reasons. One is, so it's not classically structured in a relational way. When you say structured database, to most people, they're thinking tables and columns and in relational database and set theory and all that. Graphs still have structure, but it's not the tables and columns structure. And you could wonder, and people have made this case, that they are a better representation of knowledge for LLMs and for AI generally than other things. So that's kind of thing number one conceptually, and that might be true, I think is possibly true. And the other thing that I really like about that in the context of, you know, I've been in the context of data stores for RAG is, you know, RAG, you say, oh, I have a million documents, I'm going to build the vector embeddings, I'm going to come back with the top X based on the semantic match, and that's fine. All that's very, very useful. But the reality is something gets lost in the chunking process and the, okay, well, those tend, you know, like, you don't really get the whole picture, so to speak, and maybe not even the right set of dimensions on the kind of broader picture. And it makes intuitive sense to me that if we did capture it properly in a graph form, that maybe that feeding into a RAG pipeline will actually yield better results for some use cases, I don't know, but yeah.Alessio [00:15:03]: And do you feel like at the core of it, there's this difference between imperative and declarative programs? Because if you think about HubSpot, it's like, you know, people and graph kind of goes hand in hand, you know, but I think maybe the software before was more like primary foreign key based relationship, versus now the models can traverse through the graph more easily.Dharmesh [00:15:22]: Yes. So I like that representation. There's something. It's just conceptually elegant about graphs and just from the representation of it, they're much more discoverable, you can kind of see it, there's observability to it, versus kind of embeddings, which you can't really do much with as a human. You know, once they're in there, you can't pull stuff back out. But yeah, I like that kind of idea of it. And the other thing that's kind of, because I love graphs, I've been long obsessed with PageRank from back in the early days. And, you know, one of the kind of simplest algorithms in terms of coming up, you know, with a phone, everyone's been exposed to PageRank. And the idea is that, and so I had this other idea for a project, not a company, and I have hundreds of these, called NodeRank, is to be able to take the idea of PageRank and apply it to an arbitrary graph that says, okay, I'm going to define what authority looks like and say, okay, well, that's interesting to me, because then if you say, I'm going to take my knowledge store, and maybe this person that contributed some number of chunks to the graph data store has more authority on this particular use case or prompt that's being submitted than this other one that may, or maybe this one was more. popular, or maybe this one has, whatever it is, there should be a way for us to kind of rank nodes in a graph and sort them in some, some useful way. Yeah.swyx [00:16:34]: So I think that's generally useful for, for anything. I think the, the problem, like, so even though at my conferences, GraphRag is super popular and people are getting knowledge, graph religion, and I will say like, it's getting space, getting traction in two areas, conversation memory, and then also just rag in general, like the, the, the document data. Yeah. It's like a source. Most ML practitioners would say that knowledge graph is kind of like a dirty word. The graph database, people get graph religion, everything's a graph, and then they, they go really hard into it and then they get a, they get a graph that is too complex to navigate. Yes. And so like the, the, the simple way to put it is like you at running HubSpot, you know, the power of graphs, the way that Google has pitched them for many years, but I don't suspect that HubSpot itself uses a knowledge graph. No. Yeah.Dharmesh [00:17:26]: So when is it over engineering? Basically? It's a great question. I don't know. So the question now, like in AI land, right, is the, do we necessarily need to understand? So right now, LLMs for, for the most part are somewhat black boxes, right? We sort of understand how the, you know, the algorithm itself works, but we really don't know what's going on in there and, and how things come out. So if a graph data store is able to produce the outcomes we want, it's like, here's a set of queries I want to be able to submit and then it comes out with useful content. Maybe the underlying data store is as opaque as a vector embeddings or something like that, but maybe it's fine. Maybe we don't necessarily need to understand it to get utility out of it. And so maybe if it's messy, that's okay. Um, that's, it's just another form of lossy compression. Uh, it's just lossy in a way that we just don't completely understand in terms of, because it's going to grow organically. Uh, and it's not structured. It's like, ah, we're just gonna throw a bunch of stuff in there. Let the, the equivalent of the embedding algorithm, whatever they called in graph land. Um, so the one with the best results wins. I think so. Yeah.swyx [00:18:26]: Or is this the practical side of me is like, yeah, it's, if it's useful, we don't necessarilyDharmesh [00:18:30]: need to understand it.swyx [00:18:30]: I have, I mean, I'm happy to push back as long as you want. Uh, it's not practical to evaluate like the 10 different options out there because it takes time. It takes people, it takes, you know, resources, right? Set. That's the first thing. Second thing is your evals are typically on small things and some things only work at scale. Yup. Like graphs. Yup.Dharmesh [00:18:46]: Yup. That's, yeah, no, that's fair. And I think this is one of the challenges in terms of implementation of graph databases is that the most common approach that I've seen developers do, I've done it myself, is that, oh, I've got a Postgres database or a MySQL or whatever. I can represent a graph with a very set of tables with a parent child thing or whatever. And that sort of gives me the ability, uh, why would I need anything more than that? And the answer is, well, if you don't need anything more than that, you don't need anything more than that. But there's a high chance that you're sort of missing out on the actual value that, uh, the graph representation gives you. Which is the ability to traverse the graph, uh, efficiently in ways that kind of going through the, uh, traversal in a relational database form, even though structurally you have the data, practically you're not gonna be able to pull it out in, in useful ways. Uh, so you wouldn't like represent a social graph, uh, in, in using that kind of relational table model. It just wouldn't scale. It wouldn't work.swyx [00:19:36]: Uh, yeah. Uh, I think we want to move on to MCP. Yeah. But I just want to, like, just engineering advice. Yeah. Uh, obviously you've, you've, you've run, uh, you've, you've had to do a lot of projects and run a lot of teams. Do you have a general rule for over-engineering or, you know, engineering ahead of time? You know, like, because people, we know premature engineering is the root of all evil. Yep. But also sometimes you just have to. Yep. When do you do it? Yes.Dharmesh [00:19:59]: It's a great question. This is, uh, a question as old as time almost, which is what's the right and wrong levels of abstraction. That's effectively what, uh, we're answering when we're trying to do engineering. I tend to be a pragmatist, right? So here's the thing. Um, lots of times doing something the right way. Yeah. It's like a marginal increased cost in those cases. Just do it the right way. And this is what makes a, uh, a great engineer or a good engineer better than, uh, a not so great one. It's like, okay, all things being equal. If it's going to take you, you know, roughly close to constant time anyway, might as well do it the right way. Like, so do things well, then the question is, okay, well, am I building a framework as the reusable library? To what degree, uh, what am I anticipating in terms of what's going to need to change in this thing? Uh, you know, along what dimension? And then I think like a business person in some ways, like what's the return on calories, right? So, uh, and you look at, um, energy, the expected value of it's like, okay, here are the five possible things that could happen, uh, try to assign probabilities like, okay, well, if there's a 50% chance that we're going to go down this particular path at some day, like, or one of these five things is going to happen and it costs you 10% more to engineer for that. It's basically, it's something that yields a kind of interest compounding value. Um, as you get closer to the time of, of needing that versus having to take on debt, which is when you under engineer it, you're taking on debt. You're going to have to pay off when you do get to that eventuality where something happens. One thing as a pragmatist, uh, so I would rather under engineer something than over engineer it. If I were going to err on the side of something, and here's the reason is that when you under engineer it, uh, yes, you take on tech debt, uh, but the interest rate is relatively known and payoff is very, very possible, right? Which is, oh, I took a shortcut here as a result of which now this thing that should have taken me a week is now going to take me four weeks. Fine. But if that particular thing that you thought might happen, never actually, you never have that use case transpire or just doesn't, it's like, well, you just save yourself time, right? And that has value because you were able to do other things instead of, uh, kind of slightly over-engineering it away, over-engineering it. But there's no perfect answers in art form in terms of, uh, and yeah, we'll, we'll bring kind of this layers of abstraction back on the code generation conversation, which we'll, uh, I think I have later on, butAlessio [00:22:05]: I was going to ask, we can just jump ahead quickly. Yeah. Like, as you think about vibe coding and all that, how does the. Yeah. Percentage of potential usefulness change when I feel like we over-engineering a lot of times it's like the investment in syntax, it's less about the investment in like arc exacting. Yep. Yeah. How does that change your calculus?Dharmesh [00:22:22]: A couple of things, right? One is, um, so, you know, going back to that kind of ROI or a return on calories, kind of calculus or heuristic you think through, it's like, okay, well, what is it going to cost me to put this layer of abstraction above the code that I'm writing now, uh, in anticipating kind of future needs. If the cost of fixing, uh, or doing under engineering right now. Uh, we'll trend towards zero that says, okay, well, I don't have to get it right right now because even if I get it wrong, I'll run the thing for six hours instead of 60 minutes or whatever. It doesn't really matter, right? Like, because that's going to trend towards zero to be able, the ability to refactor a code. Um, and because we're going to not that long from now, we're going to have, you know, large code bases be able to exist, uh, you know, as, as context, uh, for a code generation or a code refactoring, uh, model. So I think it's going to make it, uh, make the case for under engineering, uh, even stronger. Which is why I take on that cost. You just pay the interest when you get there, it's not, um, just go on with your life vibe coded and, uh, come back when you need to. Yeah.Alessio [00:23:18]: Sometimes I feel like there's no decision-making in some things like, uh, today I built a autosave for like our internal notes platform and I literally just ask them cursor. Can you add autosave? Yeah. I don't know if it's over under engineer. Yep. I just vibe coded it. Yep. And I feel like at some point we're going to get to the point where the models kindDharmesh [00:23:36]: of decide where the right line is, but this is where the, like the, in my mind, the danger is, right? So there's two sides to this. One is the cost of kind of development and coding and things like that stuff that, you know, we talk about. But then like in your example, you know, one of the risks that we have is that because adding a feature, uh, like a save or whatever the feature might be to a product as that price tends towards zero, are we going to be less discriminant about what features we add as a result of making more product products more complicated, which has a negative impact on the user and navigate negative impact on the business. Um, and so that's the thing I worry about if it starts to become too easy, are we going to be. Too promiscuous in our, uh, kind of extension, adding product extensions and things like that. It's like, ah, why not add X, Y, Z or whatever back then it was like, oh, we only have so many engineering hours or story points or however you measure things. Uh, that least kept us in check a little bit. Yeah.Alessio [00:24:22]: And then over engineering, you're like, yeah, it's kind of like you're putting that on yourself. Yeah. Like now it's like the models don't understand that if they add too much complexity, it's going to come back to bite them later. Yep. So they just do whatever they want to do. Yeah. And I'm curious where in the workflow that's going to be, where it's like, Hey, this is like the amount of complexity and over-engineering you can do before you got to ask me if we should actually do it versus like do something else.Dharmesh [00:24:45]: So you know, we've already, let's like, we're leaving this, uh, in the code generation world, this kind of compressed, um, cycle time. Right. It's like, okay, we went from auto-complete, uh, in the GitHub co-pilot to like, oh, finish this particular thing and hit tab to a, oh, I sort of know your file or whatever. I can write out a full function to you to now I can like hold a bunch of the context in my head. Uh, so we can do app generation, which we have now with lovable and bolt and repletage. Yeah. Association and other things. So then the question is, okay, well, where does it naturally go from here? So we're going to generate products. Make sense. We might be able to generate platforms as though I want a platform for ERP that does this, whatever. And that includes the API's includes the product and the UI, and all the things that make for a platform. There's no nothing that says we would stop like, okay, can you generate an entire software company someday? Right. Uh, with the platform and the monetization and the go-to-market and the whatever. And you know, that that's interesting to me in terms of, uh, you know, what, when you take it to almost ludicrous levels. of abstract.swyx [00:25:39]: It's like, okay, turn it to 11. You mentioned vibe coding, so I have to, this is a blog post I haven't written, but I'm kind of exploring it. Is the junior engineer dead?Dharmesh [00:25:49]: I don't think so. I think what will happen is that the junior engineer will be able to, if all they're bringing to the table is the fact that they are a junior engineer, then yes, they're likely dead. But hopefully if they can communicate with carbon-based life forms, they can interact with product, if they're willing to talk to customers, they can take their kind of basic understanding of engineering and how kind of software works. I think that has value. So I have a 14-year-old right now who's taking Python programming class, and some people ask me, it's like, why is he learning coding? And my answer is, is because it's not about the syntax, it's not about the coding. What he's learning is like the fundamental thing of like how things work. And there's value in that. I think there's going to be timeless value in systems thinking and abstractions and what that means. And whether functions manifested as math, which he's going to get exposed to regardless, or there are some core primitives to the universe, I think, that the more you understand them, those are what I would kind of think of as like really large dots in your life that will have a higher gravitational pull and value to them that you'll then be able to. So I want him to collect those dots, and he's not resisting. So it's like, okay, while he's still listening to me, I'm going to have him do things that I think will be useful.swyx [00:26:59]: You know, part of one of the pitches that I evaluated for AI engineer is a term. And the term is that maybe the traditional interview path or career path of software engineer goes away, which is because what's the point of lead code? Yeah. And, you know, it actually matters more that you know how to work with AI and to implement the things that you want. Yep.Dharmesh [00:27:16]: That's one of the like interesting things that's happened with generative AI. You know, you go from machine learning and the models and just that underlying form, which is like true engineering, right? Like the actual, what I call real engineering. I don't think of myself as a real engineer, actually. I'm a developer. But now with generative AI. We call it AI and it's obviously got its roots in machine learning, but it just feels like fundamentally different to me. Like you have the vibe. It's like, okay, well, this is just a whole different approach to software development to so many different things. And so I'm wondering now, it's like an AI engineer is like, if you were like to draw the Venn diagram, it's interesting because the cross between like AI things, generative AI and what the tools are capable of, what the models do, and this whole new kind of body of knowledge that we're still building out, it's still very young, intersected with kind of classic engineering, software engineering. Yeah.swyx [00:28:04]: I just described the overlap as it separates out eventually until it's its own thing, but it's starting out as a software. Yeah.Alessio [00:28:11]: That makes sense. So to close the vibe coding loop, the other big hype now is MCPs. Obviously, I would say Cloud Desktop and Cursor are like the two main drivers of MCP usage. I would say my favorite is the Sentry MCP. I can pull in errors and then you can just put the context in Cursor. How do you think about that abstraction layer? Does it feel... Does it feel almost too magical in a way? Do you think it's like you get enough? Because you don't really see how the server itself is then kind of like repackaging theDharmesh [00:28:41]: information for you? I think MCP as a standard is one of the better things that's happened in the world of AI because a standard needed to exist and absent a standard, there was a set of things that just weren't possible. Now, we can argue whether it's the best possible manifestation of a standard or not. Does it do too much? Does it do too little? I get that, but it's just simple enough to both be useful and unobtrusive. It's understandable and adoptable by mere mortals, right? It's not overly complicated. You know, a reasonable engineer can put a stand up an MCP server relatively easily. The thing that has me excited about it is like, so I'm a big believer in multi-agent systems. And so that's going back to our kind of this idea of an atomic agent. So imagine the MCP server, like obviously it calls tools, but the way I think about it, so I'm working on my current passion project is agent.ai. And we'll talk more about that in a little bit. More about the, I think we should, because I think it's interesting not to promote the project at all, but there's some interesting ideas in there. One of which is around, we're going to need a mechanism for, if agents are going to collaborate and be able to delegate, there's going to need to be some form of discovery and we're going to need some standard way. It's like, okay, well, I just need to know what this thing over here is capable of. We're going to need a registry, which Anthropic's working on. I'm sure others will and have been doing directories of, and there's going to be a standard around that too. How do you build out a directory of MCP servers? I think that's going to unlock so many things just because, and we're already starting to see it. So I think MCP or something like it is going to be the next major unlock because it allows systems that don't know about each other, don't need to, it's that kind of decoupling of like Sentry and whatever tools someone else was building. And it's not just about, you know, Cloud Desktop or things like, even on the client side, I think we're going to see very interesting consumers of MCP, MCP clients versus just the chat body kind of things. Like, you know, Cloud Desktop and Cursor and things like that. But yeah, I'm very excited about MCP in that general direction.swyx [00:30:39]: I think the typical cynical developer take, it's like, we have OpenAPI. Yeah. What's the new thing? I don't know if you have a, do you have a quick MCP versus everything else? Yeah.Dharmesh [00:30:49]: So it's, so I like OpenAPI, right? So just a descriptive thing. It's OpenAPI. OpenAPI. Yes, that's what I meant. So it's basically a self-documenting thing. We can do machine-generated, lots of things from that output. It's a structured definition of an API. I get that, love it. But MCPs sort of are kind of use case specific. They're perfect for exactly what we're trying to use them for around LLMs in terms of discovery. It's like, okay, I don't necessarily need to know kind of all this detail. And so right now we have, we'll talk more about like MCP server implementations, but We will? I think, I don't know. Maybe we won't. At least it's in my head. It's like a back processor. But I do think MCP adds value above OpenAPI. It's, yeah, just because it solves this particular thing. And if we had come to the world, which we have, like, it's like, hey, we already have OpenAPI. It's like, if that were good enough for the universe, the universe would have adopted it already. There's a reason why MCP is taking office because marginally adds something that was missing before and doesn't go too far. And so that's why the kind of rate of adoption, you folks have written about this and talked about it. Yeah, why MCP won. Yeah. And it won because the universe decided that this was useful and maybe it gets supplanted by something else. Yeah. And maybe we discover, oh, maybe OpenAPI was good enough the whole time. I doubt that.swyx [00:32:09]: The meta lesson, this is, I mean, he's an investor in DevTools companies. I work in developer experience at DevRel in DevTools companies. Yep. Everyone wants to own the standard. Yeah. I'm sure you guys have tried to launch your own standards. Actually, it's Houseplant known for a standard, you know, obviously inbound marketing. But is there a standard or protocol that you ever tried to push? No.Dharmesh [00:32:30]: And there's a reason for this. Yeah. Is that? And I don't mean, need to mean, speak for the people of HubSpot, but I personally. You kind of do. I'm not smart enough. That's not the, like, I think I have a. You're smart. Not enough for that. I'm much better off understanding the standards that are out there. And I'm more on the composability side. Let's, like, take the pieces of technology that exist out there, combine them in creative, unique ways. And I like to consume standards. I don't like to, and that's not that I don't like to create them. I just don't think I have the, both the raw wattage or the credibility. It's like, okay, well, who the heck is Dharmesh, and why should we adopt a standard he created?swyx [00:33:07]: Yeah, I mean, there are people who don't monetize standards, like OpenTelemetry is a big standard, and LightStep never capitalized on that.Dharmesh [00:33:15]: So, okay, so if I were to do a standard, there's two things that have been in my head in the past. I was one around, a very, very basic one around, I don't even have the domain, I have a domain for everything, for open marketing. Because the issue we had in HubSpot grew up in the marketing space. There we go. There was no standard around data formats and things like that. It doesn't go anywhere. But the other one, and I did not mean to go here, but I'm going to go here. It's called OpenGraph. I know the term was already taken, but it hasn't been used for like 15 years now for its original purpose. But what I think should exist in the world is right now, our information, all of us, nodes are in the social graph at Meta or the professional graph at LinkedIn. Both of which are actually relatively closed in actually very annoying ways. Like very, very closed, right? Especially LinkedIn. Especially LinkedIn. I personally believe that if it's my data, and if I would get utility out of it being open, I should be able to make my data open or publish it in whatever forms that I choose, as long as I have control over it as opt-in. So the idea is around OpenGraph that says, here's a standard, here's a way to publish it. I should be able to go to OpenGraph.org slash Dharmesh dot JSON and get it back. And it's like, here's your stuff, right? And I can choose along the way and people can write to it and I can prove. And there can be an entire system. And if I were to do that, I would do it as a... Like a public benefit, non-profit-y kind of thing, as this is a contribution to society. I wouldn't try to commercialize that. Have you looked at AdProto? What's that? AdProto.swyx [00:34:43]: It's the protocol behind Blue Sky. Okay. My good friend, Dan Abramov, who was the face of React for many, many years, now works there. And he actually did a talk that I can send you, which basically kind of tries to articulate what you just said. But he does, he loves doing these like really great analogies, which I think you'll like. Like, you know, a lot of our data is behind a handle, behind a domain. Yep. So he's like, all right, what if we flip that? What if it was like our handle and then the domain? Yep. So, and that's really like your data should belong to you. Yep. And I should not have to wait 30 days for my Twitter data to export. Yep.Dharmesh [00:35:19]: you should be able to at least be able to automate it or do like, yes, I should be able to plug it into an agentic thing. Yeah. Yes. I think we're... Because so much of our data is... Locked up. I think the trick here isn't that standard. It is getting the normies to care.swyx [00:35:37]: Yeah. Because normies don't care.Dharmesh [00:35:38]: That's true. But building on that, normies don't care. So, you know, privacy is a really hot topic and an easy word to use, but it's not a binary thing. Like there are use cases where, and we make these choices all the time, that I will trade, not all privacy, but I will trade some privacy for some productivity gain or some benefit to me that says, oh, I don't care about that particular data being online if it gives me this in return, or I don't mind sharing this information with this company.Alessio [00:36:02]: If I'm getting, you know, this in return, but that sort of should be my option. I think now with computer use, you can actually automate some of the exports. Yes. Like something we've been doing internally is like everybody exports their LinkedIn connections. Yep. And then internally, we kind of merge them together to see how we can connect our companies to customers or things like that.Dharmesh [00:36:21]: And not to pick on LinkedIn, but since we're talking about it, but they feel strongly enough on the, you know, do not take LinkedIn data that they will block even browser use kind of things or whatever. They go to great, great lengths, even to see patterns of usage. And it says, oh, there's no way you could have, you know, gotten that particular thing or whatever without, and it's, so it's, there's...swyx [00:36:42]: Wasn't there a Supreme Court case that they lost? Yeah.Dharmesh [00:36:45]: So the one they lost was around someone that was scraping public data that was on the public internet. And that particular company had not signed any terms of service or whatever. It's like, oh, I'm just taking data that's on, there was no, and so that's why they won. But now, you know, the question is around, can LinkedIn... I think they can. Like, when you use, as a user, you use LinkedIn, you are signing up for their terms of service. And if they say, well, this kind of use of your LinkedIn account that violates our terms of service, they can shut your account down, right? They can. And they, yeah, so, you know, we don't need to make this a discussion. By the way, I love the company, don't get me wrong. I'm an avid user of the product. You know, I've got... Yeah, I mean, you've got over a million followers on LinkedIn, I think. Yeah, I do. And I've known people there for a long, long time, right? And I have lots of respect. And I understand even where the mindset originally came from of this kind of members-first approach to, you know, a privacy-first. I sort of get that. But sometimes you sort of have to wonder, it's like, okay, well, that was 15, 20 years ago. There's likely some controlled ways to expose some data on some member's behalf and not just completely be a binary. It's like, no, thou shalt not have the data.swyx [00:37:54]: Well, just pay for sales navigator.Alessio [00:37:57]: Before we move to the next layer of instruction, anything else on MCP you mentioned? Let's move back and then I'll tie it back to MCPs.Dharmesh [00:38:05]: So I think the... Open this with agent. Okay, so I'll start with... Here's my kind of running thesis, is that as AI and agents evolve, which they're doing very, very quickly, we're going to look at them more and more. I don't like to anthropomorphize. We'll talk about why this is not that. Less as just like raw tools and more like teammates. They'll still be software. They should self-disclose as being software. I'm totally cool with that. But I think what's going to happen is that in the same way you might collaborate with a team member on Slack or Teams or whatever you use, you can imagine a series of agents that do specific things just like a team member might do, that you can delegate things to. You can collaborate. You can say, hey, can you take a look at this? Can you proofread that? Can you try this? You can... Whatever it happens to be. So I think it is... I will go so far as to say it's inevitable that we're going to have hybrid teams someday. And what I mean by hybrid teams... So back in the day, hybrid teams were, oh, well, you have some full-time employees and some contractors. Then it was like hybrid teams are some people that are in the office and some that are remote. That's the kind of form of hybrid. The next form of hybrid is like the carbon-based life forms and agents and AI and some form of software. So let's say we temporarily stipulate that I'm right about that over some time horizon that eventually we're going to have these kind of digitally hybrid teams. So if that's true, then the question you sort of ask yourself is that then what needs to exist in order for us to get the full value of that new model? It's like, okay, well... You sort of need to... It's like, okay, well, how do I... If I'm building a digital team, like, how do I... Just in the same way, if I'm interviewing for an engineer or a designer or a PM, whatever, it's like, well, that's why we have professional networks, right? It's like, oh, they have a presence on likely LinkedIn. I can go through that semi-structured, structured form, and I can see the experience of whatever, you know, self-disclosed. But, okay, well, agents are going to need that someday. And so I'm like, okay, well, this seems like a thread that's worth pulling on. That says, okay. So I... So agent.ai is out there. And it's LinkedIn for agents. It's LinkedIn for agents. It's a professional network for agents. And the more I pull on that thread, it's like, okay, well, if that's true, like, what happens, right? It's like, oh, well, they have a profile just like anyone else, just like a human would. It's going to be a graph underneath, just like a professional network would be. It's just that... And you can have its, you know, connections and follows, and agents should be able to post. That's maybe how they do release notes. Like, oh, I have this new version. Whatever they decide to post, it should just be able to... Behave as a node on the network of a professional network. As it turns out, the more I think about that and pull on that thread, the more and more things, like, start to make sense to me. So it may be more than just a pure professional network. So my original thought was, okay, well, it's a professional network and agents as they exist out there, which I think there's going to be more and more of, will kind of exist on this network and have the profile. But then, and this is always dangerous, I'm like, okay, I want to see a world where thousands of agents are out there in order for the... Because those digital employees, the digital workers don't exist yet in any meaningful way. And so then I'm like, oh, can I make that easier for, like... And so I have, as one does, it's like, oh, I'll build a low-code platform for building agents. How hard could that be, right? Like, very hard, as it turns out. But it's been fun. So now, agent.ai has 1.3 million users. 3,000 people have actually, you know, built some variation of an agent, sometimes just for their own personal productivity. About 1,000 of which have been published. And the reason this comes back to MCP for me, so imagine that and other networks, since I know agent.ai. So right now, we have an MCP server for agent.ai that exposes all the internally built agents that we have that do, like, super useful things. Like, you know, I have access to a Twitter API that I can subsidize the cost. And I can say, you know, if you're looking to build something for social media, these kinds of things, with a single API key, and it's all completely free right now, I'm funding it. That's a useful way for it to work. And then we have a developer to say, oh, I have this idea. I don't have to worry about open AI. I don't have to worry about, now, you know, this particular model is better. It has access to all the models with one key. And we proxy it kind of behind the scenes. And then expose it. So then we get this kind of community effect, right? That says, oh, well, someone else may have built an agent to do X. Like, I have an agent right now that I built for myself to do domain valuation for website domains because I'm obsessed with domains, right? And, like, there's no efficient market for domains. There's no Zillow for domains right now that tells you, oh, here are what houses in your neighborhood sold for. It's like, well, why doesn't that exist? We should be able to solve that problem. And, yes, you're still guessing. Fine. There should be some simple heuristic. So I built that. It's like, okay, well, let me go look for past transactions. You say, okay, I'm going to type in agent.ai, agent.com, whatever domain. What's it actually worth? I'm looking at buying it. It can go and say, oh, which is what it does. It's like, I'm going to go look at are there any published domain transactions recently that are similar, either use the same word, same top-level domain, whatever it is. And it comes back with an approximate value, and it comes back with its kind of rationale for why it picked the value and comparable transactions. Oh, by the way, this domain sold for published. Okay. So that agent now, let's say, existed on the web, on agent.ai. Then imagine someone else says, oh, you know, I want to build a brand-building agent for startups and entrepreneurs to come up with names for their startup. Like a common problem, every startup is like, ah, I don't know what to call it. And so they type in five random words that kind of define whatever their startup is. And you can do all manner of things, one of which is like, oh, well, I need to find the domain for it. What are possible choices? Now it's like, okay, well, it would be nice to know if there's an aftermarket price for it, if it's listed for sale. Awesome. Then imagine calling this valuation agent. It's like, okay, well, I want to find where the arbitrage is, where the agent valuation tool says this thing is worth $25,000. It's listed on GoDaddy for $5,000. It's close enough. Let's go do that. Right? And that's a kind of composition use case that in my future state. Thousands of agents on the network, all discoverable through something like MCP. And then you as a developer of agents have access to all these kind of Lego building blocks based on what you're trying to solve. Then you blend in orchestration, which is getting better and better with the reasoning models now. Just describe the problem that you have. Now, the next layer that we're all contending with is that how many tools can you actually give an LLM before the LLM breaks? That number used to be like 15 or 20 before you kind of started to vary dramatically. And so that's the thing I'm thinking about now. It's like, okay, if I want to... If I want to expose 1,000 of these agents to a given LLM, obviously I can't give it all 1,000. Is there some intermediate layer that says, based on your prompt, I'm going to make a best guess at which agents might be able to be helpful for this particular thing? Yeah.Alessio [00:44:37]: Yeah, like RAG for tools. Yep. I did build the Latent Space Researcher on agent.ai. Okay. Nice. Yeah, that seems like, you know, then there's going to be a Latent Space Scheduler. And then once I schedule a research, you know, and you build all of these things. By the way, my apologies for the user experience. You realize I'm an engineer. It's pretty good.swyx [00:44:56]: I think it's a normie-friendly thing. Yeah. That's your magic. HubSpot does the same thing.Alessio [00:45:01]: Yeah, just to like quickly run through it. You can basically create all these different steps. And these steps are like, you know, static versus like variable-driven things. How did you decide between this kind of like low-code-ish versus doing, you know, low-code with code backend versus like not exposing that at all? Any fun design decisions? Yeah. And this is, I think...Dharmesh [00:45:22]: I think lots of people are likely sitting in exactly my position right now, coming through the choosing between deterministic. Like if you're like in a business or building, you know, some sort of agentic thing, do you decide to do a deterministic thing? Or do you go non-deterministic and just let the alum handle it, right, with the reasoning models? The original idea and the reason I took the low-code stepwise, a very deterministic approach. A, the reasoning models did not exist at that time. That's thing number one. Thing number two is if you can get... If you know in your head... If you know in your head what the actual steps are to accomplish whatever goal, why would you leave that to chance? There's no upside. There's literally no upside. Just tell me, like, what steps do you need executed? So right now what I'm playing with... So one thing we haven't talked about yet, and people don't talk about UI and agents. Right now, the primary interaction model... Or they don't talk enough about it. I know some people have. But it's like, okay, so we're used to the chatbot back and forth. Fine. I get that. But I think we're going to move to a blend of... Some of those things are going to be synchronous as they are now. But some are going to be... Some are going to be async. It's just going to put it in a queue, just like... And this goes back to my... Man, I talk fast. But I have this... I only have one other speed. It's even faster. So imagine it's like if you're working... So back to my, oh, we're going to have these hybrid digital teams. Like, you would not go to a co-worker and say, I'm going to ask you to do this thing, and then sit there and wait for them to go do it. Like, that's not how the world works. So it's nice to be able to just, like, hand something off to someone. It's like, okay, well, maybe I expect a response in an hour or a day or something like that.Dharmesh [00:46:52]: In terms of when things need to happen. So the UI around agents. So if you look at the output of agent.ai agents right now, they are the simplest possible manifestation of a UI, right? That says, oh, we have inputs of, like, four different types. Like, we've got a dropdown, we've got multi-select, all the things. It's like back in HTML, the original HTML 1.0 days, right? Like, you're the smallest possible set of primitives for a UI. And it just says, okay, because we need to collect some information from the user, and then we go do steps and do things. And generate some output in HTML or markup are the two primary examples. So the thing I've been asking myself, if I keep going down that path. So people ask me, I get requests all the time. It's like, oh, can you make the UI sort of boring? I need to be able to do this, right? And if I keep pulling on that, it's like, okay, well, now I've built an entire UI builder thing. Where does this end? And so I think the right answer, and this is what I'm going to be backcoding once I get done here, is around injecting a code generation UI generation into, the agent.ai flow, right? As a builder, you're like, okay, I'm going to describe the thing that I want, much like you would do in a vibe coding world. But instead of generating the entire app, it's going to generate the UI that exists at some point in either that deterministic flow or something like that. It says, oh, here's the thing I'm trying to do. Go generate the UI for me. And I can go through some iterations. And what I think of it as a, so it's like, I'm going to generate the code, generate the code, tweak it, go through this kind of prompt style, like we do with vibe coding now. And at some point, I'm going to be happy with it. And I'm going to hit save. And that's going to become the action in that particular step. It's like a caching of the generated code that I can then, like incur any inference time costs. It's just the actual code at that point.Alessio [00:48:29]: Yeah, I invested in a company called E2B, which does code sandbox. And they powered the LM arena web arena. So it's basically the, just like you do LMS, like text to text, they do the same for like UI generation. So if you're asking a model, how do you do it? But yeah, I think that's kind of where.Dharmesh [00:48:45]: That's the thing I'm really fascinated by. So the early LLM, you know, we're understandably, but laughably bad at simple arithmetic, right? That's the thing like my wife, Normies would ask us, like, you call this AI, like it can't, my son would be like, it's just stupid. It can't even do like simple arithmetic. And then like we've discovered over time that, and there's a reason for this, right? It's like, it's a large, there's, you know, the word language is in there for a reason in terms of what it's been trained on. It's not meant to do math, but now it's like, okay, well, the fact that it has access to a Python interpreter that I can actually call at runtime, that solves an entire body of problems that it wasn't trained to do. And it's basically a form of delegation. And so the thought that's kind of rattling around in my head is that that's great. So it's, it's like took the arithmetic problem and took it first. Now, like anything that's solvable through a relatively concrete Python program, it's able to do a bunch of things that I couldn't do before. Can we get to the same place with UI? I don't know what the future of UI looks like in a agentic AI world, but maybe let the LLM handle it, but not in the classic sense. Maybe it generates it on the fly, or maybe we go through some iterations and hit cache or something like that. So it's a little bit more predictable. Uh, I don't know, but yeah.Alessio [00:49:48]: And especially when is the human supposed to intervene? So, especially if you're composing them, most of them should not have a UI because then they're just web hooking to somewhere else. I just want to touch back. I don't know if you have more comments on this.swyx [00:50:01]: I was just going to ask when you, you said you got, you're going to go back to code. What are you coding with? What's your stack? Yep.Dharmesh [00:50:06]: Uh, so Python's my language. Uh, I'm glad that it won in terms of the AI, uh, languages, lingua franca.swyx [00:50:12]: It's the second best language for everything.Dharmesh [00:50:13]: And by the way, there, I think exactly end of one of things that I disagree with Brett Taylor on, uh, when, when he was on, and just generally, I'm a massive Brett Taylor fan, uh, smart. One of my favorite people in tech, like it was like a segment in there. He was talking about like, oh, we need a, a different language than Python or whatever. That is like built for, uh, built for AI and built. It's like, no, Brett, I don't think we do actually, it's just fine. Um, it deals with just fine, just expressive enough. And it's nice to have a language that we can use as a common denominator across both humans and AI it's, it doesn't slow the AI down. Enough, but it does make it awfully useful for us to also be able to participate in that kind of future world, uh, that we can still be somewhat useful.swyx [00:50:53]: I mean, but yeah, so it's, uh, Python, uh, cursor as my, uh, kind of code gen thing. Yeah. I would also mention that I really like your code generation thing. I have another thesis I haven't written up yet about how generative UI has kind of not fulfilled its full potential. We've seen the bolts and lovables and those are great. And then Vercel has a version of generative UI that is basically function calling pre-made components. And there's some. Thing in between where you should be able to generate the UI that you want and pin it and stick to it. And that becomes your form or yeah. And so the way I put it is, um, you know, I think that the two form factors of agents that I've seen a lot of product market fit recently has been deep research and the AI builders, like the bolt lovables. I think there's some version of this where you generate the UI, but you sort of generate the Mad Libs fill in the blanks forms, and then you, you, you keep that stable. And the deep research is. Just fills that in. Yeah. Yep. And that's it. I like that.Dharmesh [00:51:49]: Yeah. Um, so I, I, I love those, uh, kind of simple, uh, simple limitations and kind of abstractions, but then if you look at the kind of, I'll say almost like the polar opposite of that. So, so right now, most of the UIs that you and I think about or conceive, or even examples are based on the primitives and the vocabulary that we have for UI right now. It's like, oh, we have text boxes. We have check boxes. We have radio buttons. We have pulldowns. We have nav. We have clicks, touches, swipes, now voice, whatever it is, the set of primitives that exist right now, we will combine them in, uh, in interesting ways, but where I think AI is going to be headed on, I think on the UI front is the same place is headed on the science front that originally it's like, oh, well, based on the things that we know right now, it'll sort of combine them, but we're like right at the cusp of it being able to actual novel research. So maybe a future version of AI comes up with a new set of primitives that actually work better for human computer interaction than things that we've done in the past, right? It's like, I don't. I don't think it's, it ended with the, uh, the checkbox, radio button and dropdown list. Right. I think there's life beyond that.Alessio [00:52:44]: Uh, yeah, I know we're going to move to business models after, but when you talked about ivory teams, one way we talk to folks about it is like you had offshoring yet on shoring, which is like, you know, move to cheaper place in the country than offshoring. You know, it's like AI shoring. Yep. You're kind of moving some roles. That's the thing people say. Yeah. Shoring. Yeah.Dharmesh [00:53:01]: That's the first time I've ever heard of that. Yeah. Yeah.Alessio [00:53:04]: I don't know, man. But I think to me, the most interesting thing about the professional networks is like with people, you have limited availability to evaluate a person. Yeah. So you have to use previous signal as kind of like a evaluation thing. With agents, theoretically, you can have kind of like proof of work. Yeah. You know, you can run simulations and like evaluate them in that way. Yep. How do you think about that when running, building agent.ai even? It's like, you know, instead of just choosing one, I could like literally just run across all of them and figure out which one is going to work best.Dharmesh [00:53:32]: I'm a big believer. So under the covers, when you build, because the primitives are so simple, you have some sort of inputs. We know that what the variables are. Every agent that's on agent.ai automatically has a REST API. That's callable in exactly the way you would expect. Automatically shows up in the MCP server, so you're able to invoke it in whatever form you decide to. And so my expectation is that in this future state, whether it's a human hiring an agent to do a particular task or evaluating a set of five agents to do a particular task and picking the best one for their particular use case, we should be able to do that. It's like, I just want to try it, and there should be a policy that the publisher or builder of the agent has that says, okay, well, I'm going to let you call me 50 times, 100 times before you have to pay or something like that. We should have effectively like an audit trail, like, okay, this agent has been called this many times. We also have kind of human ratings and reviews right now, and we have tens of thousands of reviews of the existing agents on agent.ai. Average is like 4.1 out of five stars. And all those things are nice signals to be able to have. But the kind of callable... Verifiable kind of thing, I think, is super useful. Like, if I can just call... Give me an API that says here are five agents and it solves this particular problem for me. If I have like a simple eval, I think that'd be so powerful. I wish I had that for humans, honestly. That'd be so cool.Alessio [00:54:47]: Yeah, because, I mean, when I was running engineering teams, people would try and come up with these rubrics, you know, when hiring. And it's like, they're not really helpful, but you just kind of need some ground truth. But I feel like now, say you want to hire, yeah, an AI software engineer. Yep. You can literally generate like 15. 20 examples of like your actual issues in your organization, both from a people perspective of like collaboration and like actual code generation. Yep. And just pay for it to run it. Yeah. Like today we do take home projects and we pay people. Sure. Like this should be kind of the same thing. Yeah. It's like, I'll just run you. But I feel like people are not investing in their own evals as much internally.Dharmesh [00:55:22]: I mean, that's the present company included, right? Everyone talks about evals. Everyone accepts the fact that we should be doing more with evals. I won't say nobody, but almost nobody actually does. That's the... And yeah, it's a topic for a whole other day. I'm not...swyx [00:55:36]: It's funny, I mean, because obviously HubSpot is famous for launching graders of things. Yes. You'd be perfect for it. Yeah. Somehow. agree on evals, by the way. I mean, I just force myself to be the human in the loop or, you know, someone I work with and that's okay. But obviously the scalable thing needs to be done. Just a fun fact on, or question on the agent AI, you famously, you've already talked about the chat.com acquisition and all that. Yeah. And that was around the time of custom GPTs and the GPT store launching. Yes. And I definitely feel agent AI is kind of the GPT score, but not taken seriously. Yeah. Do you feel open AI if like they woke up one day and they were like, agent AI is the thing, like we should just reinvest in GPT store instead of fear?Dharmesh [00:56:20]: I think that won't be agent.ai driven. It's an inevitability that open AI, I don't have any insider information, I'm an investor, but no inside information is because it makes too much money. It makes too much sense for them not to like, and they, they've taken multiple passes at it, right? They did the plugins back in the day, then the custom GPTs and the GPT store because, you know, being the platform that they are, I think it's inevitable that they will ultimately come up with, and they already have custom, it's going to happen. I'm not on the list of things I promised myself I would never do is compete with Solomon Altman ever, not intentionally anyway. But here you are. But yeah, here I am.swyx [00:56:58]: But I'm not really, right? Not really. It's free, so like, whatever. But, you know, at some point, if it's actually valuable.Dharmesh [00:57:06]: They're solving a much, much bigger problem. I'm like a small, tiny rounding error in the universe. But the reason that compelled me to actually create in the first place, because I knew custom GPTs existed, I did have this rule in my head that don't compete with Sam. He's literally like at the top of my list of people not to compete with. He's so good. But the thing that I needed in terms of for my own personal use, which is how agent.ai got started, because I was building a bunch of what I call solo software. Things for my own personal productivity gain. And I found myself doing more and more kind of LM driven stuff because it was better that way. You know, I sort of showed up in those solo projects a bunch. And so the thing I needed was an underlying framework to kind of build these things. And high on the list was I want to be able to straddle models because certain steps in the thing is like, oh, for this particular thing involves writing. So maybe I want to use Claude for this particular thing. Maybe I want to do this even around image generation, different types of whether. It has texture, doesn't have texture, whatever. And I want to be able to mix and match. And my sense is that whether it's OpenAI or Anthropic or whatever, they're likely going to have an affinity for their own models, right? Which makes sense for them. But I can sort of be, for my own purposes and for our user base, a little bit of the Switzerland. It's like we don't think there's like one model to rule them all based on your use case. You're going to want to mix and match and maybe even change them out. Maybe even test them back to the kind of eval idea. It's like I have this agentic workflow. And here's the thing that we've been playing with recently. Because we have. We have enough users now where they, like the LM, and I look at the bills and it's like, oh, I'm spending real money now. And this is just human nature, right? It's not just normies, but it's like, so you have this drop down of all the models that you can say, which model do you want to use in your agent.ai agent? And as it turns out, people pick the largest number. So they will pick 4.5 or whatever, whatever it is, right? It's like it's.swyx [00:58:55]: Oh my God, you're doing 4.5? Yes.Dharmesh [00:58:57]: Ouch. Yes. Yeah. But the thing I've promised myself is we will support all of them, regardless of what it costs. And like, once again, I see this as a just a research thing, you know, benefit to humanity and inference costs are going down. At least I so I tell myself late at night so I can sleep. So they pick the highest numbered one. And so we have an option in there right now that says, which is the first option. It's like, let the system pick for me. Auto-optimist. Yeah. As it turns out, people don't do that. They just pick the, because they don't trust it yet, which is fine. They shouldn't trust it completely. But one thing we discovered is that if we back channel it, and this is the thing we're testing with, is that, oh, if I can just run the exact same agent that gets run a thousand times, we'll do it on our own internal agents first. And if the ratings and reviews, because we're getting human evals all the time on these agents, we can get a dramatic multiple orders of magnitude reduction by going to a lower model with literally like no change in the quality of the output. Right. Which makes sense. Because so many of the things we're doing doesn't require the most powerful model. And it's actually bad because there is higher latency. It's not just a cost thing. But so anyway, like in that kind of future state, I think we're going to have model routing and a whole body of people working on that problem, too. It's like, help me pick the best model at runtime. Would you buy or build model routing? I buy everything that I can buy. I don't want to build anything if I don't have to.swyx [01:00:26]: One of the most impressive examples of this. I think was our Chai AI conversation, which I think about a lot. He views himself explicitly as a marketplace. You are kind of a marketplace, but he has a third angle, which is the model providers, and he lets them compete. And I think that sort of Chai three-way marketplace maybe makes a lot of sense. Like, I don't know why every AI company isn't built that way. It's a good point, actually.Dharmesh [01:00:48]: Yeah, it makes sense. I have a list of things I'm super passionate about. I'm very passionate about efficient markets or extremely irritated by inefficient markets. And so efficient markets, for the normies listening, are markets that exist where every possible efficient markets are the ones that every transaction that should occur actually does. That's an efficient market that should happen. And so then why do inefficient markets exist? Well, maybe the buyer and seller don't know about each other. Maybe there's not enough of a trust mechanism. There's no way to actually price that or come up with fair market value for fair pricing. And as you kind of knock those dominoes down, the market becomes more and more. And lots of latent value exists as a result of inefficiency. And whoever removes those inefficiencies. Yeah. And then the market recedes for high value markets makes a lot of money. That's been proven time and time again. This is one of those examples of there's an inefficiency right now because we are either over using over models or whatever. Let's just reduce that to an efficient market. The right model should be matched up with the right use case for the right price. And then we'll... Very interesting. You ever looked into DSPy? I have looked at it. Not deeply enough, though.swyx [01:01:48]: It's supposed to be, as far as I think, the only evals first framework. Yep. And if evals are so important. And by the way, the relationship between this and all that is DSPy would also help you optimize your models. Yep. Because you did the evals first. Yep. I wonder why it's not as popular, you know. But I mean, it is growing in traction, I would say. We're keeping an eye on it.Alessio [01:02:09]: Let's talk about business models. Obviously, you have kind of two, work as a service and results as a service. Yep. I'm curious how you divide the two. Yeah.Dharmesh [01:02:19]: So work as a service is... So we know about software as a service, right? So I'm licensing software that's delivered to me as a service. That's been around for decades now. So we understand that. But the consumer of that service is generally a human that's doing the actual work, whichever software you're buying. Work as a service is the software is actually doing the work, whatever that work happens to be. And so that's work as a service. So I'll come up with kind of discrete use cases, whether it's kind of classification or legal contract review or whatever the software is actually doing the thing. Results as a service is you're actually charging for the outcome, not actually the work, right? That says, okay, instead of saying, I'm going to pay you X amount of dollars to review a legal contract or this amount of time or number of uses or something like that, I'm going to actually pay you for the actual result, which is... So my take on this in the industry or the parts of the industry are super excited about this kind of results as a service or outcomes-based pricing. And I think the reason for that, I think we're over-indexing on it. And the reason we're over-indexing on it is the most popular use case on the kind of agent side right now is like customer support. Well-documented. A lot of the providers that have agents for customer support do it on a number of tickets resolved times X dollars per ticket. And the reason that that makes a lot of sense is that the customer support departments and teams sort of already have a sense for what a ticket costs to resolve through their kind of current way. And so you can come up with an approximation for A, what the kind of economic value is. There's also at least a semi-objective measure for what an acceptable value is. And that's what an acceptable resolution or outcome is, right? Like you can say, oh, well, we measured the net promoter score or CSAT for tickets or whatever. As long as the customers, 90% of the tickets were handled in a way the customer was happy. That's whatever your kind of line is. As long as the AI is able to kind of replicate that same SLA, it's like, okay, well, it's the same. They're fungible, one versus the other. I think the reason we're over-indexed, though, is that there are not that many use cases that have those two dimensions to them that are objectively measurable. And that there's a known economic value that's constant. Like, customer support tickets, because they're handled by humans, make sense. And humans have a discrete cost. And especially in retail, which is where this originally got started in B2C companies that have a high volume of customer support tickets that they're distributing across, a ticket is roughly worth the same because it takes the same amount of time for most humans to do that kind of level one, tier one support. But in other things, the value per outcome can vary dramatically, literally by orders of magnitude, in terms of what the thing is actually worth. That's kind of thing number one. Thing number two is, how do you objectively evaluate that? How do you measure? So let's say you're going to do a logo creator as a service based on results, right? And that's a completely opposite subjective thing or whatever. And so, okay, well, it may take me 100 iterations. It may take me five iterations. The quality of the output is actually not completely under my control. It's not up to the software. It could be you have weird taste or you didn't describe what you're looking for enough or whatever. It's like it was just not a solvable problem. Design kind of qualitative, subjective disciplines deal with this all the time. How do you make for a happy customer? There's a reason why they have, oh, we'll go through five iterations. But our output is we're going to charge you $5,000 or $500 or whatever it is for this logo. But that's hard, right, to kind of do at scale.swyx [01:05:29]: Just a relatable anecdote. Our podcast, actually, we just got a new logo. And we did 99 designs for it. And there are so many designers who are working really hard. But I just didn't know what I wanted. So I was just too bad. You seem great, but you know.Dharmesh [01:05:48]: that's another example of a market made efficient, right? Yeah. It's like I've been a 99designs user and customer for a dozen plus years now.swyx [01:05:55]: It's fantastic. Yeah. So many designers, like this doesn't cost that much for them to do. It's worth a lot to us. We can't design for s**t. Totally. Yeah. Yep.Dharmesh [01:06:04]: By the way, pro tip on 99designs is that on the margin, you're better off kind of committing to paying the designer that you're going to pick a winner. Whether you like it or not doesn't really matter. And that gets higher participation. And you're still going to get a bunch of crap that happens. You get a bunch of noise in it. But the kind of quality outcome is often a function of the number of iterations. And logo design is one of those examples. If you had to choose between 200 logos versus 20 logos, chances are closer that you're going to find something you like. Yeah.swyx [01:06:33]: For those interested, I have a blog post on my reflections on the 99designs thing. And that's one of those. They give an estimate of how many designs you get. Yep. And I think that the modifier for like, we will pay you, we'll pay somebody and maybe it's you, is like 30 to 60. But actually it's 200. Yep. So it's underpriced. Yep.Alessio [01:06:51]: Yep. Do you think some markets are just fundamentally going to move to more results-driven business models? Probably.Dharmesh [01:06:59]: And I don't understand enough markets well enough to know. But if we had to kind of sort or rank them, there's likely some dimension along which we could sort that. It's like, oh, these kinds of businesses, is there an objective measure of kind of truth or the outcome? Is there a way to kind of price it in terms of the low variance or variability on the value? If those things are true, whatever industries that is true in, customer support is an example, but there's likely lots of other examples where those two things are true. But then the thing I wonder, though, is that from the customer's perspective, would they rather actually pay for work as a service versus an actual, it's like maybe the way they think about it is that's sort of my arbitrage opportunity. Like I can get work done for X, but the value is actually Y. Why would I want that delta to be squozed out by the kind of provider of the software if I have a choice? I don't know. Oh, I mean, okay.swyx [01:07:51]: Attribution. There's 18 things that go into them. You're one of them. So it's hard to tell. Yes, it is. By the way, have you seen, obviously you're in this industry, not exactly HubSpot's exact part of the market, but what have you seen in attribution that is interesting? Because that directly ties into work as a service versus results. Yeah.Dharmesh [01:08:12]: Not enough because we are so, as a world, as an industry, just pick your thing. So behind. Yeah. This is why I think Web3 in the way that it was meant to be done is going to make a comeback because fundamental principles of that makes sense. I think what happened in that world was kind of a bunch of crypto bros and grifters and NFT stuff or whatever that was loosely related. There was no actual, but the idea of a blockchain, of a trackable thing, of being able to fractionalize digital assets, attribution, having an audit log, a published thing that's verifiable. All those primitives make sense, right? And maybe there's a limited, but it's not zero, set of use cases where the kind of what we would now call like the inference cost or the overhead, the tax for storing data on the blockchain. And there's certainly a tax to it. It doesn't make sense for all things, but it makes sense for some things for sure. But we just don't have attribution in any meaningful way, I don't think. Isn't it sad that it's so important and no answer? I know. It partly comes down to incentives. Yeah. So people that actually have the data or parts of the data from which attribution could be calculated or derived don't really have the incentives to make that data available. So even something as simple like on the PPC side, right, on the Google search thing, that's sort of my world or has been. We have less data now than we did back in the day in terms of like click-throughs and things like that before Google would actually send you. Here are the keywords people typed. And years ago, they even took that away. So it's hard to kind of really connect the dots back on things. And we're seeing that across. It's not just PPC, but just all sorts of things. They took that away from the Search Console. What's that? The Search Console has that. Yes. They took that away. Search Console has that. But your website, if you go to Google Analytics, you can connect it back to the Google Search Console. I see. Yeah. Yes. Okay.swyx [01:10:00]: All right. Yeah. Well, it's a known thing. You don't have to make it a rant about Google.Alessio [01:10:06]: What about software engineering? Do you think it will stay as like a work as a service? Or do you think? I think most companies hire a lot of engineers, but they don't really know what to do with them or like they don't really use them productively. Yeah. And I think now they're kind of hitting this like, you know, crisis where it's like, okay, I don't know what I will price an agent because I don't really know what my people are doing anyway. Yeah. Like, how do you think that changes?Dharmesh [01:10:27]: I think, so I'm actually bullish on engineers in terms of their kind of long-term economic value. Not despite all the movements in Cogen and all the things that we're already seeing, but because of it. Because what's going to happen as a result of AI, and people have talked about this in even other disciplines, we're going to be able to solve many more problems. The semi-math guy in me is like, okay, so we always say, oh, well, now agents are going to be doing code or whatever. And so there's going to be a million software engineers, you know, virtual digital software engineers out there. And so the value per engineer is going to go down because I'm just in that same mix that I as an engineer. What they don't recognize is that it's not just about the denominator, there's a numerator as well, which is what's the total economic value that's possible. And I would argue that's growing faster than the kind of denominator is, that the actual economic value that's possible as a result of software and what engineers can produce, you know, with the tools that they will have at hand. So I think the value of an engineer actually goes up. They're going to have the power tools, they're going to be able to solve a larger base of problems that are going to need to be solved. Yeah.Alessio [01:11:29]: It feels to me like he'll stay as like work as a service. You're paying for work. I don't think there's like a way to do that.Dharmesh [01:11:34]: And there will be a set of engineers that, and we see this all the time, you know, they're like in the media industry, you have people that are kind of writers, but then you have freelancers that, you know, write articles or write however they manifest their kind of creative talent. And both make sense, right? There's like the work for hire. There's also the kind of outcome based or like I produce this thing. And maybe they, some of those engineers actually produce agents. So they put it in a marketplace like agent did AI someday, and that's how they make their millions. Yeah.Alessio [01:11:58]: Any other thoughts just on agents? We got a lot of like misc things that we want to talk to you about. Miscellaneous.Dharmesh [01:12:03]: I think we cover a lot of territory. So I'm excited about agents. My kind of message to the world. Yeah. Would be, don't be scared. I know it's scary. Easy for me to say as a tech techno optimist, but learn it. Even if you're a normie, even if you're not an engineer, if you're not an AI person, you'll think of yourself as an AI person. Use the tools. I don't care what role you have right now, where you are in the workforce. It will be useful to you and start to get to know agents, use them, build them.swyx [01:12:29]: And I think my message for engineers is always like, there's more to go. Like we're still in the early days of figuring out what an agent's stack looks like. Yeah. And I want to push people towards agents with memory. Yeah. Agents with planning.Dharmesh [01:12:43]: Oh, we have to talk about memory. We got to talk about memory. Let's go. Let's do it. Because I think that's the next, in my mind, the next frontier is actual long-term memory, both for agents and then for agentic networks and a trustable, verifiable, I won't say privacy first, but privacy oriented way. I have an issue with the term privacy first, because a lot of times we say privacy first, when we don't really mean that. Privacy first means I value that above all things. It doesn't matter what we're talking about. And that's just not true, not for any human. Anything that wants to be used. So memory is an interesting thing, right? So the thing I'm working on right now, lots of things in play in agent.ai is around implementation of memory. And there are great projects out there, mem0 being one of them. But the thing that's interesting for me, right, is, and so we see this in ChatGPT and other things right now, where it does have the notion of a longer term memory. You can pull things back into context as needed. The thing I'm fascinated by is cross-agent memory. So if I'm an agent builder right now, it's like, okay, here are the things that I sort of know or I learned from the user in terms of pulling out the, I'll call them knowledge nuggets, for lack of a better term. And that's great. But then when the next agent builder comes out and it's the same user, shouldn't all the things that agent one learned about me, if it's going to be useful for agent two, as long as I opt into it, it's like, yeah, I don't care those things. In fact, I would find it awfully annoying to tell agent two and agent n and agent n plus one, all the same things I've already told it, because it should know, like the system should know. And this is part of the reason why I'm like a believer in these kind of networks of agents and shared state is that that user utility gets created as a result of having shared memory. Not just we should solve the memory problem for an independent agent, but then we should also be able to share that context, share that memory across the system. And that's part of the value prop for agent.ai is like, okay, when you're building, it's like, so we've got, you know, whatever million users and we're going to have growing memory about all of them. So instead of you going off on your own thing and building an agent out as this kind of disconnected node in the universe or whatever, here's the value for building on the network or on the platform, ours or someone else's, because there's more user value that gets created. It's more utility.Alessio [01:14:59]: How do you think about auth for that? Because part of memory is like selective memory. So it takes like scheduling. Yep. I want you to have access. If I have another scheduling agent, you should be able to access the events you're a part of. Yep. And like what times I have available, but it shouldn't tell you about other events on my calendar. Like what's that like?Dharmesh [01:15:15]: I have so many thoughts on this. This is like the opportunity out there, like solving these kind of fundamental, like this is going to need to exist, right? So right now the closest approximation we have is auth, auth 2.0, right? And everyone has, it's like, okay, approve. And it's a very, very coarse set of scopes, right? Like based on the provider of the auth server, be it Google, whoever it is, HubSpot, it doesn't matter. It's like, oh, I pick a set of scopes and they could have defined the scopes to be super granular. Fine. But it's sort of up to them. But that is going to move so slowly, right? So for instance, the use case I have right now, like I use email for everything. I use it as a, like an event and data bus for my life, right? And why I mean that, like literally, it's like, I'm like anything that I do, if there's a way to kind of get that into email, because I know it's an open protocol, right? It's like, okay, I will be able to get to that data in useful ways. And this is before. So I have 3 million that I've built a vector store off of that has solved my own personal use cases. So I'll give you the example, but obviously I'm not going to build all my own software for everything. But if a startup comes along and says, Dharmesh, can you make your email inbox available in exchange for these things? I'm like, hell no. Like that's the, literally my kind of like everything, like my life is in here, right? So you need to share subsets. Yes. And so I think there's a, and maybe this is not the actual implementation, but imagine if someone said, okay, I have a trusted intermediary for that first trust, however defined that says, okay, I'm going to OAuth into this thing. And it gets to control that. I can say in natural language, I only want to pass email to this provider where the label is one of X or that's within the last thing and no more than 50 emails in a day or whatever. So I don't have them dumping the entire 3 million backlog, whatever controls I want to put on it. It's unlikely that the, all the OAuth server side right now, the Googles, even the big ones, small ones doesn't really matter. Are going to do that. But this is an opportunity for someone and they're going to need to get to some scale, build some level of trust that says, okay, I'm going to hand over the keys to this intermediary. Yeah. But then it opens up a bunch of utility because it gives me control, more fine, fine grainswyx [01:17:15]: control. Yeah. I'd say Langchain has, has an interesting one. There are a bunch of people who has tried to track crack AI email. Every single one of them who has tried has pivoted away. Yep. And I'm waiting for Superhuman to do it. Yep. I don't know why they haven't, but you know, at some point.Alessio [01:17:29]: They have some cool AI stuff. Yeah. Yeah. I think the pace needs to increase, but I think this goes back to like open graph. Yeah. Right. Which is like, I think Google is not incentivized to build better scopes. Nope. And like, they're just not going to do it. Nope. So.Dharmesh [01:17:42]: We can't even get like, we haven't been able to get semantic search out of Google for like, still. Not totally. Yeah. Just now they made the announcement this week. What do you mean? Semantic search? In Gmail. Oh, I see. Yeah. So, okay. So they have all the, they have my 3 million emails. Why don't they have a vector store where I can just like basic. Yeah. Yeah.Dharmesh [01:18:01]: In real time.swyx [01:18:03]: Like, I don't think my email is that big a deal, but. Yeah. My standard thing on memory is, it sounds like you are using mem0. I am. There's also memgpt, now Letta, which give a workshop at my conference. There's Zep, which uses a graph database, just kind of open source, kind of interesting. Yep. And LangMem from LangGraph, which I would highlight. Also, like it's really interesting, this developing philosophy that people seem to be agreeing on, on a hierarchy of memories. Mm-hmm. Mm-hmm. So, from memory to episodic memory to, I think it's just overall sort of background processing. Like, we have independently reinvented that AI should sleep. Yep. To do the deep REM processing of memories. Yep. It's kind of interesting. Yep.Dharmesh [01:18:43]: Yeah, that is. It's the other, I mean, just on the notion of memory and hierarchies. So, you know, I talked about the memory we're working on right now is at the user level and it's cross agent, right? Yeah. But the other kind of one step up would be, so once again, going back to this kind of hybrid digital teams. Yeah. Is that you can imagine to say, oh, well, my team has this kind of shared team. I don't want to share with the world or this set of agents across this group of people. I want to have shared state like we would have in a Slack channel or something like that. That should sort of exist as an option, right? Yeah. And the platforms should provide that.swyx [01:19:15]: And the B folks I should also mention have mentioned that they're working on that as well. Okay. So, imagine being able to share, you know, selective conversations with people. Like, that's nice. Yeah. Yeah. VerbalLess has, I guess, voice-based shielding. I don't think they have the action. I'm an investor in that too.Dharmesh [01:19:32]: Oh, really? Okay. Trying to think about all the things I've said, Invest in OpenAI, Perplexity, Langraph, Kru.ai, Limitless, a bunch of them. So, if I've said anything, by the way, I have no insider knowledge. I have no... I'm not trying to plug or pitch or anything like that. No, no, no.swyx [01:19:48]: I think it's understood. We're often... Like, you know, if you have skin in the game, you've probably invested or me or me not... I'm not an investor in B, but I'm just a friend. And I think you should be able to speak freely of your opinions regardless. Okay, we have some miscellaneous questions that may be zooming out from Agent AI. First of all, you mentioned this and I have to ask, you have so many AI projects you'll never get to. What's one or two that you want other people to work on?Dharmesh [01:20:15]: Oh, wow.swyx [01:20:16]: Drop some from your list.Dharmesh [01:20:18]: Other people to work on. Because you'll never get to it. Yeah, what I need to do is I've had this thought before. So I have this is like maybe like pick one a week or something like that and give the domain away. Like I have people submit their one pager or something like that. It's like, if you can convince me that you have at least enough of an idea, enough like willingness to kind of commit to actually doing something. It's the ones that you keep mentioning, but you haven't gotten to it for whatever reason. Yep, yep. Traffic, like some of them, I don't have the underlying business model. We're going to have to come back to this, maybe do a follow-up episode. I don't, like they're just not jumping to mine. You don't need the business model, just... Yeah, so I own Scout.ai. I think that's an interesting... By the way, pretty much all of them, there was an idea at the time. It's like it was one of those late night, it's like, oh, I could do this. Is the domain available? And I'll go grab it. I'm trying to think what else I have on the AI space. I have a lot of like non-profit domain names as well for like non-profit like OpenGraph. I'm not sure why things are not jumping to my head. Yeah, I have agent.com, which obviously is tied to agent.ai.swyx [01:21:24]: Oh, that's going to be big. That's going to be big. Oh my God. That's going to be like a 30, $50 million.Dharmesh [01:21:29]: It's going to be big. It's going to be, I think, end up being bigger than chat.com, which was 15.swyx [01:21:38]: Yeah, it's more work oriented. Yep. That's interesting.Alessio [01:21:41]: Yeah, do you want to talk about the chat.com thing? I would love just the backstories. Like, did you just call up Sam one day and be like, I got the domain? Yeah. Did they? Can I get back to you?Dharmesh [01:21:52]: No, I'll give you, it's a good story. Back in the original ChatGPT days, the first thought I had in my head, which lots of people had in their head, is that OpenAI is going to build a platform and ChatGPT is actually just a demo app to show off the thing. And there's been precedence for tech companies that have had, you know, demo apps to kind of help normies understand the underlying technology. And even after the kind of boost or whatever. So my original thought was, well, someone should actually create like an actual real world. And so I'm like, and that product should be called chat.com because GPT is not a consumer friendly thing at all. Like that's an acronym, not pretty, it doesn't roll off the tongue. And so like, I'll build ChatGPT because that was just a demo app back then. So I, you know, got chat.com. And then as it turns out, ChatGPT is like a real product. And I was at an event here in San Francisco that Sam spoke at where he launched plugins. I think it was the announcement at that time. Yeah. And that's the thing is like, I had sort of suspected, it's like, okay, things sort of be like, there's no way. There's no way that OpenAI is going to launch plugins for ChatGPT if they were not thinking of it as an actual platform. So it's not just about the GPT APIs. This is like a real thing. I'm like, crap. Like this violates the first rule of Dharmesh, which is don't compete with Sam. I knew when I bought the domain that there was competition for the domain. There were other companies looking to buy it. I don't know who they were. I had suspicions. So I bought it and then I'm like, okay, well, I'll reach out to Sam. I was like, hey, Sam, I happen to have got, I don't know. I don't know if he was or wasn't kind of in the running or trying to acquire it or not, but I have chat.com. I don't, not looking to make a profit or whatever. If you want it, you will obviously do something much better, bigger with it. I don't want to be in the compete with Sam game effectively is what I said. And so they did want it.swyx [01:23:38]: And yeah, we struck a deal. Looks like it's been a very good deal if the valuations are, you know, to be, to be real. Yeah. Who knows? Who knows?Alessio [01:23:48]: It's one of those weird things. Like, yeah. Yeah. The agent that AI domain evaluator said that late in that space is for between five and 15 K. Okay.swyx [01:23:55]: So does that feel right? Well, it's missed the, it's missing this one.Dharmesh [01:24:00]: Does not incorporate the transactional data. I have not published that one yet. Uh, that's because it's also operationally very intensive, uh, that other one. But anyway, we, we actually had it donated by a listener, so I don't know what the real cost is, but it's missing that it's linked to an influencer by way of AI, which I've offered. I'm an investor in, in, yes, I bought that. Uh, and I've told him that like, whenever you're ready, you let me know, I'll sell it to you at cost. Uh, yeah.swyx [01:24:25]: So, yeah, I mean, that, that is some value add since you may buy a lot of domains.Dharmesh [01:24:29]: What, what are your favorite, uh, domain buying tips apart from have a really good domain broker, which I assume you have, uh, no, I actually don't, uh, I do, I do my own deals. Um, I have a, like a very cards face up approach to life. Um, so there's, so, you know, some people would tell you, it's like, oh, well, if someone, they know that it's, you're behind the transaction. Yeah. So, you know, the price is going to go up, sure, but it's still like willing seller or willing buyer or whatever. It doesn't mean I'm going to have to necessarily pay that price. Uh, it's like, okay. But the upside to it, uh, cause I always, you know, reach out as myself when I'm, when there's a domain out there. Um, and they can look you up. They can look me up. But then I also come off as like legit, like, okay, well, there's very few people are not going to return my email. When I say I'm interested in a domain that they may have for sale, um, or had not considered selling, but you know, would you consider selling? Uh, so yeah. And some of the, like, uh. So I own some of my favorites. I still own prompt.com, by the way, that, that could be a big one. Um, but I owned, and this is one, uh, I don't regret it. I went into a good, I owned a playground.com. And so the original idea behind playground.com was at the time, uh, open AI had their, uh, playground where you can play around with the models and things like that. Right. It's like, okay, well, there should be a platform neutral thing. There should be a playground across all the LLMs. Then you can, and there are obviously products and, uh, startups that, that do that now. And so that was my original thing. It's like, oh, there should be playground.com and you can go test out all the models and play around with them just like you can with, uh, with open AI's, uh, GPT stuff. And then, uh, so sale was out there with, uh, with, with playground, uh, the company, uh, and I think he reached out, it might've reached out to me over, over Twitter or something like that. So we knew of, of each other. I'd never, I've still never, never met him. And he asked me whether I would consider, and that was a tough one because I'm like, I actually have the business idea already in my head. I think it's a great idea. I think it's a great domain name. Uh, and it's like a really simple English word that has like relevance and a whole new context now. But once again, uh, I took, uh, took equity. So it's like, uh, look on the bright side. That's like, I, so domains that get me into deals that I would never been able to likely get into two other ways. So, yeah.Alessio [01:26:35]: Yeah. We should securitize your GoDaddy account and just make it a fund. It's a fund.Dharmesh [01:26:41]: It's basically a fund. Yeah. Um, and by the way, and so back to the kind of, uh, I hope you don't use GoDaddy by the way. Vested, uh, I don't know if it's public yet. Um, but in a company that's going to treat domains as a fractionalizable, uh, tradable asset, because that's the kind of the original NFT in a way, right? It's like, okay, well, and then if you can make both fractionalizing, but also just to transfer, like right now, it's so painful when you buy a domain, you go through an escrow service and there's just all of this. It's like, I just want like instantaneous, like charge me in Bitcoin or credit card, whatever it is. And then I should show up and I should be able to reroute the DNS. Like that should be minutes, not weeks or days. Um, anyway, so.Alessio [01:27:19]: Yeah, that's what ENS on Ethereum is basically the same, but it should bring that for normies. Yeah, exactly. They should bring it. Yeah. The ICANN and all of that is, uh, as its own, its own thing.swyx [01:27:30]: I have a question on, on just, uh, you know, you keep bringing up your Sam Altman rule. One of my favorite, favorite, favorite, my first millions of all time was actually without you there, but talking about you. Okay. Cause, uh, Sean was describing you as a fierce nerd, which I'm sure you, you, you were there. Uh, um, and, uh, I think Sam also is a fierce nerd and, and he is, uh, uh, I was, I was listening to this Jessica Livingston podcast where what she had him on and described him as a formidable person. I think you're also very formidable and I just wonder what makes you formidable. What makes you a fierce nerd? What, what keeps you this driven? Yeah.Dharmesh [01:28:09]: Sam's fiercer and nerdier just for the record. Um, but I think part of it is just like the strength of my conviction, I guess. Like I'm, I'm willing to. Work harder and grind it out, uh, more than people that are smarter than me. And I'm only slightly stupider than people that are willing to work harder than me. Right. Like I'm just the right mix of, uh, the kind of grinded it, kind of work at it, stick to it for extended periods of time. If I think I'm right, I will latch out, latch on and not let go until I can either like prove to myself that it's not. Um, so even like the natural language thing, it's like, you know, it took 20 years, but eventually I got to a point where, uh, the world caught up and it became possible. Uh, but yeah, I think. And part of it is, uh, I think this is partly, I think what makes me like, I'm a nice guy. Uh, sometimes they're the most dangerous kind, right? It's like, okay, well, I, I don't make enemies or whatever, but so my advice would be my, this is my take on competition. I don't think of it as like war. I think of it as, uh, their opponents. All right. And this is, it's not worried up. It's like, it's, it's a game, right? And you can use whatever analogy I happen to play a fair amount of chess. I'm a student of the game. That's partly, I think what, uh, makes me. Effective, uh, I'm solving for the long-term, uh, so I'm kind of hard to deter. So for those of you out there looking to kind of compete with HubSpot, uh, no, uh, I'm going to be here 18 years. I'm going to be here for another 18 years. So, but not that you shouldn't do it. It's a big market.swyx [01:29:34]: Uh, I'm not trying to sway anyone, but yeah, I think like something I struggled with, with this conviction, you said you pursue things to conviction, but like you start out not knowing anything. Yeah. And so how do you develop a conviction when there's. You, you find it along the way, or you, you stumble along the way, then you lose conviction and then you stop working on it, you know, like, how do you keep going?Dharmesh [01:29:57]: The way I've sort of approached it is that, um, so I don't generally tend to have conviction around a solution or a product. I have conviction around a problem, uh, that says this is an actual real problem that needs to be solved. And I may have an idea for how to be solved, uh, you know, right now, and that I may get dissuaded. It's like, ah, I'm not smart enough. Technology's not good enough. Whatever the constraints are, but it's the problem I have conviction around. It's like, oh, that problem still hasn't gone away. Uh, so like I sort of filed away in the back of my brain and I'll revisit it's like, okay, well, you know, the kind of board changes, uh, and then it changes really fast now with AI, like things that weren't possible before are now possible. So you kind of go back to your roster of things that you believe or believed and say, maybe now, uh, now is the time maybe then it wasn't the time, uh, but I'm a big believer in kind of attaching yourself. Passionately, uh, with conviction to problems that matter, um, that, and there are some that are just too highfalutin for me that I'm not going to ever be able to kind of take on. I have the humility to recognize that. Yeah.swyx [01:30:59]: I feel like I need a, um, updated founder's version of a serenity prayer. Like give me the confidence to like do what I think I I'm capable of, but like not to overestimate myself, you know? Uh, you know, anyway, uh, when you say board changes, how do you keep up on AI? A lot of YouTube, as it turns out. Yeah, a lot. Um, okay. Fireship. I don't know what fireship is. It's a current meme right now. Whenever OpenAI drops something, you know, they love this, like live streams of, of stuff from on the OpenAI channel. The top comment is always, I will wait for the fireship video because fireship just summarizes their thing in five minutes.Dharmesh [01:31:35]: No, I, so my kind of MO, so I, by the way, I keep very weird hours. Uh, so my average go to bedtime, uh, is roughly 2 AM. Oh boy. But I do get average seven, seven and a half hours in. Uh, I don't, I don't use alarm clocks cause I don't, I don't, uh, have meetings, uh, uh, in the morning at all, uh, or try not to at least, uh, so my late night thing is, uh, is I'll watch probably like a couple of hours of YouTube videos off in the background while I'm coding. Um,swyx [01:32:04]: that's how you've seen our talks.Dharmesh [01:32:06]: I have. Yeah, I've seen. Yeah. Okay.swyx [01:32:08]: Yep.Dharmesh [01:32:09]: , um, and so I, and there's so much good material out there and the, and the thing I love about kind of YouTube and this, by the way, in terms of like use cases and things that agents that should exist that, uh, don't yet, I would love to, uh, technology exists now to build this is to be able to take a YouTube video of like a talk about, let's say on Latent Space or not, uh, but on the, um, AI engineer event and say, just pull the slides out for me, uh, cause I want to put it into a deck for use or whatever, some form of, uh, kind of distillation or translation into a different, uh, different format. Oh, I see. Cool slides. Got it. Pull the slides out of a video. Um, so I think that's interesting. I have, yeah. So by the way, on the kind of agent.ai thing, like one of the commonly used, uh, actions, uh, primitives that we have is the ability to kind of get a transcript from a video. And that seems like such a trivial thing or whatever, but it's like, like, if you don't know how to do it programmatically or whatever, if you're just a normie, it's like, okay, well I know it's there, but I can copy it and paste it. But like, how do I actually like get to the, the transcript for you and then, uh, getting to the transcript and then being able to encode it and say, I can. Actually. Uh, give you timestamps. So if you have a use case that says, oh, I want to know exactly when this was, I want to create an aggregate video clip. This was the actual original, um, agent that I built for my wife that she wanted to pull multiple clips together without using video editing softwares. Cause she wanted to have this, uh, aggregate thing. Uh, she's on the nonprofit side to like send to a friend.swyx [01:33:27]: Uh, anyway, there are video understanding models that have come out from meta, but the easiest one by far is going to be Gemini. They just launched YouTube support. Yep.Dharmesh [01:33:36]: So, um, they're doing good work over there. By the way, in terms of. The coolest thing AI wise recently, I'll say last week to 10 days has been the new, um, image model, Gemini flash, experimental, whatever they call it, uh, because it lets you effectively do editing, um, and just, and so, you know, my son is doing a eighth grade research project on AI image generation, right? So he's kind of gone deep on, uh, stable diffusion in the algorithms and things like that. I don't know much about it, but one thing I do know, I know enough about stable diffusion to know why editing is like near impossible that you can't recreate. Because it's like, you can't go back that way. It's going to be a different thing because it's sort of spinning the roulette wheel another time. The next time you try to, you know, a similar prompt. And so the fact that they were able to pull it off, it's still, it's still a very much a V one because you know, if you, I, you know, one of the test case, like, Oh, take the HubSpot logo and replace the, Oh, which is like this kind of sprocket with a donut and it will do it, but it won't size it to the degree that will actually fit into the actual thing. It's like, okay. Um, but yeah, but that's where it's headed.swyx [01:34:36]: Do you know the backstory behind that one? No. Uh, mostly. Most of Mustafa, who was part of, so they had image generation in Lama three, uh, lawyers didn't approve it. Mustafa quit meta and joined Gemini and didn't shift it. Uh, and it is rumored. And that's all I can say is that they got rid of diffusion. They, they, they did auto-aggressive image generation. And I think it's been interesting, these two worlds colliding because diffusion was really about the images and auto-aggressive was really about languages and people were kind of seeing like, how are they going to merge? And. And on the mid-journey side, David Holtz was very much betting on text diffusion being, uh, being their path forward. Uh, but it seems like the auto-aggressive paradigm is one like next token isDharmesh [01:35:17]: So Hill and playground are doing like exceptional work on that kind of domain of, uh, I don't know if it's auto-aggressive, but around kind of image editing and not just the kind of text to image and actually building like a UI for like a Photoshop kind of thing for actual generation of images versus, uh, just doing text. It's fascinating.swyx [01:35:32]: I just thought diffusion was kind of dead. Like there wasn't that much, it was just like bigger models. You know, higher detail and now auto-aggressive come along and now like the whole field is open. Yeah. Um, and I think like, if there was any real threat to like Photoshop or Canva, it's this thing. Yeah.Alessio [01:35:47]: Just to wrap up the conversation, you have a great post called, sorry, you must pass, which if I did the math right, you first wrote in 2007, the first version, and then you re-updated it post COVID, you mentioned you made a lot of changes to your schedule and your life based on the pandemic. How do you make decisions today? You know, in the, as anything changed, like since you, because you updated this in 2022 and I think now we're kind of like, you know, five years removed from COVID and all of that. I'm curious if you made any changes. Yeah.Dharmesh [01:36:17]: So the, so that post, sorry, must pass was the issue that happened, um, is my schedule just, and life just got overwhelmed. Right. It's like, it's just, I just, uh, too many kind of dots and connections and I love interacting with new people online. I love ideas. I love startups. There's. But as it turns out, uh, every time you say yes to anything, uh, you are by definition saying no to something else. Um, this, uh, you know, despite my best app, you know, attempts to change the laws of the universe, uh, I have not been able to do that. So that post was a reaction to that because what would happen for me, uh, would be when I did say no, I would feel this guilt because it's like, okay, well, whatever happened to me, it's like, oh, can you spend 15 minutes and just review this startup idea or whatever? It's like, uh, and sometimes it would like be someone that was second degree removed, like intro through a friend or something like that. Yeah. And I felt, uh, you know, real guilt. And so this was a very kind of honest, vulnerable, here's what's going on in my life. So, so this is not a judgment on you at all, whatever your project or whatever your thing you're working on, but I have sort of come to this realization that I just can't do it. So I'm sorry, but I, so my default thing right now, and lots of people will disagree with this kind of default position is that I have to pass because unless, and Derek Sivers said this really well, it's like either a hell yes or it's a no, right? So, and I'm going to, there's going to be a limited number of the, the hell yeses, um, that I'm going to be able to kind of inject into this. Um, so yeah, that, and that's of all the blog posts I've ever written, that has been the most useful for me. So I, um, and so, and I send it and I still send it out personally, right? I don't have a, I don't automate my email responses at all yet. Um, don't do automated social media posts. Um, but yeah, that one's been very, and I, so I encourage everyone wherever your line happens to be. I think this, um, lots of people have this guilt issue and that's one of the most unproductive emotions, uh, in, in human psychology. It's like no good comes from guilt. Not really. And unless you're like a sociopath or something like that, um, maybe you need, um, anyway, you don't need more guilt.swyx [01:38:14]: I would also say, so I, um, I would just encourage people to blog more because a lot of times people want like to pick your brain and then they ask you the same five questions that everyone else has asked. So if you blogged it, then you can just hear.Dharmesh [01:38:26]: So one of the things I'm working on, uh, and there are startups that are working on this as well. Uh, but I started before then is like a Dharmesh.ai, right? That's just captures. Yeah. And it's interesting. So that's one of the agents, um, on, on agent.ai, uh, on the underlying platform. Oh, there, there's a Dharmesh.ai? It's out there. It's Dharmesh.ai. Yeah. Nice. It's pure text space. No video, no audio right now. Um, but, uh, the, the thing that's like, I found it useful in terms of just the, how, how do I give it knowledge? So I have a kind of a private email address because a lot of the interactions that I will have, or if I do answer questions, because I, the other thing I, by the way, I don't do any phone calls like at all. Even like. No Zooms. Like at all. I mean, I'll get on Zooms with teams, but no one-on-one meetings, no one-on-one, uh, it just doesn't scale. So I've moved as much as possible to an async world. It's like, I will, as long as I can control the schedule, like I will take 20 minutes and write a thoughtful response, but I reserve the right, uh, anonymously with no attribution to kind of share that, uh, either with my model or with the world, um, you know, through a blog post or something. But it's been like useful because, uh, now that I have that kind of email backlog, I can go back and say, okay, I'm going to try to answer this question. Go through the vector store. Uh, and it's shockingly good. Uh, and I'm still irritated that Gmail doesn't do that out of the box. It's like they're in Google. Um, I think it's, it's gotta be coming now. It's there. I think they're finally, uh, the giant has been woken up. I think they're, uh, they're kind of, it's gotten faster now.swyx [01:39:45]: You know, it's one of the biggest giants in the world ever. Yeah. So, yeah. When I first told Alessio, you know, you were one of our dream guests. I never, I never expected, actually expected to book you because of, sorry, my spouse. So we were just like, ah, let's send an email. And then like, he'll say no and we'll move on with all day. Uh, so I just have to say like, uh, yeah, we're very honored.Dharmesh [01:40:05]: Oh, I'm just thrilled to be here. A huge fan of first time, first time guest, but, uh, yeah. Thank you for all that you do for the, for the community. I, I, I speak for a lot of them. You guys taught me a lot of, uh, what I think I know. So, uh, yeah.swyx [01:40:20]: Appreciate it. Yeah. I mean, uh, I am explicitly inspired by, by, um, by HubSpot. Oh, thank you. Inbound marketing. Uh, I think it's a stroke of genius and like the. The AI engineering is explicitly modeled after that. So like you created your own industry, you know, subsection of an industry that became a huge thing because you got the trend, right. And that's what AI engineering is supposed to be if we get it right. Um, how do we screw this up? How do we square what up? How, how do I screw this up? How do we screw AI engineering up?Dharmesh [01:40:47]: Oh, um, you know, yeah, the common failure modes, right. Is, um, so the original thing that makes inbound marketing work, the kind of kernel of the idea was to kind of, uh, to solve for the customer, solve for the audience, solve for the other side, uh, because the thing that was broken about marketing was marketing was a very self-centered, I have this budget. I'm going to blast you and interrupt your life and interrupt your day. And because I want you to buy this thing from me, right. And inbound marketing was the exact opposite. It's like use whatever limited budget you have and put something useful in the world that your target customer, uh, whoever it happens to be, will find valuable. Um, anyway, so the, the common failure mode is, um, is that you lose that, uh, I don't think you will, but it's very, very common, right? It's like, ah, like now I'm just going to like turn the crank and squeeze it just a little bit more like it's, uh, but you, you, the right reason, I think, uh, folks like me, uh, you know, appreciate that community so much is used you to have that genuine want to act. And there's nothing wrong with making money. There's nothing wrong with having spot, none of that, but at the, at the core of it, it's like, we want to lift the overall level of awareness for this group of people and create value and create goodness in the world. Um, I think if you hold onto that over the fullness of time, uh, the market becomes more efficient rewards. Yeah. Uh, that generosity, uh, that's my kind of fundamental life belief. So I think you guys are doing well. Thank you for your help and support. Yeah. My pleasure. Yeah.Alessio [01:42:06]: And just to wrap in very Dharmesh fashion, you have a URL for the Sorry Must Pass blog, which is sorrymustpass.org. So yeah, I thought that was a good, good nugget. Um, yeah, thanks so much for coming on. Oh, thanks. Thanks for having me. Get full access to Latent.Space at www.latent.space/subscribe

  --------  

  1:38:24
• Building Snipd: The AI Podcast App for Learning

  We are working with Amplify on the 2025 State of AI Engineering Survey to be presented at the AIE World’s Fair in SF! Join the survey to shape the future of AI Eng!We first met Snipd (affiliate link! we get a free month, you get a free month. but this is not a sponsored pod, we’ve never done one) over a year ago, and were immediately impressed by the design, but were doubtful about the behavior of snipping as the title behavior:Podcast apps are enormously sticky - Spotify spent almost $1b in podcast acquisitions and exclusive content just to get an 8% bump in market share among normies.However, after a disappointing Overcast 2.0 rewrite with no AI features in the last 3 years, I finally bit the bullet and switched to Snipd. It’s 2025, your podcast app should be able to let you search transcripts of your podcasts. Snipd is the best implementation of this so far.And yet they keep shipping:What impressed us wasn’t just how this tiny team of 4 was able to bootstrap a consumer AI app against massive titans and do so well; but also how seriously they think about learning through podcasts and improving retention of knowledge over time, aka “Duolingo for podcasts”. As an educational AI podcast, that’s a mission we can get behind.Full Video PodFind us on YouTube! This was the first pod we’ve ever shot outdoors!Show Notes* How does Shazam work?* Flutter/FlutterFlow* wav2vec paper* Perplexity Online LLM* Google Search Grounding* Comparing Snipd transcription with our Bee episode* NIPS 2017 Flo Rida* Gustav Söderström - Background AudioTimestamps* [00:00:03] Takeaways from AI Engineer NYC* [00:00:17] Weather in New York.* [00:00:26] Swyx and Snipd.* [00:01:01] Kevin's AI summit experience.* [00:01:31] Zurich and AI.* [00:03:25] SigLIP authors join OpenAI.* [00:03:39] Zurich is very costly.* [00:04:06] The Snipd origin story.* [00:05:24] Introduction to machine learning.* [00:09:28] Snipd and user knowledge extraction.* [00:13:48] App's tech stack, Flutter, Python.* [00:15:11] How speakers are identified.* [00:18:29] The concept of "backgroundable" video.* [00:29:05] Voice cloning technology.* [00:31:03] Using AI agents.* [00:34:32] Snipd's future is multi-modal AI.* [00:36:37] Snipd and existing user behaviour.* [00:42:10] The app, summary, and timestamps.* [00:55:25] The future of AI and podcasting.* [1:14:55] Voice AITranscriptswyx [00:00:03]: Hey, I'm here in New York with Kevin Ben-Smith of Snipd. Welcome.Kevin [00:00:07]: Hi. Hi. Amazing to be here.swyx [00:00:09]: Yeah. This is our first ever, I think, outdoors podcast recording.Kevin [00:00:14]: It's quite a location for the first time, I have to say.swyx [00:00:18]: I was actually unsure because, you know, it's cold. It's like, I checked the temperature. It's like kind of one degree Celsius, but it's not that bad with the sun. No, it's quite nice. Yeah. Especially with our beautiful tea. With the tea. Yeah. Perfect. We're going to talk about Snips. I'm a Snips user. I'm a Snips user. I had to basically, you know, apart from Twitter, it's like the number one use app on my phone. Nice. When I wake up in the morning, I open Snips and I, you know, see what's new. And I think in terms of time spent or usage on my phone, I think it's number one or number two. Nice. Nice. So I really had to talk about it also because I think people interested in AI want to think about like, how can we, we're an AI podcast, we have to talk about the AI podcast app. But before we get there, we just finished. We just finished the AI Engineer Summit and you came for the two days. How was it?Kevin [00:01:07]: It was quite incredible. I mean, for me, the most valuable was just being in the same room with like-minded people who are building the future and who are seeing the future. You know, especially when it comes to AI agents, it's so often I have conversations with friends who are not in the AI world. And it's like so quickly it happens that you, it sounds like you're talking in science fiction. And it's just crazy talk. It was, you know, it's so refreshing to talk with so many other people who already see these things and yeah, be inspired then by them and not always feel like, like, okay, I think I'm just crazy. And like, this will never happen. It really is happening. And for me, it was very valuable. So day two, more relevant, more relevant for you than day one. Yeah. Day two. So day two was the engineering track. Yeah. That was definitely the most valuable for me. Like also as a producer. Practitioner myself, especially there were one or two talks that had to do with voice AI and AI agents with voice. Okay. So that was quite fascinating. Also spoke with the speakers afterwards. Yeah. And yeah, they were also very open and, and, you know, this, this sharing attitudes that's, I think in general, quite prevalent in the AI community. I also learned a lot, like really practical things that I can now take away with me. Yeah.swyx [00:02:25]: I mean, on my side, I, I think I watched only like half of the talks. Cause I was running around and I think people saw me like towards the end, I was kind of collapsing. I was on the floor, like, uh, towards the end because I, I needed to get, to get a rest, but yeah, I'm excited to watch the voice AI talks myself.Kevin [00:02:43]: Yeah. Yeah. Do that. And I mean, from my side, thanks a lot for organizing this conference for bringing everyone together. Do you have anything like this in Switzerland? The short answer is no. Um, I mean, I have to say the AI community in, especially Zurich, where. Yeah. Where we're, where we're based. Yeah. It is quite good. And it's growing, uh, especially driven by ETH, the, the technical university there and all of the big companies, they have AI teams there. Google, like Google has the biggest tech hub outside of the U S in Zurich. Yeah. Facebook is doing a lot in reality labs. Uh, Apple has a secret AI team, open AI and then SwapBit just announced that they're coming to Zurich. Yeah. Um, so there's a lot happening. Yeah.swyx [00:03:23]: So, yeah, uh, I think the most recent notable move, I think the entire vision team from Google. Uh, Lucas buyer, um, and, and all the other authors of Siglip left Google to join open AI, which I thought was like, it's like a big move for a whole team to move all at once at the same time. So I've been to Zurich and it just feels expensive. Like it's a great city. Yeah. It's great university, but I don't see it as like a business hub. Is it a business hub? I guess it is. Right.Kevin [00:03:51]: Like it's kind of, well, historically it's, uh, it's a finance hub, finance hub. Yeah. I mean, there are some, some large banks there, right? Especially UBS, uh, the, the largest wealth manager in the world, but it's really becoming more of a tech hub now with all of the big, uh, tech companies there.swyx [00:04:08]: I guess. Yeah. Yeah. And, but we, and research wise, it's all ETH. Yeah. There's some other things. Yeah. Yeah. Yeah.Kevin [00:04:13]: It's all driven by ETH. And then, uh, it's sister university EPFL, which is in Lausanne. Okay. Um, which they're also doing a lot, but, uh, it's, it's, it's really ETH. Uh, and otherwise, no, I mean, it's a beautiful, really beautiful city. I can recommend. To anyone. To come, uh, visit Zurich, uh, uh, let me know, happy to show you around and of course, you know, you, you have the nature so close, you have the mountains so close, you have so, so beautiful lakes. Yeah. Um, I think that's what makes it such a livable city. Yeah.swyx [00:04:42]: Um, and the cost is not, it's not cheap, but I mean, we're in New York city right now and, uh, I don't know, I paid $8 for a coffee this morning, so, uh, the coffee is cheaper in Zurich than the New York city. Okay. Okay. Let's talk about Snipt. What is Snipt and, you know, then we'll talk about your origin story, but I just, let's, let's get a crisp, what is Snipt? Yeah.Kevin [00:05:03]: I always see two definitions of Snipt, so I'll give you one really simple, straightforward one, and then a second more nuanced, um, which I think will be valuable for the rest of our conversation. So the most simple one is just to say, look, we're an AI powered podcast app. So if you listen to podcasts, we're now providing this AI enhanced experience. But if you look at the more nuanced, uh, podcast. Uh, perspective, it's actually, we, we've have a very big focus on people who like your audience who listened to podcasts to learn something new. Like your audience, you want, they want to learn about AI, what's happening, what's, what's, what's the latest research, what's going on. And we want to provide a, a spoken audio platform where you can do that most effectively. And AI is basically the way that we can achieve that. Yeah.swyx [00:05:53]: Means to an end. Yeah, exactly. When you started. Was it always meant to be AI or is it, was it more about the social sharing?Kevin [00:05:59]: So the first version that we ever released was like three and a half years ago. Okay. Yeah. So this was before ChatGPT. Before Whisper. Yeah. Before Whisper. Yeah. So I think a lot of the features that we now have in the app, they weren't really possible yet back then. But we already from the beginning, we always had the focus on knowledge. That's the reason why, you know, we in our team, why we listen to podcasts, but we did have a bit of a different approach. Like the idea in the very beginning was, so the name is Snips and you can create these, what we call Snips, which is basically a small snippet, like a clip from a, from a podcast. And we did envision sort of like a, like a social TikTok platform where some people would listen to full episodes and they would snip certain, like the best parts of it. And they would post that in a feed and other users would consume this feed of Snips. And use that as a discovery tool or just as a means to an end. And yeah, so you would have both people who create Snips and people who listen to Snips. So our big hypothesis in the beginning was, you know, it will be easy to get people to listen to these Snips, but super difficult to actually get them to create them. So we focused a lot of, a lot of our effort on making it as seamless and easy as possible to create a Snip. Yeah.swyx [00:07:17]: It's similar to TikTok. You need CapCut for there to be videos on TikTok. Exactly.Kevin [00:07:23]: And so for, for Snips, basically whenever you hear an amazing insight, a great moment, you can just triple tap your headphones. And our AI actually then saves the moment that you just listened to and summarizes it to create a note. And this is then basically a Snip. So yeah, we built, we built all of this, launched it. And what we found out was basically the exact opposite. So we saw that people use the Snips to discover podcasts, but they really, you know, they don't. You know, really love listening to long form podcasts, but they were creating Snips like crazy. And this was, this was definitely one of these aha moments when we realized like, hey, we should be really doubling down on the knowledge of learning of, yeah, helping you learn most effectively and helping you capture the knowledge that you listen to and actually do something with it. Because this is in general, you know, we, we live in this world where there's so much content and we consume and consume and consume. And it's so easy to just at the end of the podcast. You just start listening to the next podcast. And five minutes later, you've forgotten everything. 90%, 99% of what you've actually just learned. Yeah.swyx [00:08:31]: You don't know this, but, and most people don't know this, but this is my fourth podcast. My third podcast was a personal mixtape podcast where I Snipped manually sections of podcasts that I liked and added my own commentary on top of them and published them as small episodes. Nice. So those would be maybe five to 10 minute Snips. Yeah. And then I added something that I thought was a good story or like a good insight. And then I added my own commentary and published it as a separate podcast. It's cool. Is that still live? It's still live, but it's not active, but you can go back and find it. If you're, if, if you're curious enough, you'll see it. Nice. Yeah. You have to show me later. It was so manual because basically what my process would be, I hear something interesting. I note down the timestamp and I note down the URL of the podcast. I used to use Overcast. So it would just link to the Overcast page. And then. Put in my note taking app, go home. Whenever I feel like publishing, I will take one of those things and then download the MP3, clip out the MP3 and record my intro, outro and then publish it as a, as a podcast. But now Snips, I mean, I can just kind of double click or triple tap.Kevin [00:09:39]: I mean, those are very similar stories to what we hear from our users. You know, it's, it's normal that you're doing, you're doing something else while you're listening to a podcast. Yeah. A lot of our users, they're driving, they're working out, walking their dog. So in those moments when you hear something amazing, it's difficult to just write them down or, you know, you have to take out your phone. Some people take a screenshot, write down a timestamp, and then later on you have to go back and try to find it again. Of course you can't find it anymore because there's no search. There's no command F. And, um, these, these were all of the issues that, that, that we encountered also ourselves as users. And given that our background was in AI, we realized like, wait, hey, this is. This should not be the case. Like podcast apps today, they're still, they're basically repurposed music players, but we actually look at podcasts as one of the largest sources of knowledge in the world. And once you have that different angle of looking at it together with everything that AI is now enabling, you realize like, hey, this is not the way that we, that podcast apps should be. Yeah.swyx [00:10:41]: Yeah. I agree. You mentioned something that you said your background is in AI. Well, first of all, who's the team and what do you mean your background is in AI?Kevin [00:10:48]: Those are two very different things. I'm going to ask some questions. Yeah. Um, maybe starting with, with my backstory. Yeah. My backstory actually goes back, like, let's say 12 years ago or something like that. I moved to Zurich to study at ETH and actually I studied something completely different. I studied mathematics and economics basically with this specialization for quant finance. Same. Okay. Wow. All right. So yeah. And then as you know, all of these mathematical models for, um, asset pricing, derivative pricing, quantitative trading. And for me, the thing that, that fascinates me the most was the mathematical modeling behind it. Uh, mathematics, uh, statistics, but I was never really that passionate about the finance side of things.swyx [00:11:32]: Oh really? Oh, okay. Yeah. I mean, we're different there.Kevin [00:11:36]: I mean, one just, let's say symptom that I noticed now, like, like looking back during that time. Yeah. I think I never read an academic paper about the subject in my free time. And then it was towards the end of my studies. I was already working for a big bank. One of my best friends, he comes to me and says, Hey, I just took this course. You have to, you have to do this. You have to take this lecture. Okay. And I'm like, what, what, what is it about? It's called machine learning and I'm like, what, what, what kind of stupid name is that? Uh, so you sent me the slides and like over a weekend I went through all of the slides and I just, I just knew like freaking hell. Like this is it. I'm, I'm in love. Wow. Yeah. Okay. And that was then over the course of the next, I think like 12 months, I just really got into it. Started reading all about it, like reading blog posts, starting building my own models.swyx [00:12:26]: Was this course by a famous person, famous university? Was it like the Andrew Wayne Coursera thing? No.Kevin [00:12:31]: So this was a ETH course. So a professor at ETH. Did he teach in English by the way? Yeah. Okay.swyx [00:12:37]: So these slides are somewhere available. Yeah. Definitely. I mean, now they're quite outdated. Yeah. Sure. Well, I think, you know, reflecting on the finance thing for a bit. So I, I was, used to be a trader, uh, sell side and buy side. I was options trader first and then I was more like a quantitative hedge fund analyst. We never really use machine learning. It was more like a little bit of statistical modeling, but really like you, you fit, you know, your regression.Kevin [00:13:03]: No, I mean, that's, that's what it is. And, uh, or you, you solve partial differential equations and have then numerical methods to, to, to solve these. That's, that's for you. That's your degree. And that's, that's not really what you do at work. Right. Unless, well, I don't know what you do at work. In my job. No, no, we weren't solving the partial differential. Yeah.swyx [00:13:18]: You learn all this in school and then you don't use it.Kevin [00:13:20]: I mean, we, we, well, let's put it like that. Um, in some things, yeah, I mean, I did code algorithms that would do it, but it was basically like, it was the most basic algorithms and then you just like slightly improve them a little bit. Like you just tweak them here and there. Yeah. It wasn't like starting from scratch, like, Oh, here's this new partial differential equation. How do we know?swyx [00:13:43]: Yeah. Yeah. I mean, that's, that's real life, right? Most, most of it's kind of boring or you're, you're using established things because they're established because, uh, they tackle the most important topics. Um, yeah. Portfolio management was more interesting for me. Um, and, uh, we, we were sort of the first to combine like social data with, with quantitative trading. And I think, uh, I think now it's very common, but, um, yeah. Anyway, then you, you went, you went deep on machine learning and then what? You quit your job? Yeah. Yeah. Wow.Kevin [00:14:12]: I quit my job because, uh, um, I mean, I started using it at the bank as well. Like try, like, you know, I like desperately tried to find any kind of excuse to like use it here or there, but it just was clear to me, like, no, if I want to do this, um, like I just have to like make a real cut. So I quit my job and joined an early stage, uh, tech startup in Zurich where then built up the AI team over five years. Wow. Yeah. So yeah, we built various machine learning, uh, things for, for banks from like models for, for sales teams to identify which clients like which product to sell to them and with what reasons all the way to, we did a lot, a lot with bank transactions. One of the actually most fun projects for me was we had an, an NLP model that would take the booking text of a transaction, like a credit card transaction and pretty fired. Yeah. Because it had all of these, you know, like numbers in there and abbreviations and whatnot. And sometimes you look at it like, what, what is this? And it was just, you know, it would just change it to, I don't know, CVS. Yeah.swyx [00:15:15]: Yeah. But I mean, would you have hallucinations?Kevin [00:15:17]: No, no, no. The way that everything was set up, it wasn't like, it wasn't yet fully end to end generative, uh, neural network as what you would use today. Okay.swyx [00:15:30]: Awesome. And then when did you go like full time on Snips? Yeah.Kevin [00:15:33]: So basically that was, that was afterwards. I mean, how that started was the friend of mine who got me into machine learning, uh, him and I, uh, like he also got me interested into startups. He's had a big impact on my life. And the two of us were just a jam on, on like ideas for startups every now and then. And his background was also in AI data science. And we had a couple of ideas, but given that we were working full times, we were thinking about, uh, so we participated in Hack Zurich. That's, uh, Europe's biggest hackathon, um, or at least was at the time. And we said, Hey, this is just a weekend. Let's just try out an idea, like hack something together and see how it works. And the idea was that we'd be able to search through podcast episodes, like within a podcast. Yeah. So we did that. Long story short, uh, we managed to do it like to build something that we realized, Hey, this actually works. You can, you can find things again in podcasts. We had like a natural language search and we pitched it on stage. And we actually won the hackathon, which was cool. I mean, we, we also, I think we had a good, um, like a good, good pitch or a good example. So we, we used the famous Joe Rogan episode with Elon Musk where Elon Musk smokes a joint. Okay. Um, it's like a two and a half hour episode. So we were on stage and then we just searched for like smoking weed and it would find that exact moment. It will play it. And it just like, come on with Elon Musk, just like smoking. Oh, so it was video as well? No, it was actually completely based on audio. But we did have the video for the presentation. Yeah. Which had a, had of course an amazing effect. Yeah. Like this gave us a lot of activation energy, but it wasn't actually about winning the hackathon. Yeah. But the interesting thing that happened was after we pitched on stage, several of the other participants, like a lot of them came up to us and started saying like, Hey, can I use this? Like I have this issue. And like some also came up and told us about other problems that they have, like very adjacent to this with a podcast. Where's like, like this. Like, could, could I use this for that as well? And that was basically the, the moment where I realized, Hey, it's actually not just us who are having these issues with, with podcasts and getting to the, making the most out of this knowledge. Yeah. The other people. Yeah. That was now, I guess like four years ago or something like that. And then, yeah, we decided to quit our jobs and start, start this whole snip thing. Yeah. How big is the team now? We're just four people. Yeah. Just four people. Yeah. Like four. We're all technical. Yeah. Basically two on the, the backend side. So one of my co-founders is this person who got me into machine learning and startups. And we won the hackathon together. So we have two people for the backend side with the AI and all of the other backend things. And two for the front end side, building the app.swyx [00:18:18]: Which is mostly Android and iOS. Yeah.Kevin [00:18:21]: It's iOS and Android. We also have a watch app for, for Apple, but yeah, it's mostly iOS. Yeah.swyx [00:18:27]: The watch thing, it was very funny because in the, in the Latent Space discord, you know, most of us have been slowly adopting snips. You came to me like a year ago and you introduced snip to me. I was like, I don't know. I'm, you know, I'm very sticky to overcast and then slowly we switch. Why watch?Kevin [00:18:43]: So it goes back to a lot of our users, they do something else while, while listening to a podcast, right? Yeah. And one of the, us giving them the ability to then capture this knowledge, even though they're doing something else at the same time is one of the killer features. Yeah. Maybe I can actually, maybe at some point I should maybe give a bit more of an overview of what the, all of the features that we have. Sure. So this is one of the killer features and for one big use case that people use this for is for running. Yeah. So if you're a big runner, a big jogger or cycling, like really, really cycling competitively and a lot of the people, they don't want to take their phone with them when they go running. So you load everything onto the watch. So you can download episodes. I mean, if you, if you have an Apple watch that has internet access, like with a SIM card, you can also directly stream. That's also possible. Yeah. So of course it's a, it's basically very limited to just listening and snipping. And then you can see all of your snips later on your phone. Let me tell you this error I just got.swyx [00:19:47]: Error playing episode. Substack, the host of this podcast, does not allow this podcast to be played on an Apple watch. Yeah.Kevin [00:19:52]: That's a very beautiful thing. So we found out that all of the podcasts hosted on Substack, you cannot play them on an Apple watch. Why is this restriction? What? Like, don't ask me. We try to reach out to Substack. We try to reach out to some of the bigger podcasters who are hosting the podcast on Substack to also let them know. Substack doesn't seem to care. This is not specific to our app. You can also check out the Apple podcast app. Yeah. It's the same problem. It's just that we actually have identified it. And we tell the user what's going on.swyx [00:20:25]: I would say we host our podcast on Substack, but they're not very serious about their podcasting tools. I've told them before, I've been very upfront with them. So I don't feel like I'm shitting on them in any way. And it's kind of sad because otherwise it's a perfect creative platform. But the way that they treat podcasting as an afterthought, I think it's really disappointing.Kevin [00:20:45]: Maybe given that you mentioned all these features, maybe I can give a bit of a better overview of the features that we have. Let's do that. Let's do that. So I think we're mostly in our minds. Maybe for some of the listeners.swyx [00:20:55]: I mean, I'll tell you my version. Yeah. They can correct me, right? So first of all, I think the main job is for it to be a podcast listening app. It should be basically a complete superset of what you normally get on Overcast or Apple Podcasts or anything like that. You pull your show list from ListenNotes. How do you find shows? You've got to type in anything and you find them, right?Kevin [00:21:18]: Yeah. We have a search engine that is powered by ListenNotes. Yeah. But I mean, in the meantime, we have a huge database of like 99% of all podcasts out there ourselves. Yeah.swyx [00:21:27]: What I noticed, the default experience is you do not auto-download shows. And that's one very big difference for you guys versus other apps, where like, you know, if I'm subscribed to a thing, it auto-downloads and I already have the MP3 downloaded overnight. For me, I have to actively put it onto my queue, then it auto-downloads. And actually, I initially didn't like that. I think I maybe told you that I was like, oh, it's like a feature that I don't like. Like, because it means that I have to choose to listen to it in order to download and not to... It's like opt-in. There's a difference between opt-in and opt-out. So I opt-in to every episode that I listen to. And then, like, you know, you open it and depends on whether or not you have the AI stuff enabled. But the default experience is no AI stuff enabled. You can listen to it. You can see the snips, the number of snips and where people snip during the episode, which roughly correlates to interest level. And obviously, you can snip there. I think that's the default experience. I think snipping is really cool. Like, I use it to share a lot on Discord. I think we have tons and tons of just people sharing snips and stuff. Tweeting stuff is also like a nice, pleasant experience. But like the real features come when you actually turn on the AI stuff. And so the reason I got snipped, because I got fed up with Overcast not implementing any AI features at all. Instead, they spent two years rewriting their app to be a little bit faster. And I'm like, like, it's 2025. I should have a podcast that has transcripts that I can search. Very, very basic thing. Overcast will basically never have it.Kevin [00:22:49]: Yeah, I think that was a good, like, basic overview. Maybe I can add a bit to it with the AI features that we have. So one thing that we do every time a new podcast comes out, we transcribe the episode. We do speaker diarization. We identify the speaker names. Each guest, we extract a mini bio of the guest, try to find a picture of the guest online, add it. We break the podcast down into chapters, as in AI generated chapters. That one. That one's very handy. With a quick description per title and quick description per each chapter. We identify all books that get mentioned on a podcast. You can tell I don't use that one. It depends on the podcast. There are some podcasts where the guests often recommend like an amazing book. So later on, you can you can find that again.swyx [00:23:42]: So you literally search for the word book or I just read blah, blah, blah.Kevin [00:23:46]: No, I mean, it's all LLM based. Yeah. So basically, we have we have an LLM that goes through the entire transcript and identifies if a user mentions a book, then we use perplexity API together with various other LLM orchestration to go out there on the Internet, find everything that there is to know about the book, find the cover, find who or what the author is, get a quick description of it for the author. We then check on which other episodes the author appeared on.swyx [00:24:15]: Yeah, that is killer.Kevin [00:24:17]: Because that for me, if. If there's an interesting book, the first thing I do is I actually listen to a podcast episode with a with a writer because he usually gives a really great overview already on a podcast.swyx [00:24:28]: Sometimes the podcast is with the person as a guest. Sometimes his podcast is about the person without him there. Do you pick up both?Kevin [00:24:37]: So, yes, we pick up both in like our latest models. But actually what we show you in the app, the goal is to currently only show you the guest to separate that. In the future, we want to show the other things more.swyx [00:24:47]: For what it's worth, I don't mind. Yeah, I don't think like if I like if I like somebody, I'll just learn about them regardless of whether they're there or not.Kevin [00:24:55]: Yeah, I mean, yes and no. We we we have seen there are some personalities where this can break down. So, for example, the first version that we released with this feature, it picked up much more often a person, even if it was not a guest. Yeah. For example, the best examples for me is Sam Altman and Elon Musk. Like they're just mentioned on every second podcast and it has like they're not on there. And if you're interested in it, you can go to Elon Musk. And actually like learning from them. Yeah, I see. And yeah, we updated our our algorithms, improved that a lot. And now it's gotten much better to only pick it up if they're a guest. And yeah, so this this is maybe to come back to the features, two more important features like we have the ability to chat with an episode. Yes. Of course, you can do the old style of searching through a transcript with a keyword search. But I think for me, this is this is how you used to do search and extracting knowledge in the in the past. Old school. And the A.I. Web. Way is is basically an LLM. So you can ask the LLM, hey, when do they talk about topic X? If you're interested in only a certain part of the episode, you can ask them for four to give a quick overview of the episode. Key takeaways afterwards also to create a note for you. So this is really like very open, open ended. And yeah. And then finally, the snipping feature that we mentioned just to reiterate. Yeah. I mean, here the the feature is that whenever you hear an amazing idea, you can trip. It's up your headphones or click a button in the app and the A.I. summarizes the insight you just heard and saves that together with the original transcript and audio in your knowledge library. I also noticed that you you skip dynamic content. So dynamic content, we do not skip it automatically. Oh, sorry. You detect. But we detect it. Yeah. I mean, that's one of the thing that most people don't don't actually know that like the way that ads get inserted into podcasts or into most podcasts is actually that every time you listen. To a podcast, you actually get access to a different audio file and on the server, a different ad is inserted into the MP3 file automatically. Yeah. Based on IP. Exactly. And that's what that means is if we transcribe an episode and have a transcript with timestamps like words, word specific timestamps, if you suddenly get a different audio file, like the whole time says I messed up and that's like a huge issue. And for that, we actually had to build another algorithm that would dynamically on the floor. I re sync the audio that you're listening to the transcript that we have. Yeah. Which is a fascinating problem in and of itself.swyx [00:27:24]: You sync by matching up the sound waves? Or like, or do you sync by matching up words like you basically do partial transcription?Kevin [00:27:33]: We are not matching up words. It's happening on the basically a bytes level matching. Yeah. Okay.swyx [00:27:40]: It relies on this. It relies on the exact match at some point.Kevin [00:27:46]: So it's actually. We're actually not doing exact matches, but we're doing fuzzy matches to identify the moment. It's basically, we basically built Shazam for podcasts. Just as a little side project to solve this issue.swyx [00:28:02]: Actually, fun fact, apparently the Shazam algorithm is open. They published the paper, it's talked about it. I haven't really dived into the paper. I thought it was kind of interesting that basically no one else has built Shazam.Kevin [00:28:16]: Yeah, I mean, well, the one thing is the algorithm. If you now talk about Shazam, the other thing is also having the database behind it and having the user mindset that if they have this problem, they come to you, right?swyx [00:28:29]: Yeah, I'm very interested in the tech stack. There's a big data pipeline. Could you share what is the tech stack?Kevin [00:28:35]: What are the most interesting or challenging pieces of it? So the general tech stack is our entire backend is, or 90% of our backend is written in Python. Okay. Hosting everything on Google Cloud Platform. And our front end is written with, well, we're using the Flutter framework. So it's written in Dart and then compiled natively. So we have one code base that handles both Android and iOS. You think that was a good decision? It's something that a lot of people are exploring. So up until now, yes. Okay. Look, it has its pros and cons. Some of the, you know, for example, earlier, I mentioned we have a Apple Watch app. Yeah. I mean, there's no Flutter for that, right? So that you build native. And then of course you have to sort of like sync these things together. I mean, I'm not the front end engineer, so I'm not just relaying this information, but our front end engineers are very happy with it. It's enabled us to be quite fast and be on both platforms from the very beginning. And when I talk with people and they hear that we are using Flutter, usually they think like, ah, it's not performant. It's super junk, janky and everything. And then they use it. They use our app and they're always super surprised. Or if they've already used our app, I couldn't tell them. They're like, what? Yeah. Um, so there is actually a lot that you can do with it.swyx [00:29:51]: The danger, the concern, there's a few concerns, right? One, it's Google. So when were they, when are they going to abandon it? Two, you know, they're optimized for Android first. So iOS is like a second, second thought, or like you can feel that it is not a native iOS app. Uh, but you guys put a lot of care into it. And then maybe three, from my point of view, JavaScript, as a JavaScript guy, React Native was supposed to be there. And I think that it hasn't really fulfilled that dream. Um, maybe Expo is trying to do that, but, um, again, it is not, does not feel as productive as Flutter. And I've, I spent a week on Flutter and dot, and I'm an investor in Flutter flow, which is the local, uh, Flutter, Flutter startup. That's doing very, very well. I think a lot of people are still Flutter skeptics. Yeah. Wait. So are you moving away from Flutter?Kevin [00:30:41]: I don't know. We don't have plans to do that. Yeah.swyx [00:30:43]: You're just saying about that. What? Yeah. Watch out. Okay. Let's go back to the stack.Kevin [00:30:47]: You know, that was just to give you a bit of an overview. I think the more interesting things are, of course, on the AI side. So we, like, as I mentioned earlier, when we started out, it was before chat GPT for the chat GPT moment before there was the GPT 3.5 turbo, uh, API. So in the beginning, we actually were running everything ourselves, open source models, try to fine tune them. They worked. There was us, but let's, let's be honest. They weren't. What was the sort of? Before Whisper, the transcription. Yeah, we were using wave to work like, um, there was a Google one, right? No, it was a Facebook, Facebook one. That was actually one of the papers. Like when that came out for me, that was one of the reasons why I said we, we should try something to start a startup in the audio space. For me, it was a bit like before that I had been following the NLP space, uh, quite closely. And as, as I mentioned earlier, we, we did some stuff at the startup as well, that I was working up. But before, and wave to work was the first paper that I had at least seen where the whole transformer architecture moved over to audio and bit more general way of saying it is like, it was the first time that I saw the transformer architecture being applied to continuous data instead of discrete tokens. Okay. And it worked amazingly. Ah, and like the transformer architecture plus self-supervised learning, like these two things moved over. And then for me, it was like, Hey, this is now going to take off similarly. It's the text space has taken off. And with these two things in place, even if some features that we want to build are not possible yet, they will be possible in the near term, uh, with this, uh, trajectory. So that was a little side, side note. No, it's in the meantime. Yeah. We're using whisper. We're still hosting some of the models ourselves. So for example, the whole transcription speaker diarization pipeline, uh,swyx [00:32:38]: You need it to be as cheap as possible.Kevin [00:32:40]: Yeah, exactly. I mean, we're doing this at scale where we have a lot of audio.swyx [00:32:44]: We're what numbers can you disclose? Like what, what are just to give people an idea because it's a lot. So we have more than a million podcasts that we've already processed when you say a million. So processing is basically, you have some kind of list of podcasts that you will auto process and others where a paying pay member can choose to press the button and transcribe it. Right. Is that the rough idea? Yeah, exactly.Kevin [00:33:08]: Yeah. And if, when you press that button or we also transcribe it. Yeah. So first we do the, we do the transcription. We do the. The, the speaker diarization. So basically you identify speech blocks that belong to the same speaker. This is then all orchestrated within, within LLM to identify which speech speech block belongs to which speaker together with, you know, we identify, as I mentioned earlier, we identify the guest name and the bio. So all of that comes together with an LLM to actually then assign assigned speaker names to, to each block. Yeah. And then most of the rest of the, the pipeline we've now used, we've now migrated to LLM. So we use mainly open AI, Google models, so the Gemini models and the open AI models, and we use some perplexity basically for those things where we need, where we need web search. Yeah. That's something I'm still hoping, especially open AI will also provide us an API. Oh, why? Well, basically for us as a consumer, the more providers there are.swyx [00:34:07]: The more downtime.Kevin [00:34:08]: The more competition and it will lead to better, better results. And, um, lower costs over time. I don't, I don't see perplexity as expensive. If you use the web search, the price is like $5 per a thousand queries. Okay. Which is affordable. But, uh, if you compare that to just a normal LLM call, um, it's, it's, uh, much more expensive. Have you tried Exa? We've, uh, looked into it, but we haven't really tried it. Um, I mean, we, we started with perplexity and, uh, it works, it works well. And if I remember. Correctly, Exa is also a bit more expensive.swyx [00:34:45]: I don't know. I don't know. They seem to focus on the search thing as a search API, whereas perplexity, maybe more consumer-y business that is higher, higher margin. Like I'll put it like perplexity is trying to be a product, Exa is trying to be infrastructure. Yeah. So that, that'll be my distinction there. And then the other thing I will mention is Google has a search grounding feature. Yeah. Which you, which you might want. Yeah.Kevin [00:35:07]: Yeah. We've, uh, we've also tried that out. Um, not as good. So we, we didn't, we didn't go into. Too much detail in like really comparing it, like quality wise, because we actually already had the perplexity one and it, and it's, and it's working. Yeah. Um, I think also there, the price is actually higher than perplexity. Yeah. Really? Yeah.swyx [00:35:26]: Google should cut their prices.Kevin [00:35:29]: Maybe it was the same price. I don't want to say something incorrect, but it wasn't cheaper. It wasn't like compelling. And then, then there was no reason to switch. So, I mean, maybe like in general, like for us, given that we do work with a lot of content, price is actually something that we do look at. Like for us, it's not just about taking the best model for every task, but it's really getting the best, like identifying what kind of intelligence level you need and then getting the best price for that to be able to really scale this and, and provide us, um, yeah, let our users use these features with as many podcasts as possible. Yeah.swyx [00:36:03]: I wanted to double, double click on diarization. Yeah. Uh, it's something that I don't think people do very well. So you know, I'm, I'm a, I'm a B user. I don't have it right now. And, and they were supposed to speak, but they dropped out last minute. Um, but, uh, we've had them on the podcast before and it's not great yet. Do you use just PI Anode, the default stuff, or do you find any tricks for diarization?Kevin [00:36:27]: So we do use the, the open source packages, but we have tweaked it a bit here and there. For example, if you mentioned the BAI guys, I actually listened to the podcast episode was super nice. Thank you. And when you started talking about speaker diarization, and I just have to think about, uh, I don't know.Kevin [00:36:49]: Is it possible? I don't know. I don't know. F**k this. Yeah, no, I don't know.Kevin [00:36:55]: Yeah. We are the best. This is a.swyx [00:37:07]: I don't know. This is the best. I don't know. This is the best. Yeah. Yeah. Yeah. You're doing good.Kevin [00:37:12]: So, so yeah. This is great. This is good. Yeah. No, so that of course helps us. Another thing that helps us is that we know certain structural aspects of the podcast. For example, how often does someone speak? Like if someone, like let's say there's a one hour episode and someone speaks for 30 seconds, that person is most probably not the guest and not the host. It's probably some ad, like some speaker from an ad. So we have like certain of these heuristics that we can use and we leverage to improve things. And in the past, we've also changed the clustering algorithm. So basically how a lot of the speaker diarization works is you basically create an embedding for the speech that's happening. And then you try to somehow cluster these embeddings and then find out this is all one speaker. This is all another speaker. And there we've also tweaked a couple of things where we again used heuristics that we could apply from knowing how podcasts function. And that's also actually why I was feeling so much with the BAI guys, because like all of these heuristics, like for them, it's probably almost impossible to use any heuristics because it can just be any situation, anything.Kevin [00:38:34]: So that's one thing that we do. Yeah, another thing is that we actually combine it with LLM. So the transcript, LLMs and the speaker diarization, like bringing all of these together to recalibrate some of the switching points. Like when does the speaker stop? When does the next one start?swyx [00:38:51]: The LLMs can add errors as well. You know, I wouldn't feel safe using them to be so precise.Kevin [00:38:58]: I mean, at the end of the day, like also just to not give a wrong impression, like the speaker diarization is also not perfect that we're doing, right? I basically don't really notice it.swyx [00:39:08]: Like I use it for search.Kevin [00:39:09]: Yeah, it's not perfect yet, but it's gotten quite good. Like, especially if you compare, if you look at some of the, like if you take a latest episode and you compare it to an episode that came out a year ago, we've improved it quite a bit.swyx [00:39:23]: Well, it's beautifully presented. Oh, I love that I can click on the transcript and it goes to the timestamp. So simple, but you know, it should exist. Yeah, I agree. I agree. So this, I'm loading a two hour episode of Detect Me Right Home, where there's a lot of different guests calling in and you've identified the guest name. And yeah, so these are all LLM based. Yeah, it's really nice.Kevin [00:39:49]: Yeah, like the speaker names.swyx [00:39:50]: I would say that, you know, obviously I'm a power user of all these tools. You have done a better job than Descript. Okay, wow. Descript is so much funding. They had their open AI invested in them and they still suck. So I don't know, like, you know, keep going. You're doing great. Yeah, thanks. Thanks.Kevin [00:40:12]: I mean, I would, I would say that, especially for anyone listening who's interested in building a consumer app with AI, I think the, like, especially if your background is in AI and you love working with AI and doing all of that, I think the most important thing is just to keep reminding yourself of what's actually the job to be done here. Like, what does actually the consumer want? Like, for example, you now were just delighted by the ability to click on this word and it jumps there. Yeah. Like, this is not, this is not rocket science. This is, like, you don't have to be, like, I don't know, Android Kapathi to come up with that and build that, right? And I think that's, that's something that's super important to keep in mind.swyx [00:40:52]: Yeah, yeah. Amazing. I mean, there's so many features, right? It's, it's so packed. There's quotes that you pick up. There's summarization. Oh, by the way, I'm going to use this as my official feature request. I want to customize what, how it's summarized. I want to, I want to have a custom prompt. Yeah. Because your summarization is good, but, you know, I have different preferences, right? Like, you know.Kevin [00:41:14]: So one thing that you can already do today, I completely get your feature request. And I think it just.swyx [00:41:18]: I'm sure people have asked it.Kevin [00:41:19]: I mean, maybe just in general as a, as a, how I see the future, you know, like in the future, I think all, everything will be personalized. Yeah, yeah. Like, not, this is not specific to us. Yeah. And today we're still in a, in a phase where the cost of LLMs, at least if you're working with, like, such long context windows. As us, I mean, there's a lot of tokens in, if you take an entire podcast, so you still have to take that cost into consideration. So if for every single user, we regenerate it entirely, it gets expensive. But in the future, this, you know, cost will continue to go down and then it will just be personalized. So that being said, you can already today, if you go to the player screen. Okay. And open up the chat. Yeah. You can go to the, to the chat. Yes. And just ask for a summary in your style.swyx [00:42:13]: Yeah. Okay. I mean, I, I listen to consume, you know? Yeah. Yeah. I, I've never really used this feature. I don't know. I think that's, that's me being a slow adopter. No, no. I mean, that's. It has, when does the conversation start? Okay.Kevin [00:42:26]: I mean, you can just type anything. I think what you're, what you're describing, I mean, maybe that is also an interesting topic to talk about. Yes. Where, like, basically I told you, like, look, we have this chat. You can just ask for it. Yeah. And this is, this is how ChatGPT works today. But if you're building a consumer app, you have to move beyond the chat box. People do not want to always type out what they want. So your feature request was, even though theoretically it's already possible, what you are actually asking for is, hey, I just want to open up the app and it should just be there in a nicely formatted way. Beautiful way such that I can read it or consume it without any issues. Interesting. And I think that's in general where a lot of the, the. Opportunities lie currently in the market. If you want to build a consumer app, taking the capability and the intelligence, but finding out what the actual user interface is the best way how a user can engage with this intelligence in a natural way.swyx [00:43:24]: Is this something I've been thinking about as kind of like AI that's not in your face? Because right now, you know, we like to say like, oh, use Notion has Notion AI. And we have the little thing there. And there's, or like some other. Any other platform has like the sparkle magic wand emoji, like that's our AI feature. Use this. And it's like really in your face. A lot of people don't like it. You know, it should just kind of become invisible, kind of like an invisible AI.Kevin [00:43:49]: 100%. I mean, the, the way I see it as AI is, is the electricity of, of the future. And like no one, like, like we don't talk about, I don't know, this, this microphone uses electricity, this phone, you don't think about it that way. It's just in there, right? It's not an electricity enabled product. No, it's just a product. Yeah. It will be the same with AI. I mean, now. It's still a, something that you use to market your product. I mean, we do, we do the same, right? Because it's still something that people realize, ah, they're doing something new, but at some point, no, it'll just be a podcast app and it will be normal that it has all of this AI in there.swyx [00:44:24]: I noticed you do something interesting in your chat where you source the timestamps. Yeah. Is that part of this prompt? Is there a separate pipeline that adds source sources?Kevin [00:44:33]: This is, uh, actually part of the prompt. Um, so this is all prompt engine. Engineering, um, uh, you should be able to click on it. Yeah, I clicked on it. Um, this is all prompt engineering with how to provide the, the context, you know, we, because we provide all of the transcript, how to provide the context and then, yeah, I get them all to respond in a correct way with a certain format and then rendering that on the front end. This is one of the examples where I would say it's so easy to create like a quick demo of this. I mean, you can just go to chat to be deep, paste this thing in and say like, yeah, do this. Okay. Like 15 minutes and you're done. Yeah. But getting this to like then production level that it actually works 99% of the time. Okay. This is then where, where the difference lies. Yeah. So, um, for this specific feature, like we actually also have like countless regexes that they're just there to correct certain things that the LLM is doing because it doesn't always adhere to the format correctly. And then it looks super ugly on the front end. So yeah, we have certain regexes that correct that. And maybe you'd ask like, why don't you use an LLM for that? Because that's sort of the, again, the AI native way, like who uses regexes anymore. But with the chat for user experience, it's very important that you have the streaming because otherwise you need to wait so long until your message has arrived. So we're streaming live the, like, just like ChatGPT, right? You get the answer and it's streaming the text. So if you're streaming the text and something is like incorrect. It's currently not easy to just like pipe, like stream this into another stream, stream this into another stream and get the stream back, which corrects it, that would be amazing. I don't know, maybe you can answer that. Do you know of any?swyx [00:46:19]: There's no API that does this. Yeah. Like you cannot stream in. If you own the models, you can, uh, you know, whatever token sequence has, has been emitted, start loading that into the next one. If you fully own the models, uh, I don't, it's probably not worth it. That's what you do. It's better. Yeah. I think. Yeah. Most engineers who are new to AI research and benchmarking actually don't know how much regexing there is that goes on in normal benchmarks. It's just like this ugly list of like a hundred different, you know, matches for some criteria that you're looking for. No, it's very cool. I think it's, it's, it's an example of like real world engineering. Yeah. Do you have a tooling that you're proud of that you've developed for yourself?Kevin [00:47:02]: Is it just a test script or is it, you know? I think it's a bit more, I guess the term that has come up is, uh, vibe coding, uh, vibe coding, some, no, sorry, that's actually something else in this case, but, uh, no, no, yes, um, vibe evals was a term that in one of the talks actually on, on, um, I think it might've been the first, the first or the first day at the conference, someone brought that up. Yeah. Uh, because yeah, a lot of the talks were about evals, right. Which is so important. And yeah, I think for us, it's a bit more vibe. Evals, you know, that's also part of, you know, being a startup, we can take risks, like we can take the cost of maybe sometimes it failing a little bit or being a little bit off and our users know that and they appreciate that in return, like we're moving fast and iterating and building, building amazing things, but you know, a Spotify or something like that, half of our features will probably be in a six month review through legal or I don't know what, uh, before they could sell them out.swyx [00:48:04]: Let's just say Spotify is not very good at podcasting. Um, I have a documented, uh, dislike for, for their podcast features, just overall, really, really well integrated any other like sort of LLM focused engineering challenges or problems that you, that you want to highlight.Kevin [00:48:20]: I think it's not unique to us, but it goes again in the direction of handling the uncertainty of LLMs. So for example, with last year, at the end of the year, we did sort of a snipped wrapped. And one of the things we thought it would be fun to, just to do something with, uh, with an LLM and something with the snips that, that a user has. And, uh, three, let's say unique LLM features were that we assigned a personality to you based on the, the snips that, that you have. It was, I mean, it was just all, I guess, a bit of a fun, playful way. I'm going to look up mine. I forgot mine already.swyx [00:48:57]: Um, yeah, I don't know whether it's actually still in the, in the, we all took screenshots of it.Kevin [00:49:01]: Ah, we posted it in the, in the discord. And the, the second one, it was, uh, we had a learning scorecard where we identified the topics that you snipped on the most, and you got like a little score for that. And the third one was a, a quote that stood out. And the quote is actually a very good example of where we would run that for user. And most of the time it was an interesting quote, but every now and then it was like a super boring quotes that you think like, like how, like, why did you select that? Like, come on for there. The solution was actually just to say, Hey, give me five. So it extracted five quotes as a candidate, and then we piped it into a different model as a judge, LLM as a judge, and there we use a, um, a much better model because with the, the initial model, again, as, as I mentioned also earlier, we do have to look at the, like the, the costs because it's like, we have so much text that goes into it. So we, there we use a bit more cheaper model, but then the judge can be like a really good model to then just choose one out of five. This is a practical example.swyx [00:50:03]: I can't find it. Bad search in discord. Yeah. Um, so, so you do recommend having a much smarter model as a judge, uh, and that works for you. Yeah. Yeah. Interesting. I think this year I'm very interested in LM as a judge being more developed as a concept, I think for things like, you know, snips, raps, like it's, it's fine. Like, you know, it's, it's, it's, it's entertaining. There's no right answer.Kevin [00:50:29]: I mean, we also have it. Um, we also use the same concept for our books feature where we identify the, the mention. Books. Yeah. Because there it's the same thing, like 90% of the time it, it works perfectly out of the box one shot and every now and then it just, uh, starts identifying books that were not really mentioned or that are not books or made, yeah, starting to make up books. And, uh, they are basically, we have the same thing of like another LLM challenging it. Um, yeah. And actually with the speakers, we do the same now that I think about it. Yeah. Um, so I'm, I think it's a, it's a great technique. Interesting.swyx [00:51:05]: You run a lot of calls.Kevin [00:51:07]: Yeah.swyx [00:51:08]: Okay. You know, you mentioned costs. You move from self hosting a lot of models to the, to the, you know, big lab models, open AI, uh, and Google, uh, non-topic.Kevin [00:51:18]: Um, no, we love Claude. Like in my opinion, Claude is the, the best one when it comes to the way it formulates things. The personality. Yeah. The personality. Okay. I actually really love it. But yeah, the cost is. It's still high.swyx [00:51:36]: So you cannot, you tried Haiku, but you're, you're like, you have to have Sonnet.Kevin [00:51:40]: Uh, like basically we like with Haiku, we haven't experimented too much. We obviously work a lot with 3.5 Sonnet. Uh, also, you know, coding. Yeah. For coding, like in cursor, just in general, also brainstorming. We use it a lot. Um, I think it's a great brainstorm partner, but yeah, with, uh, with, with a lot of things that we've done done, we opted for different models.swyx [00:52:00]: What I'm trying to drive at is how much cheaper can you get if you go from cloud to cloud? Closed models to open models. And maybe it's like 0% cheaper, maybe it's 5% cheaper, or maybe it's like 50% cheaper. Do you have a sense?Kevin [00:52:13]: It's very difficult to, to judge that. I don't really have a sense, but I can, I can give you a couple of thoughts that have gone through our minds over the time, because obviously we do realize like, given that we, we have a couple of tasks where there are just so many tokens going in, um, at some point it will make sense to, to offload some of that. Uh, to an open source model, but going back to like, we're, we're a startup, right? Like we're not an AI lab or whatever, like for us, actually the most important thing is to iterate fast because we need to learn from our users, improve that. And yeah, just this velocity of this, these iterations. And for that, the closed models hosted by open AI, Google is, uh, and swapping, they're just unbeatable because you just, it's just an API call. Yeah. Um, so you don't need to worry about. Yeah. So much complexity behind that. So this is, I would say the biggest reason why we're not doing more in this space, but there are other thoughts, uh, also for the future. Like I see two different, like we basically have two different usage patterns of LLMs where one is this, this pre-processing of a podcast episode, like this initial processing, like the transcription, speaker diarization, chapterization. We do that once. And this, this usage pattern it's, it's quite predictable. Because we know how many podcasts get released when, um, so we can sort of have a certain capacity and we can, we, we're running that 24 seven, it's one big queue running 24 seven.swyx [00:53:44]: What's the queue job runner? Uh, is it a Django, just like the Python one?Kevin [00:53:49]: No, that, that's just our own, like our database and the backend talking to the database, picking up jobs, finding it back. I'm just curious in orchestration and queues. I mean, we, we of course have like, uh, a lot of other orchestration where we're, we're, where we use, uh, the Google pub sub, uh, thing, but okay. So we have this, this, this usage pattern of like very predictable, uh, usage, and we can max out the, the usage. And then there's this other pattern where it's, for example, the snippet where it's like a user, it's a user action that triggers an LLM call and it has to be real time. And there can be moments where it's by usage and there can be moments when there's very little usage for that. There. So that's, that's basically where these LLM API calls are just perfect because you don't need to worry about scaling this up, scaling this down, um, handling, handling these issues. Serverless versus serverful.swyx [00:54:44]: Yeah, exactly. Okay.Kevin [00:54:45]: Like I see them a bit, like I see open AI and all of these other providers, I see them a bit as the, like as the Amazon, sorry, AWS of, of AI. So it's a bit similar how like back before AWS, you would have to have your, your servers and buy new servers or get rid of servers. And then with AWS, it just became so much easier to just ramp stuff up and down. Yeah. And this is like the taking it even, even, uh, to the next level for AI. Yeah.swyx [00:55:18]: I am a big believer in this. Basically it's, you know, intelligence on demand. Yeah. We're probably not using it enough in our daily lives to do things. I should, we should be able to spin up a hundred things at once and go through things and then, you know, stop. And I feel like we're still trying to figure out how to use LLMs in our lives effectively. Yeah. Yeah.Kevin [00:55:38]: 100%. I think that goes back to the whole, like that, that's for me where the big opportunity is for, if you want to do a startup, um, it's not about, but you can let the big labs handleswyx [00:55:48]: the challenge of more intelligence, but, um, it's the... Existing intelligence. How do you integrate? How do you actually incorporate it into your life? AI engineering. Okay, cool. Cool. Cool. Cool. Um, the one, one other thing I wanted to touch on was multimodality in frontier models. Dwarcash had a interesting application of Gemini recently where he just fed raw audio in and got diarized transcription out or timestamps out. And I think that will come. So basically what we're saying here is another wave of transformers eating things because right now models are pretty much single modality things. You know, you have whisper, you have a pipeline and everything. Yeah. You can't just say, Oh, no, no, no, we only fit like the raw, the raw files. Do you think that will be realistic for you? I 100% agree. Okay.Kevin [00:56:38]: Basically everything that we talked about earlier with like the speaker diarization and heuristics and everything, I completely agree. Like in the, in the future that would just be put everything into a big multimodal LLM. Okay. And it will output, uh, everything that you want. Yeah. So I've also experimented with that. Like just... With, with Gemini 2? With Gemini 2.0 Flash. Yeah. Just for fun. Yeah. Yeah. Because the big difference right now is still like the cost difference of doing speaker diarization this way or doing transcription this way is a huge difference to the pipeline that we've built up. Huh. Okay.swyx [00:57:15]: I need to figure out what, what that cost is because in my mind 2.0 Flash is so cheap. Yeah. But maybe not cheap enough for you.Kevin [00:57:23]: Uh, no, I mean, if you compare it to, yeah, whisper and speaker diarization and especially self-hosting it and... Yeah. Yeah. Yeah.swyx [00:57:30]: Yeah.Kevin [00:57:30]: Okay. But we will get there, right? Like this is just a question of time.swyx [00:57:33]: And, um, at some point, as soon as that happens, we'll be the first ones to switch. Yeah. Awesome. Anything else that you're like sort of eyeing on the horizon as like, we are thinking about this feature, we're thinking about incorporating this new functionality of AI into our, into our app? Yeah.Kevin [00:57:50]: I mean, we, there's so many areas that we're thinking about, like our challenge is a bit more... Choosing. Yeah. Choosing. Yeah. So, I mean, I think for me, like looking into like the next couple of years, like the big areas that interest us a lot, basically four areas, like one is content. Um, right now it's, it's podcasts. I mean, you did mention, I think you mentioned like you can also upload audio books and YouTube videos. YouTube. I actually use the YouTube one a fair amount. But in the future, we, we want to also have audio books natively in the app. And, uh, we want to enable AI generated content. Like just think of, take deep research and notebook analysis. Like put these together. That should be, that should be in our app. The second area is discovery. I think in general. Yeah.swyx [00:58:38]: I noticed that you don't have, so you have download counts and most snips. Right. Something like that. Yeah. Yeah.Kevin [00:58:45]: On the discovery side, we want to do much, much more. I think in general, discovery as a paradigm in all apps is, will undergo a change thanks Thanks to AI. You know, there has been a lot of talk. Before Elon bought Twitter, there was a lot of talk about bring your own algorithm to Twitter. And that was Jack Dorsey's big thing. He talked a lot about that. And I actually think this is coming, but with a bit of a twist. So I think what actually AI will enable is not that you bring your own algorithm, but you will be able to talk. You will be able to communicate with the algorithm. So you can just tell the algorithm, like, hey, you keep showing me cat videos. And I know I freaking love them. And that's why you keep showing them to me. But please, for the next two hours, I really want to get more into AI stuff. Do not show me cat videos. And then it will just adapt. And of course, the question is, you know, like big platforms like, I don't know, let's say TikTok. They do not have the incentive to offer that.swyx [00:59:49]: Exactly. That's what I was going to say.Kevin [00:59:50]: But we actually, we are driven by helping you learn, get the most, like achieve your goals. And so for us, it's actually very much our incentive. Like, hey, you know, you should be able to guide it. Yeah. So that was a long way of saying that I think there will happen a lot in recommendations. Order by.swyx [01:00:12]: The most popular. Yeah. I think collaborative filtering will be the first step, right? For Rexis and then some LLM fancy stuff.Kevin [01:00:20]: Yeah. Maybe to go back to the question that you had before. So the other, like these were the first two areas. Yeah. The two are voice, voices and interfaces and voice AI. Well, how is this going to exist? Yeah. So maybe I can tell you a bit first, like why I find it so interesting for us. Yeah. Because voice as an interface, like historically, there has been so much talk about it and it always fell flat. The reason why I'm excited about it this time around is with any consumer app, I like to ask myself, what is the... moment in my life, what is the trigger in my life that gets me to open this app and start using it? So, for example, I don't know, take Airbnb. It's the trigger is like, ah, you want to travel and then and then you, you do that and then you open up the app. Apps that do not have this already existing natural trigger in your life, it's very difficult for a consumer app to then get the user to open the app again. You need a hook. Yeah. There's basically only one app. One super successful app that has been able to do that without this natural trigger, and that is Duolingo. So Duolingo, like everyone wants to learn a language, but there's, you don't have this natural moment during your day where it's like, ah, now I need to open up this app. You have the notifications. Exactly. The owl memes. Exactly. So they, I mean, they gamified the s**t super successful, super beautiful. They are the GOATs in this arena. But the much easier is actually... No, there is already this trigger and then you don't have to do all of the streaks and leaderboards and everything. Okay. That's a bit of a context. Now, if you look at what we're doing and our goal of getting people to really maximize what they get out of their listening, we are interested in, there are a couple of features where we know we can sort of 10x the value that people get out of a podcast. Okay. But we need them to do something for that. There is friction involved. Because it's all about learning, right? It's about thinking for yourself. Like, those are the moments when you actually start, yeah, really 10x-ing the value that you got out of the podcast instead of just consuming it.swyx [01:02:37]: Applying the knowledge. Yeah. Okay.Kevin [01:02:39]: Basically, being forced to think about like, what was actually the main takeaway for you from this episode? Okay. Like, there's something that I like doing myself for every episode that I listen to, I try to boil it down to, like, try to decide one single takeaway. Yeah. Even though there might have been 10. Yeah. There might have been 10 amazing things. Pick one. One most important one. Yeah. And this is an active process that is like a forcing function in your brain to challenge all of the insights and really come up with the one thing that is applicable to you and your life and what you might want to do with it. So it also helps you to turn it into action. This is basically a feature that we're interested in, but you have to get the user to use that, right? So when do you get the user to use that? Yeah. So if this is all text-based, then we're basically playing the same game as Duolingo, where at some point you're going to get a notification from Snip and be like, hey, Swyx, come on, you know you should do this. Maybe there's a blue owl.Kevin [01:03:40]: But if you have voice, you can basically hook into the existing habits that the user already has. So you already have this habit that you listen to a podcast. You're already doing that. Yeah. And once an episode ends, instead of just jumping into the next episode, you can now actually have your AI companion come on and you can have a quick conversation. You can go through these things. And how that looks like in detail, we need to figure that out. But just this paradigm of you're staying in the flow. This also relates to what you were saying, like AI that is invisible. You're staying in the flow of what you're already doing. But now we can insert a completely new experience. That helps you get the most out of real estate. Yeah.swyx [01:04:27]: I think your framing of this is very powerful. Because I think this is where you are a product person more than an engineer. Because an engineer would just be like, oh, it's just chat with your podcast. It's like chat with PDF, chat with podcast. Okay, cool. But you're framing it in a different light that actually makes sense to me now, as opposed to previously. I don't chat with my podcast. Why? I just listen to the podcast. But for you, it's more about retention and learning and all that. And because you're very serious about it, that's why you started the company. So you're focused on that. Whereas I'm still stuck in that consume, consume, consume mentality. And I know it's not good, but this is my default. Which is why I was a little bit lost when you were saying all the things about Duolingo. And you're saying the things about the trigger. This is my trigger for listening to the podcast is I'm by myself. That's my trigger. But you're saying the trigger is not about listening to the podcast. The trigger is remembering and retaining and processing the podcast I just listened to.Kevin [01:05:41]: So what I meant, you already have this trigger that gets you to start listening to a podcast. Yes. This you already have. And so do, I don't know. Millions of people. Yeah. So there are more than half a billion monthly active podcast listeners. Okay. So you already have this trigger that gets you to start listening. But you do not have this trigger. As you just said yourself, basically, you do not have this trigger that gets you to regularly process this information. And voice basically for me is the ability to hook into your existing trigger with the trigger that I was talking about is basically your podcast. And you're just still listening. So we just continue and we can now spend, you know, this can be two minutes. Like I'm not saying now this is like a 60 minute process. I think like two minutes, three minutes that can just come on completely naturally. And if we manage to do that and you start noticing as a user, like freaking hell, like I'm just now spending three minutes with this AI companion. But like. Your retention is more. I'm taking this much away. And it's not. And like retention is one thing. But you're like. Yeah. You start to take what you've learned and apply it to what's important to you. Like you're thinking. Yeah. And if we get you to notice that feeling, then yeah, then we've won. Yeah.swyx [01:07:05]: I would say like a lot of people rely on Anki, Anki notes like flashcards and all that to do that. But making the notes is also a chore. And I think this could be very, very interesting. I think that I'm just noticing that it's kind of like a different usage mode. Like you already talked about this. You know, the name of Snips is very Snip centric. And I actually originally also resisted adopting Snip because of that. But now you're like, you know, you observe that people are listening to long form episodes and you're talking at the end. Like the ideal implementation of this is I browse through a bunch of Snips of the things that I'm subscribed to. I listen to the Snips. I talk with it. And then maybe it double clicks on the podcast and it goes and finds other timestamps that are relevant to the thing that I want to talk about. Just. I don't know that. I don't know if that's interesting.Kevin [01:07:53]: I think these are all areas that we should explore. Yeah.swyx [01:07:57]: Like we're still quite open about how this will look like in detail. What are your thoughts on voice cloning? Everyone wants to continue. I have had my voice clones and people have talked to me, the AI version of me. Is that too creepy?Kevin [01:08:13]: I don't think it's too creepy in the future. Okay. With a lot of these things in our society is going through a change. And things seem quite weird now that in the future will seem normal. I think already voice cloning has become much more normalized. I remember I was at the, I think it was 2017 Nips conference. San Diego?swyx [01:08:42]: No, LA. LA. It was the Flo Rida one? Yeah. Yeah. Flo Rida. Yeah.Kevin [01:08:47]: So everyone says that was peak Nips. Yeah. I remember there was this talk or workshop by Liar Bird. They actually got acquired by Descript later. They were doing voice cloning and they were showing off their tech. And there was this huge discussion later on, like all of the moral implications and ethical implications. And it really felt like this would never be accepted by society. And you look now, you have 11 labs and just anyone can just clone their voice. And no one really talks about it as like, oh my God, the world is going to end. Yeah. So I think society will get used to that. In our case, I think there are some interesting applications where we'd also be super interested in working together with creators, like podcast creators, to play a bit around with this concept. I think that would be super cool if someone can come onto Snipped, go to the Latent Spaceswyx [01:09:42]: podcast and start chatting with AI Swyx. Yeah. No, I think we'd be there. Yeah. We want to, obviously, I think as an AI podcast, we should be first consumers of these things. Yeah. I would say that one observation I've made about podcasting, this is the general state of the market. And you can ask me your questions, things you want to ask about podcasters. We are focusing a lot more on YouTube this year. YouTube is the best podcasting platform. It is not MP3s. It is not Apple Podcasts. It is not Spotify. It's YouTube. And it's just the social layer of recommendations and the existing habit that people have of logging onto YouTube and getting that. That's my observation. You can riff on that. The only thing I would just say is like, when you were listing your list of priorities, you said audio books first over YouTube.Kevin [01:10:26]: And I would switch that if I were you. Yeah. Like as in YouTube, video, video podcasts. I mean, it's obvious that video podcasts are here to stay. Not just here to stay, bigger. Yeah. What I want to do with Snipped is obviously also add video to the platform. Oh, yeah. The way I see video is I do believe it's... Yeah. I like this concept of backgroundable video. I didn't come up with this concept. It was actually Gustav Söderström. The CPO of Spotify. Exactly. Exactly. When I speak with people, it remains true that they listen to podcasts when they do something else at the same time. Like this is like 90% of their consumption. Also if they listen to on YouTube. But every now and then it's nice to have the video. It's nice if you're, for example, just watching a clip. It's nice if they sometimes mention something, like they show some slides or they show something where you need to have the visual with it. It helps you connect much more with the host as a listener. But the biggest benefit I see with video is discovery. I think that is also why YouTube has become the biggest podcast player out there because they have the discovery. And discovery in video is just so much easier and so much better. And so much more engaging. So this is the area where I'm most interested about when it comes to video and snips. That we can provide a much better, much more engaging and much more fun discovery experience. For consumers? Yeah, for consumers.swyx [01:12:01]: Okay. I think that you almost have like three different audiences. The vast majority of people for you is the people listening to podcasts. Right? Of course. Then there's a second layer of people who create snips. Right? Who add extra data, annotation value to your platform. By the way, we use the snip count as a proxy for popularity, right? Because we have download counts, but for example, platforms like Spotify re-host our MP3 file. So we don't get any download count for Spotify. Snip count is active, like I opt in to listen to you and I shared this. Those are really, really good metrics. But the third audience that you haven't really touched is the podcast creators like myself. And for me, discovery from that point of view, not from your point of view, discovery for me is like, I want to be discovered. And I think YouTube is still there. Twitter, obviously for me, Substack, Hacker News. I really try very hard to rank on Hacker News. I think when TikTok took this very seriously, they prioritized the creators of the content. And for you, the creator of the content was the snips. But there may be a world for you in which you prioritize the creators of the podcast.Kevin [01:13:10]: Yeah. Interesting observation. What are some of your ideas or thoughts? Do you have some specific?swyx [01:13:18]: Riverside is the closest that has come to it. Descript is number two. Descript bought a Riverside competitor and as far as I can tell, it's not been very successful. Descript just has a very, very good niche, very, very good editing angle and then just hasn't done anything interesting since then. Although Underlord is good, it's not great. Your chapterization is better than Descript's. Again, they should be able to beat you. They're not. And Riverside is good also. Very, very good. Very, very, very good. So we actually recently started a second series of podcasts within Latent Space that is YouTube only because you only find it on YouTube. And it's also shorter. So this is like a one and a half hour, two hour thing. Remote only, 30 minutes, chop, chop. Send it on to Riverside. Riverside, pretty good for that. Not great. It doesn't do good thumbnails. It doesn't do good. The editing is still a little bit rough. It has this auto editor where whoever's actively speaking, it focuses on the editor, on the active speaker. And then sometimes it goes back to the multi-speaker view, that kind of stuff. People like that. Okay. But the shorts are still not great. I still need to manually download it and then republish it to YouTube. The shorts I still need to pick. They mostly suck. There's still a lot of rough edges there that ideally, me as a creator, you know what I want. You definitely know what I want. I sit down, record, press a button, done. We're still not there.Kevin [01:14:46]: I think you guys could do it. Okay. So if I can translate that for you, it's really about the simplifying the creation process of the podcast. Yeah.swyx [01:14:55]: And I'll tell you what, this will increase the quality because the reason that most podcasts or YouTube videos are s**t is they are made by people who don't have life experience, who are not that important in the world. They're not doing important jobs. And so what you want to actually enable is CEOs to each of them make their own podcasts who are busy. They're not going to sit there and figure out Riverside. A lot of the reason that people like Latent Space is it takes an idiot like me who could be doing a lot more with my life, making a lot more money, having a real job somewhere else. I just choose to do this because I like it. But otherwise, they will never get access to me and the access to the people that I have access to. So that's my pitch. Cool.swyx [01:15:44]: Anything else that you normally want to talk to podcasters about?Kevin [01:15:46]: I think we've covered everything. I guess like last messages, you know, go try out Snipped. Yeah. It's a premium version so you can use and try out everything for free. Also happy to provide you with a link that you can add to the show notes. Try out the premium version also for free for a month if people want to do that. Yeah. Give it a shot.swyx [01:16:08]: I would say. Yeah. Thanks for coming on. I would say that after you demoed me, I did not convert for another four to six months because I found it very challenging to switch over. And I think that's the main thing. Like you basically had you have import OPML. Right. But there's no way to import like all the existing like half listened to episodes or like my rankings or whatever. And for that, for listeners who are. I have a blog post where I talked about my switch. Just treat it as a chance to clean house.swyx [01:16:45]: That's a good point. Do things and, you know, just refocus here. First start. 2025. Yeah. Great. Well, thank you for working on Snipped. Thank you for coming on. You know, we usually spend a lot of time talking to like big companies like venture startups, B2B, SaaS, you know, that kind of stuff. But I think your journey is like, you know, it's a small team building a B2C consumer app. It's the kind of stuff that we like to also feature because a lot of people want to build what you're doing. And they don't see role models that are successful, that are confidence, that are like having success in this market, which is very challenging. So, yeah, thanks for thanks for sharing some of your thoughts. Thanks.Kevin [01:17:26]: Yeah, thanks. Thanks for having me. And thank you for creating an amazing podcast and an amazing conference as well.swyx [01:17:32]: Thank you. Get full access to Latent.Space at www.latent.space/subscribe

  --------  

  1:17:47
• ⚡️The new OpenAI Agents Platform

  While everyone is now repeating that 2025 is the “Year of the Agent”, OpenAI is heads down building towards it. In the first 2 months of the year they released Operator and Deep Research (arguably the most successful agent archetype so far), and today they are bringing a lot of those capabilities to the API:* Responses API* Web Search Tool* Computer Use Tool* File Search Tool* A new open source Agents SDK with integrated Observability ToolsWe cover all this and more in today’s lightning pod on YouTube!More details here:Responses APIIn our Michelle Pokrass episode we talked about the Assistants API needing a redesign. Today OpenAI is launching the Responses API, “a more flexible foundation for developers building agentic applications”. It’s a superset of the chat completion API, and the suggested starting point for developers working with OpenAI models. One of the big upgrades is the new set of built-in tools for the responses API: Web Search, Computer Use, and Files. Web Search ToolWe previously had Exa AI on the podcast to talk about web search for AI. OpenAI is also now joining the race; the Web Search API is actually a new “model” that exposes two 4o fine-tunes: gpt-4o-search-preview and gpt-4o-mini-search-preview. These are the same models that power ChatGPT Search, and are priced at $30/1000 queries and $25/1000 queries respectively. The killer feature is inline citations: you do not only get a link to a page, but also a deep link to exactly where your query was answered in the result page. Computer Use ToolThe model that powers Operator, called Computer-Using-Agent (CUA), is also now available in the API. The computer-use-preview model is SOTA on most benchmarks, achieving 38.1% success on OSWorld for full computer use tasks, 58.1% on WebArena, and 87% on WebVoyager for web-based interactions.As you will notice in the docs, `computer-use-preview` is both a model and a tool through which you can specify the environment. Usage is priced at $3/1M input tokens and $12/1M output tokens, and it’s currently only available to users in tiers 3-5.File Search ToolFile Search was also available in the Assistants API, and it’s now coming to Responses too. OpenAI is bringing search + RAG all under one umbrella, and we’ll definitely see more people trying to find new ways to build all-in-one apps on OpenAI. Usage is priced at $2.50 per thousand queries and file storage at $0.10/GB/day, with the first GB free.Agent SDK: Swarms++!https://github.com/openai/openai-agents-pythonTo bring it all together, after the viral reception to Swarm, OpenAI is releasing an officially supported agents framework (which was previewed at our AI Engineer Summit) with 4 core pieces:* Agents: Easily configurable LLMs with clear instructions and built-in tools.* Handoffs: Intelligently transfer control between agents.* Guardrails: Configurable safety checks for input and output validation.* Tracing & Observability: Visualize agent execution traces to debug and optimize performance.Multi-agent workflows are here to stay!OpenAI is now explicitly designs for a set of common agentic patterns: Workflows, Handoffs, Agents-as-Tools, LLM-as-a-Judge, Parallelization, and Guardrails. OpenAI previewed this in part 2 of their talk at NYC:Further coverage of the launch from Kevin Weil, WSJ, and OpenAIDevs, AMA here.Show Notes* Assistants API* Swarm (OpenAI)* Fine-Tuning in AI* 2024 OpenAI DevDay Recap with Romain* Michelle Pokrass episode (API lead)Timestamps* 00:00 Intros* 02:31 Responses API * 08:34 Web Search API * 17:14 Files Search API * 18:46 Files API vs RAG * 20:06 Computer Use / Operator API * 22:30 Agents SDKAnd of course you can catch up with the full livestream here:TranscriptAlessio [00:00:03]: Hey, everyone. Welcome back to another Latent Space Lightning episode. This is Alessio, partner and CTO at Decibel, and I'm joined by Swyx, founder of Small AI.swyx [00:00:11]: Hi, and today we have a super special episode because we're talking with our old friend Roman. Hi, welcome.Romain [00:00:19]: Thank you. Thank you for having me.swyx [00:00:20]: And Nikunj, who is most famously, if anyone has ever tried to get any access to anything on the API, Nikunj is the guy. So I know your emails because I look forward to them.Nikunj [00:00:30]: Yeah, nice to meet all of you.swyx [00:00:32]: I think that we're basically convening today to talk about the new API. So perhaps you guys want to just kick off. What is OpenAI launching today?Nikunj [00:00:40]: Yeah, so I can kick it off. We're launching a bunch of new things today. We're going to do three new built-in tools. So we're launching the web search tool. This is basically chat GPD for search, but available in the API. We're launching an improved file search tool. So this is you bringing your data to OpenAI. You upload it. We, you know, take care of parsing it, chunking it. We're embedding it, making it searchable, give you this like ready vector store that you can use. So that's the file search tool. And then we're also launching our computer use tool. So this is the tool behind the operator product in chat GPD. So that's coming to developers today. And to support all of these tools, we're going to have a new API. So, you know, we launched chat completions, like I think March 2023 or so. It's been a while. So we're looking for an update over here to support all the new things that the models can do. And so we're launching this new API. It is, you know, it works with tools. We think it'll be like a great option for all the future agentic products that we build. And so that is also launching today. Actually, the last thing we're launching is the agents SDK. We launched this thing called Swarm last year where, you know, it was an experimental SDK for people to do multi-agent orchestration and stuff like that. It was supposed to be like educational experimental, but like people, people really loved it. They like ate it up. And so we are like, all right, let's, let's upgrade this thing. Let's give it a new name. And so we're calling it the agents SDK. It's going to have built-in tracing in the OpenAI dashboard. So lots of cool stuff going out. So, yeah.Romain [00:02:14]: That's a lot, but we said 2025 was the year of agents. So there you have it, like a lot of new tools to build these agents for developers.swyx [00:02:20]: Okay. I guess, I guess we'll just kind of go one by one and we'll leave the agents SDK towards the end. So responses API, I think the sort of primary concern that people have and something I think I've voiced to you guys when, when, when I was talking with you in the, in the planning process was, is chat completions going away? So I just wanted to let it, let you guys respond to the concerns that people might have.Romain [00:02:41]: Chat completion is definitely like here to stay, you know, it's a bare metal API we've had for quite some time. Lots of tools built around it. So we want to make sure that it's maintained and people can confidently keep on building on it. At the same time, it was kind of optimized for a different world, right? It was optimized for a pre-multi-modality world. We also optimized for kind of single turn. It takes two problems. It takes prompt in, it takes response out. And now with these agentic workflows, we, we noticed that like developers and companies want to build longer horizon tasks, you know, like things that require multiple returns to get the task accomplished. And computer use is one of those, for instance. And so that's why the responses API came to life to kind of support these new agentic workflows. But chat completion is definitely here to stay.swyx [00:03:27]: And assistance API, we've, uh, has a target sunset date of first half of 2020. So this is kind of like, in my mind, there was a kind of very poetic mirroring of the API with the models. This, I kind of view this as like kind of the merging of assistance API and chat completions, right. Into one unified responses. So it's kind of like how GPT and the old series models are also unifying.Romain [00:03:48]: Yeah, that's exactly the right, uh, that's the right framing, right? Like, I think we took the best of what we learned from the assistance API, especially like being able to access tools very, uh, very like conveniently, but at the same time, like simplifying the way you have to integrate, like, you no longer have to think about six different objects to kind of get access to these tools with the responses API. You just get one API request and suddenly you can weave in those tools, right?Nikunj [00:04:12]: Yeah, absolutely. And I think we're going to make it really easy and straightforward for assistance API users to migrate over to responsive. Right. To the API without any loss of functionality or data. So our plan is absolutely to add, you know, assistant like objects and thread light objects to that, that work really well with the responses API. We'll also add like the code interpreter tool, which is not launching today, but it'll come soon. And, uh, we'll add async mode to responses API, because that's another difference with, with, uh, assistance. I will have web hooks and stuff like that, but I think it's going to be like a pretty smooth transition. Uh, once we have all of that in place. And we'll be. Like a full year to migrate and, and help them through any issues they, they, they face. So overall, I feel like assistance users are really going to benefit from this longer term, uh, with this more flexible, primitive.Alessio [00:05:01]: How should people think about when to use each type of API? So I know that in the past, the assistance was maybe more stateful, kind of like long running, many tool use kind of like file based things. And the chat completions is more stateless, you know, kind of like traditional completion API. Is that still the mental model that people should have? Or like, should you buy the.Nikunj [00:05:20]: So the responses API is going to support everything that it's at launch, going to support everything that chat completion supports, and then over time, it's going to support everything that assistance supports. So it's going to be a pretty good fit for anyone starting out with open AI. Uh, they should be able to like go to responses responses, by the way, also has a stateless mode, so you can pass in store false and they'll make the whole API stateless, just like chat completions. You're really trying to like get this unification. A story in so that people don't have to juggle multiple endpoints. That being said, like chat completions, just like the most widely adopted API, it's it's so popular. So we're still going to like support it for years with like new models and features. But if you're a new user, you want to or if you want to like existing, you want to tap into some of these like built in tools or something, you should feel feel totally fine migrating to responses and you'll have more capabilities and performance than the tech completions.swyx [00:06:16]: I think the messaging that I agree that I think resonated the most. When I talked to you was that it is a strict superset, right? Like you should be able to do everything that you could do in chat completions and with assistants. And the thing that I just assumed that because you're you're now, you know, by default is stateful, you're actually storing the chat logs or the chat state. I thought you'd be charging me for it. So, you know, to me, it was very surprising that you figured out how to make it free.Nikunj [00:06:43]: Yeah, it's free. We store your state for 30 days. You can turn it off. But yeah, it's it's free. And the interesting thing on state is that it just like makes particularly for me, it makes like debugging things and building things so much simpler, where I can like create a responses object that's like pretty complicated and part of this more complex application that I've built, I can just go into my dashboard and see exactly what happened that mess up my prompt that is like not called one of these tools that misconfigure one of the tools like the visual observability of everything that you're doing is so, so helpful. So I'm excited, like about people trying that out and getting benefits from it, too.swyx [00:07:19]: Yeah, it's a it's really, I think, a really nice to have. But all I'll say is that my friend Corey Quinn says that anything that can be used as a database will be used as a database. So be prepared for some abuse.Romain [00:07:34]: All right. Yeah, that's a good one. Some of that I've tried with the metadata. That's some people are very, very creative at stuffing data into an object. Yeah.Nikunj [00:07:44]: And we do have metadata with responses. Exactly. Yeah.Alessio [00:07:48]: Let's get through it. All of these. So web search. I think the when I first said web search, I thought you were going to just expose a API that then return kind of like a nice list of thing. But the way it's name is like GPD for all search preview. So I'm guessing you have you're using basically the same model that is in the chat GPD search, which is fine tune for search. I'm guessing it's a different model than the base one. And it's impressive the jump in performance. So just to give an example, in simple QA, GPD for all is 38% accuracy for all search is 90%. But we always talk about. How tools are like models is not everything you need, like tools around it are just as important. So, yeah, maybe give people a quick review on like the work that went into making this special.Nikunj [00:08:29]: Should I take that?Alessio [00:08:29]: Yeah, go for it.Nikunj [00:08:30]: So firstly, we're launching web search in two ways. One in responses API, which is our API for tools. It's going to be available as a web search tool itself. So you'll be able to go tools, turn on web search and you're ready to go. We still wanted to give chat completions people access to real time information. So in that. Chat completions API, which does not support built in tools. We're launching the direct access to the fine tuned model that chat GPD for search uses, and we call it GPD for search preview. And how is this model built? Basically, we have our search research team has been working on this for a while. Their main goal is to, like, get information, like get a bunch of information from all of our data sources that we use to gather information for search and then pick the right things and then cite them. As accurately as possible. And that's what the search team has really focused on. They've done some pretty cool stuff. They use like synthetic data techniques. They've done like all series model distillation to, like, make these four or fine tunes really good. But yeah, the main thing is, like, can it remain factual? Can it answer questions based on what it retrieves and get cited accurately? And that's what this like fine tune model really excels at. And so, yeah, so we're excited that, like, it's going to be directly available in chat completions along with being available as a tool. Yeah.Alessio [00:09:49]: Just to clarify, if I'm using the responses API, this is a tool. But if I'm using chat completions, I have to switch model. I cannot use 01 and call search as a tool. Yeah, that's right. Exactly.Romain [00:09:58]: I think what's really compelling, at least for me and my own uses of it so far, is that when you use, like, web search as a tool, it combines nicely with every other tool and every other feature of the platform. So think about this for a second. For instance, imagine you have, like, a responses API call with the web search tool, but suddenly you turn on function calling. You also turn on, let's say, structure. So you can have, like, the ability to structure any data from the web in real time in the JSON schema that you need for your application. So it's quite powerful when you start combining those features and tools together. It's kind of like an API for the Internet almost, you know, like you get, like, access to the precise schema you need for your app. Yeah.Alessio [00:10:39]: And then just to wrap up on the infrastructure side of it, I read on the post that people, publisher can choose to appear in the web search. So are people by default in it? Like, how can we get Latent Space in the web search API?Nikunj [00:10:53]: Yeah. Yeah. I think we have some documentation around how websites, publishers can control, like, what shows up in a web search tool. And I think you should be able to, like, read that. I think we should be able to get Latent Space in for sure. Yeah.swyx [00:11:10]: You know, I think so. I compare this to a broader trend that I started covering last year of online LLMs. Actually, Perplexity, I think, was the first. It was the first to say, to offer an API that is connected to search, and then Gemini had the sort of search grounding API. And I think you guys, I actually didn't, I missed this in the original reading of the docs, but you even give like citations with like the exact sub paragraph that is matching, which I think is the standard nowadays. I think my question is, how do we take what a knowledge cutoff is for something like this, right? Because like now, basically there's no knowledge cutoff is always live, but then there's a difference between what the model has sort of internalized in its back propagation and what is searching up its rag.Romain [00:11:53]: I think it kind of depends on the use case, right? And what you want to showcase as the source. Like, for instance, you take a company like Hebbia that has used this like web search tool. They can combine like for credit firms or law firms, they can find like, you know, public information from the internet with the live sources and citation that sometimes you do want to have access to, as opposed to like the internal knowledge. But if you're building something different, well, like, you just want to have the information. If you want to have an assistant that relies on the deep knowledge that the model has, you may not need to have these like direct citations. So I think it kind of depends on the use case a little bit, but there are many, uh, many companies like Hebbia that will need that access to these citations to precisely know where the information comes from.swyx [00:12:34]: Yeah, yeah, uh, for sure. And then one thing on the, on like the breadth, you know, I think a lot of the deep research, open deep research implementations have this sort of hyper parameter about, you know, how deep they're searching and how wide they're searching. I don't see that in the docs. But is that something that we can tune? Is that something you recommend thinking about?Nikunj [00:12:53]: Super interesting. It's definitely not a parameter today, but we should explore that. It's very interesting. I imagine like how you would do it with the web search tool and responsive API is you would have some form of like, you know, agent orchestration over here where you have a planning step and then each like web search call that you do like explicitly goes a layer deeper and deeper and deeper. But it's not a parameter that's available out of the box. But it's a cool. It's a cool thing to think about. Yeah.swyx [00:13:19]: The only guidance I'll offer there is a lot of these implementations offer top K, which is like, you know, top 10, top 20, but actually don't really want that. You want like sort of some kind of similarity cutoff, right? Like some matching score cuts cutoff, because if there's only five things, five documents that match fine, if there's 500 that match, maybe that's what I want. Right. Yeah. But also that might, that might make my costs very unpredictable because the costs are something like $30 per a thousand queries, right? So yeah. Yeah.Nikunj [00:13:49]: I guess you could, you could have some form of like a context budget and then you're like, go as deep as you can and pick the best stuff and put it into like X number of tokens. There could be some creative ways of, of managing cost, but yeah, that's a super interesting thing to explore.Alessio [00:14:05]: Do you see people using the files and the search API together where you can kind of search and then store everything in the file so the next time I'm not paying for the search again and like, yeah, how should people balance that?Nikunj [00:14:17]: That's actually a very interesting question. And let me first tell you about how I've seen a really cool way I've seen people use files and search together is they put their user preferences or memories in the vector store and so a query comes in, you use the file search tool to like get someone's like reading preferences or like fashion preferences and stuff like that, and then you search the web for information or products that they can buy related to those preferences and you then render something beautiful to show them, like, here are five things that you might be interested in. So that's how I've seen like file search, web search work together. And by the way, that's like a single responses API call, which is really cool. So you just like configure these things, go boom, and like everything just happens. But yeah, that's how I've seen like files and web work together.Romain [00:15:01]: But I think that what you're pointing out is like interesting, and I'm sure developers will surprise us as they always do in terms of how they combine these tools and how they might use file search as a way to have memory and preferences, like Nikum says. But I think like zooming out, what I find very compelling and powerful here is like when you have these like neural networks. That have like all of the knowledge that they have today, plus real time access to the Internet for like any kind of real time information that you might need for your app and file search, where you can have a lot of company, private documents, private details, you combine those three, and you have like very, very compelling and precise answers for any kind of use case that your company or your product might want to enable.swyx [00:15:41]: It's a difference between sort of internal documents versus the open web, right? Like you're going to need both. Exactly, exactly. I never thought about it doing memory as well. I guess, again, you know, anything that's a database, you can store it and you will use it as a database. That sounds awesome. But I think also you've been, you know, expanding the file search. You have more file types. You have query optimization, custom re-ranking. So it really seems like, you know, it's been fleshed out. Obviously, I haven't been paying a ton of attention to the file search capability, but it sounds like your team has added a lot of features.Nikunj [00:16:14]: Yeah, metadata filtering was like the main thing people were asking us for for a while. And I'm super excited about it. I mean, it's just so critical once your, like, web store size goes over, you know, more than like, you know, 5,000, 10,000 records, you kind of need that. So, yeah, metadata filtering is coming, too.Romain [00:16:31]: And for most companies, it's also not like a competency that you want to rebuild in-house necessarily, you know, like, you know, thinking about embeddings and chunking and, you know, how of that, like, it sounds like very complex for something very, like, obvious to ship for your users. Like companies like Navant, for instance. They were able to build with the file search, like, you know, take all of the FAQ and travel policies, for instance, that you have, you, you put that in file search tool, and then you don't have to think about anything. Now your assistant becomes naturally much more aware of all of these policies from the files.swyx [00:17:03]: The question is, like, there's a very, very vibrant RAG industry already, as you well know. So there's many other vector databases, many other frameworks. Probably if it's an open source stack, I would say like a lot of the AI engineers that I talk to want to own this part of the stack. And it feels like, you know, like, when should we DIY and when should we just use whatever OpenAI offers?Nikunj [00:17:24]: Yeah. I mean, like, if you're doing something completely from scratch, you're going to have more control, right? Like, so super supportive of, you know, people trying to, like, roll up their sleeves, build their, like, super custom chunking strategy and super custom retrieval strategy and all of that. And those are things that, like, will be harder to do with OpenAI tools. OpenAI tool has, like, we have an out-of-the-box solution. We give you the tools. We use some knobs to customize things, but it's more of, like, a managed RAG service. So my recommendation would be, like, start with the OpenAI thing, see if it, like, meets your needs. And over time, we're going to be adding more and more knobs to make it even more customizable. But, you know, if you want, like, the completely custom thing, you want control over every single thing, then you'd probably want to go and hand roll it using other solutions. So we're supportive of both, like, engineers should pick. Yeah.Alessio [00:18:16]: And then we got computer use. Which I think Operator was obviously one of the hot releases of the year. And we're only two months in. Let's talk about that. And that's also, it seems like a separate model that has been fine-tuned for Operator that has browser access.Nikunj [00:18:31]: Yeah, absolutely. I mean, the computer use models are exciting. The cool thing about computer use is that we're just so, so early. It's like the GPT-2 of computer use or maybe GPT-1 of computer use right now. But it is a separate model that has been, you know, the computer. The computer use team has been working on, you send it screenshots and it tells you what action to take. So the outputs of it are almost always tool calls and you're inputting screenshots based on whatever computer you're trying to operate.Romain [00:19:01]: Maybe zooming out for a second, because like, I'm sure your audience is like super, super like AI native, obviously. But like, what is computer use as a tool, right? And what's operator? So the idea for computer use is like, how do we let developers also build agents that can complete tasks for the users, but using a computer? Okay. Or a browser instead. And so how do you get that done? And so that's why we have this custom model, like optimized for computer use that we use like for operator ourselves. But the idea behind like putting it as an API is that imagine like now you want to, you want to automate some tasks for your product or your own customers. Then now you can, you can have like the ability to spin up one of these agents that will look at the screen and act on the screen. So that means able, the ability to click, the ability to scroll. The ability to type and to report back on the action. So that's what we mean by computer use and wrapping it as a tool also in the responses API. So now like that gives a hint also at the multi-turned thing that we were hinting at earlier, the idea that like, yeah, maybe one of these actions can take a couple of minutes to complete because there's maybe like 20 steps to complete that task. But now you can.swyx [00:20:08]: Do you think a computer use can play Pokemon?Romain [00:20:11]: Oh, interesting. I guess we tried it. I guess we should try it. You know?swyx [00:20:17]: Yeah. There's a lot of interest. I think Pokemon really is a good agent benchmark, to be honest. Like it seems like Claude is, Claude is running into a lot of trouble.Romain [00:20:25]: Sounds like we should make that a new eval, it looks like.swyx [00:20:28]: Yeah. Yeah. Oh, and then one more, one more thing before we move on to agents SDK. I know you have a hard stop. There's all these, you know, blah, blah, dash preview, right? Like search preview, computer use preview, right? And you see them all like fine tunes of 4.0. I think the question is, are we, are they all going to be merged into the main branch or are we basically always going to have subsets? Of these models?Nikunj [00:20:49]: Yeah, I think in the early days, research teams at OpenAI like operate with like fine tune models. And then once the thing gets like more stable, we sort of merge it into the main line. So that's definitely the vision, like going out of preview as we get more comfortable with and learn about all the developer use cases and we're doing a good job at them. We'll sort of like make them part of like the core models so that you don't have to like deal with the bifurcation.Romain [00:21:12]: You should think of it this way as exactly what happened last year when we introduced vision capabilities, you know. Yes. Vision capabilities were in like a vision preview model based off of GPT-4 and then vision capabilities now are like obviously built into GPT-4.0. You can think about it the same way for like the other modalities like audio and those kind of like models, like optimized for search and computer use.swyx [00:21:34]: Agents SDK, we have a few minutes left. So let's just assume that everyone has looked at Swarm. Sure. I think that Swarm has really popularized the handoff technique, which I thought was like, you know, really, really interesting for sort of a multi-agent. What is new with the SDK?Nikunj [00:21:50]: Yeah. Do you want to start? Yeah, for sure. So we've basically added support for types. We've made this like a lot. Yeah. Like we've added support for types. We've added support for guard railing, which is a very common pattern. So in the guardrail example, you basically have two things happen in parallel. The guardrail can sort of block the execution. It's a type of like optimistic generation that happens. And I think we've added support for tracing. So I think that's really cool. So you can basically look at the traces that the Agents SDK creates in the OpenAI dashboard. We also like made this pretty flexible. So you can pick any API from any provider that supports the ChatCompletions API format. So it supports responses by default, but you can like easily plug it in to anyone that uses the ChatCompletions API. And similarly, on the tracing side, you can support like multiple tracing providers. By default, it sort of points to the OpenAI dashboard. But, you know, there's like so many tracing providers. There's so many tracing companies out there. And we'll announce some partnerships on that front, too. So just like, you know, adding lots of core features and making it more usable, but still centered around like handoffs is like the main, main concept.Romain [00:22:59]: And by the way, it's interesting, right? Because Swarm just came to life out of like learning from customers directly that like orchestrating agents in production was pretty hard. You know, simple ideas could quickly turn very complex. Like what are those guardrails? What are those handoffs, et cetera? So that came out of like learning from customers. And it was initially shipped. It was not as a like low-key experiment, I'd say. But we were kind of like taken by surprise at how much momentum there was around this concept. And so we decided to learn from that and embrace it. To be like, okay, maybe we should just embrace that as a core primitive of the OpenAI platform. And that's kind of what led to the Agents SDK. And I think now, as Nikuj mentioned, it's like adding all of these new capabilities to it, like leveraging the handoffs that we had, but tracing also. And I think what's very compelling for developers is like instead of having one agent to rule them all and you stuff like a lot of tool calls in there that can be hard to monitor, now you have the tools you need to kind of like separate the logic, right? And you can have a triage agent that based on an intent goes to different kind of agents. And then on the OpenAI dashboard, we're releasing a lot of new user interface logs as well. So you can see all of the tracing UIs. Essentially, you'll be able to troubleshoot like what exactly happened. In that workflow, when the triage agent did a handoff to a secondary agent and the third and see the tool calls, et cetera. So we think that the Agents SDK combined with the tracing UIs will definitely help users and developers build better agentic workflows.Alessio [00:24:28]: And just before we wrap, are you thinking of connecting this with also the RFT API? Because I know you already have, you kind of store my text completions and then I can do fine tuning of that. Is that going to be similar for agents where you're storing kind of like my traces? And then help me improve the agents?Nikunj [00:24:43]: Yeah, absolutely. Like you got to tie the traces to the evals product so that you can generate good evals. Once you have good evals and graders and tasks, you can use that to do reinforcement fine tuning. And, you know, lots of details to be figured out over here. But that's the vision. And I think we're going to go after it like pretty hard and hope we can like make this whole workflow a lot easier for developers.Alessio [00:25:05]: Awesome. Thank you so much for the time. I'm sure you'll be busy on Twitter tomorrow with all the developer feedback. Yeah.Romain [00:25:12]: Thank you so much for having us. And as always, we can't wait to see what developers will build with these tools and how we can like learn as quickly as we can from them to make them even better over time.Nikunj [00:25:21]: Yeah.Romain [00:25:22]: Thank you, guys.Nikunj [00:25:23]: Thank you.Romain [00:25:23]: Thank you both. Awesome. Get full access to Latent.Space at www.latent.space/subscribe

  --------  

  25:38
• ⚡️How Claude 3.7 Plays Pokémon

  Special lightning pod with David Hershey from Anthropic, the person behind Claude Plays Pokémon. Sonnet 3.7 is currently trying to complete Pokémon Red live on Twitch thanks to a special harness that David built so that it can see the screen, navigate through it, remember facts about the game, and more. (Since recording, it has successfully escaped Mt Moon! You can follow along on Twitch: https://www.twitch.tv/claudeplayspokemon) Get full access to Latent.Space at www.latent.space/subscribe

  --------  

  37:38
• Open Operator, Serverless Browsers and the Future of Computer-Using Agents

  Today's episode is with Paul Klein, founder of Browserbase. We talked about building browser infrastructure for AI agents, the future of agent authentication, and their open source framework Stagehand.* [00:00:00] Introductions* [00:04:46] AI-specific challenges in browser infrastructure* [00:07:05] Multimodality in AI-Powered Browsing* [00:12:26] Running headless browsers at scale* [00:18:46] Geolocation when proxying* [00:21:25] CAPTCHAs and Agent Auth* [00:28:21] Building “User take over” functionality* [00:33:43] Stagehand: AI web browsing framework* [00:38:58] OpenAI's Operator and computer use agents* [00:44:44] Surprising use cases of Browserbase* [00:47:18] Future of browser automation and market competition* [00:53:11] Being a solo founderTranscriptAlessio [00:00:04]: Hey everyone, welcome to the Latent Space podcast. This is Alessio, partner and CTO at Decibel Partners, and I'm joined by my co-host Swyx, founder of Smol.ai.swyx [00:00:12]: Hey, and today we are very blessed to have our friends, Paul Klein, for the fourth, the fourth, CEO of Browserbase. Welcome.Paul [00:00:21]: Thanks guys. Yeah, I'm happy to be here. I've been lucky to know both of you for like a couple of years now, I think. So it's just like we're hanging out, you know, with three ginormous microphones in front of our face. It's totally normal hangout.swyx [00:00:34]: Yeah. We've actually mentioned you on the podcast, I think, more often than any other Solaris tenant. Just because like you're one of the, you know, best performing, I think, LLM tool companies that have started up in the last couple of years.Paul [00:00:50]: Yeah, I mean, it's been a whirlwind of a year, like Browserbase is actually pretty close to our first birthday. So we are one years old. And going from, you know, starting a company as a solo founder to... To, you know, having a team of 20 people, you know, a series A, but also being able to support hundreds of AI companies that are building AI applications that go out and automate the web. It's just been like, really cool. It's been happening a little too fast. I think like collectively as an AI industry, let's just take a week off together. I took my first vacation actually two weeks ago, and Operator came out on the first day, and then a week later, DeepSeat came out. And I'm like on vacation trying to chill. I'm like, we got to build with this stuff, right? So it's been a breakneck year. But I'm super happy to be here and like talk more about all the stuff we're seeing. And I'd love to hear kind of what you guys are excited about too, and share with it, you know?swyx [00:01:39]: Where to start? So people, you've done a bunch of podcasts. I think I strongly recommend Jack Bridger's Scaling DevTools, as well as Turner Novak's The Peel. And, you know, I'm sure there's others. So you covered your Twilio story in the past, talked about StreamClub, you got acquired to Mux, and then you left to start Browserbase. So maybe we just start with what is Browserbase? Yeah.Paul [00:02:02]: Browserbase is the web browser for your AI. We're building headless browser infrastructure, which are browsers that run in a server environment that's accessible to developers via APIs and SDKs. It's really hard to run a web browser in the cloud. You guys are probably running Chrome on your computers, and that's using a lot of resources, right? So if you want to run a web browser or thousands of web browsers, you can't just spin up a bunch of lambdas. You actually need to use a secure containerized environment. You have to scale it up and down. It's a stateful system. And that infrastructure is, like, super painful. And I know that firsthand, because at my last company, StreamClub, I was CTO, and I was building our own internal headless browser infrastructure. That's actually why we sold the company, is because Mux really wanted to buy our headless browser infrastructure that we'd built. And it's just a super hard problem. And I actually told my co-founders, I would never start another company unless it was a browser infrastructure company. And it turns out that's really necessary in the age of AI, when AI can actually go out and interact with websites, click on buttons, fill in forms. You need AI to do all of that work in an actual browser running somewhere on a server. And BrowserBase powers that.swyx [00:03:08]: While you're talking about it, it occurred to me, not that you're going to be acquired or anything, but it occurred to me that it would be really funny if you became the Nikita Beer of headless browser companies. You just have one trick, and you make browser companies that get acquired.Paul [00:03:23]: I truly do only have one trick. I'm screwed if it's not for headless browsers. I'm not a Go programmer. You know, I'm in AI grant. You know, browsers is an AI grant. But we were the only company in that AI grant batch that used zero dollars on AI spend. You know, we're purely an infrastructure company. So as much as people want to ask me about reinforcement learning, I might not be the best guy to talk about that. But if you want to ask about headless browser infrastructure at scale, I can talk your ear off. So that's really my area of expertise. And it's a pretty niche thing. Like, nobody has done what we're doing at scale before. So we're happy to be the experts.swyx [00:03:59]: You do have an AI thing, stagehand. We can talk about the sort of core of browser-based first, and then maybe stagehand. Yeah, stagehand is kind of the web browsing framework. Yeah.What is Browserbase? Headless Browser Infrastructure ExplainedAlessio [00:04:10]: Yeah. Yeah. And maybe how you got to browser-based and what problems you saw. So one of the first things I worked on as a software engineer was integration testing. Sauce Labs was kind of like the main thing at the time. And then we had Selenium, we had Playbrite, we had all these different browser things. But it's always been super hard to do. So obviously you've worked on this before. When you started browser-based, what were the challenges? What were the AI-specific challenges that you saw versus, there's kind of like all the usual running browser at scale in the cloud, which has been a problem for years. What are like the AI unique things that you saw that like traditional purchase just didn't cover? Yeah.AI-specific challenges in browser infrastructurePaul [00:04:46]: First and foremost, I think back to like the first thing I did as a developer, like as a kid when I was writing code, I wanted to write code that did stuff for me. You know, I wanted to write code to automate my life. And I do that probably by using curl or beautiful soup to fetch data from a web browser. And I think I still do that now that I'm in the cloud. And the other thing that I think is a huge challenge for me is that you can't just create a web site and parse that data. And we all know that now like, you know, taking HTML and plugging that into an LLM, you can extract insights, you can summarize. So it was very clear that now like dynamic web scraping became very possible with the rise of large language models or a lot easier. And that was like a clear reason why there's been more usage of headless browsers, which are necessary because a lot of modern websites don't expose all of their page content via a simple HTTP request. You know, they actually do require you to run this type of code for a specific time. JavaScript on the page to hydrate this. Airbnb is a great example. You go to airbnb.com. A lot of that content on the page isn't there until after they run the initial hydration. So you can't just scrape it with a curl. You need to have some JavaScript run. And a browser is that JavaScript engine that's going to actually run all those requests on the page. So web data retrieval was definitely one driver of starting BrowserBase and the rise of being able to summarize that within LLM. Also, I was familiar with if I wanted to automate a website, I could write one script and that would work for one website. It was very static and deterministic. But the web is non-deterministic. The web is always changing. And until we had LLMs, there was no way to write scripts that you could write once that would run on any website. That would change with the structure of the website. Click the login button. It could mean something different on many different websites. And LLMs allow us to generate code on the fly to actually control that. So I think that rise of writing the generic automation scripts that can work on many different websites, to me, made it clear that browsers are going to be a lot more useful because now you can automate a lot more things without writing. If you wanted to write a script to book a demo call on 100 websites, previously, you had to write 100 scripts. Now you write one script that uses LLMs to generate that script. That's why we built our web browsing framework, StageHand, which does a lot of that work for you. But those two things, web data collection and then enhanced automation of many different websites, it just felt like big drivers for more browser infrastructure that would be required to power these kinds of features.Alessio [00:07:05]: And was multimodality also a big thing?Paul [00:07:08]: Now you can use the LLMs to look, even though the text in the dome might not be as friendly. Maybe my hot take is I was always kind of like, I didn't think vision would be as big of a driver. For UI automation, I felt like, you know, HTML is structured text and large language models are good with structured text. But it's clear that these computer use models are often vision driven, and they've been really pushing things forward. So definitely being multimodal, like rendering the page is required to take a screenshot to give that to a computer use model to take actions on a website. And it's just another win for browser. But I'll be honest, that wasn't what I was thinking early on. I didn't even think that we'd get here so fast with multimodality. I think we're going to have to get back to multimodal and vision models.swyx [00:07:50]: This is one of those things where I forgot to mention in my intro that I'm an investor in Browserbase. And I remember that when you pitched to me, like a lot of the stuff that we have today, we like wasn't on the original conversation. But I did have my original thesis was something that we've talked about on the podcast before, which is take the GPT store, the custom GPT store, all the every single checkbox and plugin is effectively a startup. And this was the browser one. I think the main hesitation, I think I actually took a while to get back to you. The main hesitation was that there were others. Like you're not the first hit list browser startup. It's not even your first hit list browser startup. There's always a question of like, will you be the category winner in a place where there's a bunch of incumbents, to be honest, that are bigger than you? They're just not targeted at the AI space. They don't have the backing of Nat Friedman. And there's a bunch of like, you're here in Silicon Valley. They're not. I don't know.Paul [00:08:47]: I don't know if that's, that was it, but like, there was a, yeah, I mean, like, I think I tried all the other ones and I was like, really disappointed. Like my background is from working at great developer tools, companies, and nothing had like the Vercel like experience. Um, like our biggest competitor actually is partly owned by private equity and they just jacked up their prices quite a bit. And the dashboard hasn't changed in five years. And I actually used them at my last company and tried them and I was like, oh man, like there really just needs to be something that's like the experience of these great infrastructure companies, like Stripe, like clerk, like Vercel that I use in love, but oriented towards this kind of like more specific category, which is browser infrastructure, which is really technically complex. Like a lot of stuff can go wrong on the internet when you're running a browser. The internet is very vast. There's a lot of different configurations. Like there's still websites that only work with internet explorer out there. How do you handle that when you're running your own browser infrastructure? These are the problems that we have to think about and solve at BrowserBase. And it's, it's certainly a labor of love, but I built this for me, first and foremost, I know it's super cheesy and everyone says that for like their startups, but it really, truly was for me. If you look at like the talks I've done even before BrowserBase, and I'm just like really excited to try and build a category defining infrastructure company. And it's, it's rare to have a new category of infrastructure exists. We're here in the Chroma offices and like, you know, vector databases is a new category of infrastructure. Is it, is it, I mean, we can, we're in their office, so, you know, we can, we can debate that one later. That is one.Multimodality in AI-Powered Browsingswyx [00:10:16]: That's one of the industry debates.Paul [00:10:17]: I guess we go back to the LLMOS talk that Karpathy gave way long ago. And like the browser box was very clearly there and it seemed like the people who were building in this space also agreed that browsers are a core primitive of infrastructure for the LLMOS that's going to exist in the future. And nobody was building something there that I wanted to use. So I had to go build it myself.swyx [00:10:38]: Yeah. I mean, exactly that talk that, that honestly, that diagram, every box is a startup and there's the code box and then there's the. The browser box. I think at some point they will start clashing there. There's always the question of the, are you a point solution or are you the sort of all in one? And I think the point solutions tend to win quickly, but then the only ones have a very tight cohesive experience. Yeah. Let's talk about just the hard problems of browser base you have on your website, which is beautiful. Thank you. Was there an agency that you used for that? Yeah. Herb.paris.Paul [00:11:11]: They're amazing. Herb.paris. Yeah. It's H-E-R-V-E. I highly recommend for developers. Developer tools, founders to work with consumer agencies because they end up building beautiful things and the Parisians know how to build beautiful interfaces. So I got to give prep.swyx [00:11:24]: And chat apps, apparently are, they are very fast. Oh yeah. The Mistral chat. Yeah. Mistral. Yeah.Paul [00:11:31]: Late chat.swyx [00:11:31]: Late chat. And then your videos as well, it was professionally shot, right? The series A video. Yeah.Alessio [00:11:36]: Nico did the videos. He's amazing. Not the initial video that you shot at the new one. First one was Austin.Paul [00:11:41]: Another, another video pretty surprised. But yeah, I mean, like, I think when you think about how you talk about your company. You have to think about the way you present yourself. It's, you know, as a developer, you think you evaluate a company based on like the API reliability and the P 95, but a lot of developers say, is the website good? Is the message clear? Do I like trust this founder? I'm building my whole feature on. So I've tried to nail that as well as like the reliability of the infrastructure. You're right. It's very hard. And there's a lot of kind of foot guns that you run into when running headless browsers at scale. Right.Competing with Existing Headless Browser Solutionsswyx [00:12:10]: So let's pick one. You have eight features here. Seamless integration. Scalability. Fast or speed. Secure. Observable. Stealth. That's interesting. Extensible and developer first. What comes to your mind as like the top two, three hardest ones? Yeah.Running headless browsers at scalePaul [00:12:26]: I think just running headless browsers at scale is like the hardest one. And maybe can I nerd out for a second? Is that okay? I heard this is a technical audience, so I'll talk to the other nerds. Whoa. They were listening. Yeah. They're upset. They're ready. The AGI is angry. Okay. So. So how do you run a browser in the cloud? Let's start with that, right? So let's say you're using a popular browser automation framework like Puppeteer, Playwright, and Selenium. Maybe you've written a code, some code locally on your computer that opens up Google. It finds the search bar and then types in, you know, search for Latent Space and hits the search button. That script works great locally. You can see the little browser open up. You want to take that to production. You want to run the script in a cloud environment. So when your laptop is closed, your browser is doing something. The browser is doing something. Well, I, we use Amazon. You can see the little browser open up. You know, the first thing I'd reach for is probably like some sort of serverless infrastructure. I would probably try and deploy on a Lambda. But Chrome itself is too big to run on a Lambda. It's over 250 megabytes. So you can't easily start it on a Lambda. So you maybe have to use something like Lambda layers to squeeze it in there. Maybe use a different Chromium build that's lighter. And you get it on the Lambda. Great. It works. But it runs super slowly. It's because Lambdas are very like resource limited. They only run like with one vCPU. You can run one process at a time. Remember, Chromium is super beefy. It's barely running on my MacBook Air. I'm still downloading it from a pre-run. Yeah, from the test earlier, right? I'm joking. But it's big, you know? So like Lambda, it just won't work really well. Maybe it'll work, but you need something faster. Your users want something faster. Okay. Well, let's put it on a beefier instance. Let's get an EC2 server running. Let's throw Chromium on there. Great. Okay. I can, that works well with one user. But what if I want to run like 10 Chromium instances, one for each of my users? Okay. Well, I might need two EC2 instances. Maybe 10. All of a sudden, you have multiple EC2 instances. This sounds like a problem for Kubernetes and Docker, right? Now, all of a sudden, you're using ECS or EKS, the Kubernetes or container solutions by Amazon. You're spending up and down containers, and you're spending a whole engineer's time on kind of maintaining this stateful distributed system. Those are some of the worst systems to run because when it's a stateful distributed system, it means that you are bound by the connections to that thing. You have to keep the browser open while someone is working with it, right? That's just a painful architecture to run. And there's all this other little gotchas with Chromium, like Chromium, which is the open source version of Chrome, by the way. You have to install all these fonts. You want emojis working in your browsers because your vision model is looking for the emoji. You need to make sure you have the emoji fonts. You need to make sure you have all the right extensions configured, like, oh, do you want ad blocking? How do you configure that? How do you actually record all these browser sessions? Like it's a headless browser. You can't look at it. So you need to have some sort of observability. Maybe you're recording videos and storing those somewhere. It all kind of adds up to be this just giant monster piece of your project when all you wanted to do was run a lot of browsers in production for this little script to go to google.com and search. And when I see a complex distributed system, I see an opportunity to build a great infrastructure company. And we really abstract that away with Browserbase where our customers can use these existing frameworks, Playwright, Publisher, Selenium, or our own stagehand and connect to our browsers in a serverless-like way. And control them, and then just disconnect when they're done. And they don't have to think about the complex distributed system behind all of that. They just get a browser running anywhere, anytime. Really easy to connect to.swyx [00:15:55]: I'm sure you have questions. My standard question with anything, so essentially you're a serverless browser company, and there's been other serverless things that I'm familiar with in the past, serverless GPUs, serverless website hosting. That's where I come from with Netlify. One question is just like, you promised to spin up thousands of servers. You promised to spin up thousands of browsers in milliseconds. I feel like there's no real solution that does that yet. And I'm just kind of curious how. The only solution I know, which is to kind of keep a kind of warm pool of servers around, which is expensive, but maybe not so expensive because it's just CPUs. So I'm just like, you know. Yeah.Browsers as a Core Primitive in AI InfrastructurePaul [00:16:36]: You nailed it, right? I mean, how do you offer a serverless-like experience with something that is clearly not serverless, right? And the answer is, you need to be able to run... We run many browsers on single nodes. We use Kubernetes at browser base. So we have many pods that are being scheduled. We have to predictably schedule them up or down. Yes, thousands of browsers in milliseconds is the best case scenario. If you hit us with 10,000 requests, you may hit a slower cold start, right? So we've done a lot of work on predictive scaling and being able to kind of route stuff to different regions where we have multiple regions of browser base where we have different pools available. You can also pick the region you want to go to based on like lower latency, round trip, time latency. It's very important with these types of things. There's a lot of requests going over the wire. So for us, like having a VM like Firecracker powering everything under the hood allows us to be super nimble and spin things up or down really quickly with strong multi-tenancy. But in the end, this is like the complex infrastructural challenges that we have to kind of deal with at browser base. And we have a lot more stuff on our roadmap to allow customers to have more levers to pull to exchange, do you want really fast browser startup times or do you want really low costs? And if you're willing to be more flexible on that, we may be able to kind of like work better for your use cases.swyx [00:17:44]: Since you used Firecracker, shouldn't Fargate do that for you or did you have to go lower level than that? We had to go lower level than that.Paul [00:17:51]: I find this a lot with Fargate customers, which is alarming for Fargate. We used to be a giant Fargate customer. Actually, the first version of browser base was ECS and Fargate. And unfortunately, it's a great product. I think we were actually the largest Fargate customer in our region for a little while. No, what? Yeah, seriously. And unfortunately, it's a great product, but I think if you're an infrastructure company, you actually have to have a deeper level of control over these primitives. I think it's the same thing is true with databases. We've used other database providers and I think-swyx [00:18:21]: Yeah, serverless Postgres.Paul [00:18:23]: Shocker. When you're an infrastructure company, you're on the hook if any provider has an outage. And I can't tell my customers like, hey, we went down because so-and-so went down. That's not acceptable. So for us, we've really moved to bringing things internally. It's kind of opposite of what we preach. We tell our customers, don't build this in-house, but then we're like, we build a lot of stuff in-house. But I think it just really depends on what is in the critical path. We try and have deep ownership of that.Alessio [00:18:46]: On the distributed location side, how does that work for the web where you might get sort of different content in different locations, but the customer is expecting, you know, if you're in the US, I'm expecting the US version. But if you're spinning up my browser in France, I might get the French version. Yeah.Paul [00:19:02]: Yeah. That's a good question. Well, generally, like on the localization, there is a thing called locale in the browser. You can set like what your locale is. If you're like in the ENUS browser or not, but some things do IP, IP based routing. And in that case, you may want to have a proxy. Like let's say you're running something in the, in Europe, but you want to make sure you're showing up from the US. You may want to use one of our proxy features so you can turn on proxies to say like, make sure these connections always come from the United States, which is necessary too, because when you're browsing the web, you're coming from like a, you know, data center IP, and that can make things a lot harder to browse web. So we do have kind of like this proxy super network. Yeah. We have a proxy for you based on where you're going, so you can reliably automate the web. But if you get scheduled in Europe, that doesn't happen as much. We try and schedule you as close to, you know, your origin that you're trying to go to. But generally you have control over the regions you can put your browsers in. So you can specify West one or East one or Europe. We only have one region of Europe right now, actually. Yeah.Alessio [00:19:55]: What's harder, the browser or the proxy? I feel like to me, it feels like actually proxying reliably at scale. It's much harder than spending up browsers at scale. I'm curious. It's all hard.Paul [00:20:06]: It's layers of hard, right? Yeah. I think it's different levels of hard. I think the thing with the proxy infrastructure is that we work with many different web proxy providers and some are better than others. Some have good days, some have bad days. And our customers who've built browser infrastructure on their own, they have to go and deal with sketchy actors. Like first they figure out their own browser infrastructure and then they got to go buy a proxy. And it's like you can pay in Bitcoin and it just kind of feels a little sus, right? It's like you're buying drugs when you're trying to get a proxy online. We have like deep relationships with these counterparties. We're able to audit them and say, is this proxy being sourced ethically? Like it's not running on someone's TV somewhere. Is it free range? Yeah. Free range organic proxies, right? Right. We do a level of diligence. We're SOC 2. So we have to understand what is going on here. But then we're able to make sure that like we route around proxy providers not working. There's proxy providers who will just, the proxy will stop working all of a sudden. And then if you don't have redundant proxying on your own browsers, that's hard down for you or you may get some serious impacts there. With us, like we intelligently know, hey, this proxy is not working. Let's go to this one. And you can kind of build a network of multiple providers to really guarantee the best uptime for our customers. Yeah. So you don't own any proxies? We don't own any proxies. You're right. The team has been saying who wants to like take home a little proxy server, but not yet. We're not there yet. You know?swyx [00:21:25]: It's a very mature market. I don't think you should build that yourself. Like you should just be a super customer of them. Yeah. Scraping, I think, is the main use case for that. I guess. Well, that leads us into CAPTCHAs and also off, but let's talk about CAPTCHAs. You had a little spiel that you wanted to talk about CAPTCHA stuff.Challenges of Scaling Browser InfrastructurePaul [00:21:43]: Oh, yeah. I was just, I think a lot of people ask, if you're thinking about proxies, you're thinking about CAPTCHAs too. I think it's the same thing. You can go buy CAPTCHA solvers online, but it's the same buying experience. It's some sketchy website, you have to integrate it. It's not fun to buy these things and you can't really trust that the docs are bad. What Browserbase does is we integrate a bunch of different CAPTCHAs. We do some stuff in-house, but generally we just integrate with a bunch of known vendors and continually monitor and maintain these things and say, is this working or not? Can we route around it or not? These are CAPTCHA solvers. CAPTCHA solvers, yeah. Not CAPTCHA providers, CAPTCHA solvers. Yeah, sorry. CAPTCHA solvers. We really try and make sure all of that works for you. I think as a dev, if I'm buying infrastructure, I want it all to work all the time and it's important for us to provide that experience by making sure everything does work and monitoring it on our own. Yeah. Right now, the world of CAPTCHAs is tricky. I think AI agents in particular are very much ahead of the internet infrastructure. CAPTCHAs are designed to block all types of bots, but there are now good bots and bad bots. I think in the future, CAPTCHAs will be able to identify who a good bot is, hopefully via some sort of KYC. For us, we've been very lucky. We have very little to no known abuse of Browserbase because we really look into who we work with. And for certain types of CAPTCHA solving, we only allow them on certain types of plans because we want to make sure that we can know what people are doing, what their use cases are. And that's really allowed us to try and be an arbiter of good bots, which is our long term goal. I want to build great relationships with people like Cloudflare so we can agree, hey, here are these acceptable bots. We'll identify them for you and make sure we flag when they come to your website. This is a good bot, you know?Alessio [00:23:23]: I see. And Cloudflare said they want to do more of this. So they're going to set by default, if they think you're an AI bot, they're going to reject. I'm curious if you think this is something that is going to be at the browser level or I mean, the DNS level with Cloudflare seems more where it should belong. But I'm curious how you think about it.Paul [00:23:40]: I think the web's going to change. You know, I think that the Internet as we have it right now is going to change. And we all need to just accept that the cat is out of the bag. And instead of kind of like wishing the Internet was like it was in the 2000s, we can have free content line that wouldn't be scraped. It's just it's not going to happen. And instead, we should think about like, one, how can we change? How can we change the models of, you know, information being published online so people can adequately commercialize it? But two, how do we rebuild applications that expect that AI agents are going to log in on their behalf? Those are the things that are going to allow us to kind of like identify good and bad bots. And I think the team at Clerk has been doing a really good job with this on the authentication side. I actually think that auth is the biggest thing that will prevent agents from accessing stuff, not captchas. And I think there will be agent auth in the future. I don't know if it's going to happen from an individual company, but actually authentication providers that have a, you know, hidden login as agent feature, which will then you put in your email, you'll get a push notification, say like, hey, your browser-based agent wants to log into your Airbnb. You can approve that and then the agent can proceed. That really circumvents the need for captchas or logging in as you and sharing your password. I think agent auth is going to be one way we identify good bots going forward. And I think a lot of this captcha solving stuff is really short-term problems as the internet kind of reorients itself around how it's going to work with agents browsing the web, just like people do. Yeah.Managing Distributed Browser Locations and Proxiesswyx [00:24:59]: Stitch recently was on Hacker News for talking about agent experience, AX, which is a thing that Netlify is also trying to clone and coin and talk about. And we've talked about this on our previous episodes before in a sense that I actually think that's like maybe the only part of the tech stack that needs to be kind of reinvented for agents. Everything else can stay the same, CLIs, APIs, whatever. But auth, yeah, we need agent auth. And it's mostly like short-lived, like it should not, it should be a distinct, identity from the human, but paired. I almost think like in the same way that every social network should have your main profile and then your alt accounts or your Finsta, it's almost like, you know, every, every human token should be paired with the agent token and the agent token can go and do stuff on behalf of the human token, but not be presumed to be the human. Yeah.Paul [00:25:48]: It's like, it's, it's actually very similar to OAuth is what I'm thinking. And, you know, Thread from Stitch is an investor, Colin from Clerk, Octaventures, all investors in browser-based because like, I hope they solve this because they'll make browser-based submission more possible. So we don't have to overcome all these hurdles, but I think it will be an OAuth-like flow where an agent will ask to log in as you, you'll approve the scopes. Like it can book an apartment on Airbnb, but it can't like message anybody. And then, you know, the agent will have some sort of like role-based access control within an application. Yeah. I'm excited for that.swyx [00:26:16]: The tricky part is just, there's one, one layer of delegation here, which is like, you're authoring my user's user or something like that. I don't know if that's tricky or not. Does that make sense? Yeah.Paul [00:26:25]: You know, actually at Twilio, I worked on the login identity and access. Management teams, right? So like I built Twilio's login page.swyx [00:26:31]: You were an intern on that team and then you became the lead in two years? Yeah.Paul [00:26:34]: Yeah. I started as an intern in 2016 and then I was the tech lead of that team. How? That's not normal. I didn't have a life. He's not normal. Look at this guy. I didn't have a girlfriend. I just loved my job. I don't know. I applied to 500 internships for my first job and I got rejected from every single one of them except for Twilio and then eventually Amazon. And they took a shot on me and like, I was getting paid money to write code, which was my dream. Yeah. Yeah. I'm very lucky that like this coding thing worked out because I was going to be doing it regardless. And yeah, I was able to kind of spend a lot of time on a team that was growing at a company that was growing. So it informed a lot of this stuff here. I think these are problems that have been solved with like the SAML protocol with SSO. I think it's a really interesting stuff with like WebAuthn, like these different types of authentication, like schemes that you can use to authenticate people. The tooling is all there. It just needs to be tweaked a little bit to work for agents. And I think the fact that there are companies that are already. Providing authentication as a service really sets it up. Well, the thing that's hard is like reinventing the internet for agents. We don't want to rebuild the internet. That's an impossible task. And I think people often say like, well, we'll have this second layer of APIs built for agents. I'm like, we will for the top use cases, but instead of we can just tweak the internet as is, which is on the authentication side, I think we're going to be the dumb ones going forward. Unfortunately, I think AI is going to be able to do a lot of the tasks that we do online, which means that it will be able to go to websites, click buttons on our behalf and log in on our behalf too. So with this kind of like web agent future happening, I think with some small structural changes, like you said, it feels like it could all slot in really nicely with the existing internet.Handling CAPTCHAs and Agent Authenticationswyx [00:28:08]: There's one more thing, which is the, your live view iframe, which lets you take, take control. Yeah. Obviously very key for operator now, but like, was, is there anything interesting technically there or that the people like, well, people always want this.Paul [00:28:21]: It was really hard to build, you know, like, so, okay. Headless browsers, you don't see them, right. They're running. They're running in a cloud somewhere. You can't like look at them. And I just want to really make, it's a weird name. I wish we came up with a better name for this thing, but you can't see them. Right. But customers don't trust AI agents, right. At least the first pass. So what we do with our live view is that, you know, when you use browser base, you can actually embed a live view of the browser running in the cloud for your customer to see it working. And that's what the first reason is the build trust, like, okay, so I have this script. That's going to go automate a website. I can embed it into my web application via an iframe and my customer can watch. I think. And then we added two way communication. So now not only can you watch the browser kind of being operated by AI, if you want to pause and actually click around type within this iframe that's controlling a browser, that's also possible. And this is all thanks to some of the lower level protocol, which is called the Chrome DevTools protocol. It has a API called start screencast, and you can also send mouse clicks and button clicks to a remote browser. And this is all embeddable within iframes. You have a browser within a browser, yo. And then you simulate the screen, the click on the other side. Exactly. And this is really nice often for, like, let's say, a capture that can't be solved. You saw this with Operator, you know, Operator actually uses a different approach. They use VNC. So, you know, you're able to see, like, you're seeing the whole window here. What we're doing is something a little lower level with the Chrome DevTools protocol. It's just PNGs being streamed over the wire. But the same thing is true, right? Like, hey, I'm running a window. Pause. Can you do something in this window? Human. Okay, great. Resume. Like sometimes 2FA tokens. Like if you get that text message, you might need a person to type that in. Web agents need human-in-the-loop type workflows still. You still need a person to interact with the browser. And building a UI to proxy that is kind of hard. You may as well just show them the whole browser and say, hey, can you finish this up for me? And then let the AI proceed on afterwards. Is there a future where I stream my current desktop to browser base? I don't think so. I think we're very much cloud infrastructure. Yeah. You know, but I think a lot of the stuff we're doing, we do want to, like, build tools. Like, you know, we'll talk about the stage and, you know, web agent framework in a second. But, like, there's a case where a lot of people are going desktop first for, you know, consumer use. And I think cloud is doing a lot of this, where I expect to see, you know, MCPs really oriented around the cloud desktop app for a reason, right? Like, I think a lot of these tools are going to run on your computer because it makes... I think it's breaking out. People are putting it on a server. Oh, really? Okay. Well, sweet. We'll see. We'll see that. I was surprised, though, wasn't I? I think that the browser company, too, with Dia Browser, it runs on your machine. You know, it's going to be...swyx [00:30:50]: What is it?Paul [00:30:51]: So, Dia Browser, as far as I understand... I used to use Arc. Yeah. I haven't used Arc. But I'm a big fan of the browser company. I think they're doing a lot of cool stuff in consumer. As far as I understand, it's a browser where you have a sidebar where you can, like, chat with it and it can control the local browser on your machine. So, if you imagine, like, what a consumer web agent is, which it lives alongside your browser, I think Google Chrome has Project Marina, I think. I almost call it Project Marinara for some reason. I don't know why. It's...swyx [00:31:17]: No, I think it's someone really likes the Waterworld. Oh, I see. The classic Kevin Costner. Yeah.Paul [00:31:22]: Okay. Project Marinara is a similar thing to the Dia Browser, in my mind, as far as I understand it. You have a browser that has an AI interface that will take over your mouse and keyboard and control the browser for you. Great for consumer use cases. But if you're building applications that rely on a browser and it's more part of a greater, like, AI app experience, you probably need something that's more like infrastructure, not a consumer app.swyx [00:31:44]: Just because I have explored a little bit in this area, do people want branching? So, I have the state. Of whatever my browser's in. And then I want, like, 100 clones of this state. Do people do that? Or...Paul [00:31:56]: People don't do it currently. Yeah. But it's definitely something we're thinking about. I think the idea of forking a browser is really cool. Technically, kind of hard. We're starting to see this in code execution, where people are, like, forking some, like, code execution, like, processes or forking some tool calls or branching tool calls. Haven't seen it at the browser level yet. But it makes sense. Like, if an AI agent is, like, using a website and it's not sure what path it wants to take to crawl this website. To find the information it's looking for. It would make sense for it to explore both paths in parallel. And that'd be a very, like... A road not taken. Yeah. And hopefully find the right answer. And then say, okay, this was actually the right one. And memorize that. And go there in the future. On the roadmap. For sure. Don't make my roadmap, please. You know?Alessio [00:32:37]: How do you actually do that? Yeah. How do you fork? I feel like the browser is so stateful for so many things.swyx [00:32:42]: Serialize the state. Restore the state. I don't know.Paul [00:32:44]: So, it's one of the reasons why we haven't done it yet. It's hard. You know? Like, to truly fork, it's actually quite difficult. The naive way is to open the same page in a new tab and then, like, hope that it's at the same thing. But if you have a form halfway filled, you may have to, like, take the whole, you know, container. Pause it. All the memory. Duplicate it. Restart it from there. It could be very slow. So, we haven't found a thing. Like, the easy thing to fork is just, like, copy the page object. You know? But I think there needs to be something a little bit more robust there. Yeah.swyx [00:33:12]: So, MorphLabs has this infinite branch thing. Like, wrote a custom fork of Linux or something that let them save the system state and clone it. MorphLabs, hit me up. I'll be a customer. Yeah. That's the only. I think that's the only way to do it. Yeah. Like, unless Chrome has some special API for you. Yeah.Paul [00:33:29]: There's probably something we'll reverse engineer one day. I don't know. Yeah.Alessio [00:33:32]: Let's talk about StageHand, the AI web browsing framework. You have three core components, Observe, Extract, and Act. Pretty clean landing page. What was the idea behind making a framework? Yeah.Stagehand: AI web browsing frameworkPaul [00:33:43]: So, there's three frameworks that are very popular or already exist, right? Puppeteer, Playwright, Selenium. Those are for building hard-coded scripts to control websites. And as soon as I started to play with LLMs plus browsing, I caught myself, you know, code-genning Playwright code to control a website. I would, like, take the DOM. I'd pass it to an LLM. I'd say, can you generate the Playwright code to click the appropriate button here? And it would do that. And I was like, this really should be part of the frameworks themselves. And I became really obsessed with SDKs that take natural language as part of, like, the API input. And that's what StageHand is. StageHand exposes three APIs, and it's a super set of Playwright. So, if you go to a page, you may want to take an action, click on the button, fill in the form, etc. That's what the act command is for. You may want to extract some data. This one takes a natural language, like, extract the winner of the Super Bowl from this page. You can give it a Zod schema, so it returns a structured output. And then maybe you're building an API. You can do an agent loop, and you want to kind of see what actions are possible on this page before taking one. You can do observe. So, you can observe the actions on the page, and it will generate a list of actions. You can guide it, like, give me actions on this page related to buying an item. And you can, like, buy it now, add to cart, view shipping options, and pass that to an LLM, an agent loop, to say, what's the appropriate action given this high-level goal? So, StageHand isn't a web agent. It's a framework for building web agents. And we think that agent loops are actually pretty close to the application layer because every application probably has different goals or different ways it wants to take steps. I don't think I've seen a generic. Maybe you guys are the experts here. I haven't seen, like, a really good AI agent framework here. Everyone kind of has their own special sauce, right? I see a lot of developers building their own agent loops, and they're using tools. And I view StageHand as the browser tool. So, we expose act, extract, observe. Your agent can call these tools. And from that, you don't have to worry about it. You don't have to worry about generating playwright code performantly. You don't have to worry about running it. You can kind of just integrate these three tool calls into your agent loop and reliably automate the web.swyx [00:35:48]: A special shout-out to Anirudh, who I met at your dinner, who I think listens to the pod. Yeah. Hey, Anirudh.Paul [00:35:54]: Anirudh's a man. He's a StageHand guy.swyx [00:35:56]: I mean, the interesting thing about each of these APIs is they're kind of each startup. Like, specifically extract, you know, Firecrawler is extract. There's, like, Expand AI. There's a whole bunch of, like, extract companies. They just focus on extract. I'm curious. Like, I feel like you guys are going to collide at some point. Like, right now, it's friendly. Everyone's in a blue ocean. At some point, it's going to be valuable enough that there's some turf battle here. I don't think you have a dog in a fight. I think you can mock extract to use an external service if they're better at it than you. But it's just an observation that, like, in the same way that I see each option, each checkbox in the side of custom GBTs becoming a startup or each box in the Karpathy chart being a startup. Like, this is also becoming a thing. Yeah.Paul [00:36:41]: I mean, like, so the way StageHand works is that it's MIT-licensed, completely open source. You bring your own API key to your LLM of choice. You could choose your LLM. We don't make any money off of the extract or really. We only really make money if you choose to run it with our browser. You don't have to. You can actually use your own browser, a local browser. You know, StageHand is completely open source for that reason. And, yeah, like, I think if you're building really complex web scraping workflows, I don't know if StageHand is the tool for you. I think it's really more if you're building an AI agent that needs a few general tools or if it's doing a lot of, like, web automation-intensive work. But if you're building a scraping company, StageHand is not your thing. You probably want something that's going to, like, get HTML content, you know, convert that to Markdown, query it. That's not what StageHand does. StageHand is more about reliability. I think we focus a lot on reliability and less so on cost optimization and speed at this point.swyx [00:37:33]: I actually feel like StageHand, so the way that StageHand works, it's like, you know, page.act, click on the quick start. Yeah. It's kind of the integration test for the code that you would have to write anyway, like the Puppeteer code that you have to write anyway. And when the page structure changes, because it always does, then this is still the test. This is still the test that I would have to write. Yeah. So it's kind of like a testing framework that doesn't need implementation detail.Paul [00:37:56]: Well, yeah. I mean, Puppeteer, Playwright, and Slenderman were all designed as testing frameworks, right? Yeah. And now people are, like, hacking them together to automate the web. I would say, and, like, maybe this is, like, me being too specific. But, like, when I write tests, if the page structure changes. Without me knowing, I want that test to fail. So I don't know if, like, AI, like, regenerating that. Like, people are using StageHand for testing. But it's more for, like, usability testing, not, like, testing of, like, does the front end, like, has it changed or not. Okay. But generally where we've seen people, like, really, like, take off is, like, if they're using, you know, something. If they want to build a feature in their application that's kind of like Operator or Deep Research, they're using StageHand to kind of power that tool calling in their own agent loop. Okay. Cool.swyx [00:38:37]: So let's go into Operator, the first big agent launch of the year from OpenAI. Seems like they have a whole bunch scheduled. You were on break and your phone blew up. What's your just general view of computer use agents is what they're calling it. The overall category before we go into Open Operator, just the overall promise of Operator. I will observe that I tried it once. It was okay. And I never tried it again.OpenAI's Operator and computer use agentsPaul [00:38:58]: That tracks with my experience, too. Like, I'm a huge fan of the OpenAI team. Like, I think that I do not view Operator as the company. I'm not a company killer for browser base at all. I think it actually shows people what's possible. I think, like, computer use models make a lot of sense. And I'm actually most excited about computer use models is, like, their ability to, like, really take screenshots and reasoning and output steps. I think that using mouse click or mouse coordinates, I've seen that proved to be less reliable than I would like. And I just wonder if that's the right form factor. What we've done with our framework is anchor it to the DOM itself, anchor it to the actual item. So, like, if it's clicking on something, it's clicking on that thing, you know? Like, it's more accurate. No matter where it is. Yeah, exactly. Because it really ties in nicely. And it can handle, like, the whole viewport in one go, whereas, like, Operator can only handle what it sees. Can you hover? Is hovering a thing that you can do? I don't know if we expose it as a tool directly, but I'm sure there's, like, an API for hovering. Like, move mouse to this position. Yeah, yeah, yeah. I think you can trigger hover, like, via, like, the JavaScript on the DOM itself. But, no, I think, like, when we saw computer use, everyone's eyes lit up because they realized, like, wow, like, AI is going to actually automate work for people. And I think seeing that kind of happen from both of the labs, and I'm sure we're going to see more labs launch computer use models, I'm excited to see all the stuff that people build with it. I think that I'd love to see computer use power, like, controlling a browser on browser base. And I think, like, Open Operator, which was, like, our open source version of OpenAI's Operator, was our first take on, like, how can we integrate these models into browser base? And we handle the infrastructure and let the labs do the models. I don't have a sense that Operator will be released as an API. I don't know. Maybe it will. I'm curious to see how well that works because I think it's going to be really hard for a company like OpenAI to do things like support CAPTCHA solving or, like, have proxies. Like, I think it's hard for them structurally. Imagine this New York Times headline, OpenAI CAPTCHA solving. Like, that would be a pretty bad headline, this New York Times headline. Browser base solves CAPTCHAs. No one cares. No one cares. And, like, our investors are bored. Like, we're all okay with this, you know? We're building this company knowing that the CAPTCHA solving is short-lived until we figure out how to authenticate good bots. I think it's really hard for a company like OpenAI, who has this brand that's so, so good, to balance with, like, the icky parts of web automation, which it can be kind of complex to solve. I'm sure OpenAI knows who to call whenever they need you. Yeah, right. I'm sure they'll have a great partnership.Alessio [00:41:23]: And is Open Operator just, like, a marketing thing for you? Like, how do you think about resource allocation? So, you can spin this up very quickly. And now there's all this, like, open deep research, just open all these things that people are building. We started it, you know. You're the original Open. We're the original Open operator, you know? Is it just, hey, look, this is a demo, but, like, we'll help you build out an actual product for yourself? Like, are you interested in going more of a product route? That's kind of the OpenAI way, right? They started as a model provider and then…Paul [00:41:53]: Yeah, we're not interested in going the product route yet. I view Open Operator as a model provider. It's a reference project, you know? Let's show people how to build these things using the infrastructure and models that are out there. And that's what it is. It's, like, Open Operator is very simple. It's an agent loop. It says, like, take a high-level goal, break it down into steps, use tool calling to accomplish those steps. It takes screenshots and feeds those screenshots into an LLM with the step to generate the right action. It uses stagehand under the hood to actually execute this action. It doesn't use a computer use model. And it, like, has a nice interface using the live view that we talked about, the iframe, to embed that into an application. So I felt like people on launch day wanted to figure out how to build their own version of this. And we turned that around really quickly to show them. And I hope we do that with other things like deep research. We don't have a deep research launch yet. I think David from AOMNI actually has an amazing open deep research that he launched. It has, like, 10K GitHub stars now. So he's crushing that. But I think if people want to build these features natively into their application, they need good reference projects. And I think Open Operator is a good example of that.swyx [00:42:52]: I don't know. Actually, I'm actually pretty bullish on API-driven operator. Because that's the only way that you can sort of, like, once it's reliable enough, obviously. And now we're nowhere near. But, like, give it five years. It'll happen, you know. And then you can sort of spin this up and browsers are working in the background and you don't necessarily have to know. And it just is booking restaurants for you, whatever. I can definitely see that future happening. I had this on the landing page here. This might be a slightly out of order. But, you know, you have, like, sort of three use cases for browser base. Open Operator. Or this is the operator sort of use case. It's kind of like the workflow automation use case. And it completes with UiPath in the sort of RPA category. Would you agree with that? Yeah, I would agree with that. And then there's Agents we talked about already. And web scraping, which I imagine would be the bulk of your workload right now, right?Paul [00:43:40]: No, not at all. I'd say actually, like, the majority is browser automation. We're kind of expensive for web scraping. Like, I think that if you're building a web scraping product, if you need to do occasional web scraping or you have to do web scraping that works every single time, you want to use browser automation. Yeah. You want to use browser-based. But if you're building web scraping workflows, what you should do is have a waterfall. You should have the first request is a curl to the website. See if you can get it without even using a browser. And then the second request may be, like, a scraping-specific API. There's, like, a thousand scraping APIs out there that you can use to try and get data. Scraping B. Scraping B is a great example, right? Yeah. And then, like, if those two don't work, bring out the heavy hitter. Like, browser-based will 100% work, right? It will load the page in a real browser, hydrate it. I see.swyx [00:44:21]: Because a lot of people don't render to JS.swyx [00:44:25]: Yeah, exactly.Paul [00:44:26]: So, I mean, the three big use cases, right? Like, you know, automation, web data collection, and then, you know, if you're building anything agentic that needs, like, a browser tool, you want to use browser-based.Alessio [00:44:35]: Is there any use case that, like, you were super surprised by that people might not even think about? Oh, yeah. Or is it, yeah, anything that you can share? The long tail is crazy. Yeah.Surprising use cases of BrowserbasePaul [00:44:44]: One of the case studies on our website that I think is the most interesting is this company called Benny. So, the way that it works is if you're on food stamps in the United States, you can actually get rebates if you buy certain things. Yeah. You buy some vegetables. You submit your receipt to the government. They'll give you a little rebate back. Say, hey, thanks for buying vegetables. It's good for you. That process of submitting that receipt is very painful. And the way Benny works is you use their app to take a photo of your receipt, and then Benny will go submit that receipt for you and then deposit the money into your account. That's actually using no AI at all. It's all, like, hard-coded scripts. They maintain the scripts. They've been doing a great job. And they build this amazing consumer app. But it's an example of, like, all these, like, tedious workflows that people have to do to kind of go about their business. And they're doing it for the sake of their day-to-day lives. And I had never known about, like, food stamp rebates or the complex forms you have to do to fill them. But the world is powered by millions and millions of tedious forms, visas. You know, Emirate Lighthouse is a customer, right? You know, they do the O1 visa. Millions and millions of forms are taking away humans' time. And I hope that Browserbase can help power software that automates away the web forms that we don't need anymore. Yeah.swyx [00:45:49]: I mean, I'm very supportive of that. I mean, forms. I do think, like, government itself is a big part of it. I think the government itself should embrace AI more to do more sort of human-friendly form filling. Mm-hmm. But I'm not optimistic. I'm not holding my breath. Yeah. We'll see. Okay. I think I'm about to zoom out. I have a little brief thing on computer use, and then we can talk about founder stuff, which is, I tend to think of developer tooling markets in impossible triangles, where everyone starts in a niche, and then they start to branch out. So I already hinted at a little bit of this, right? We mentioned more. We mentioned E2B. We mentioned Firecrawl. And then there's Browserbase. So there's, like, all this stuff of, like, have serverless virtual computer that you give to an agent and let them do stuff with it. And there's various ways of connecting it to the internet. You can just connect to a search API, like SERP API, whatever other, like, EXA is another one. That's what you're searching. You can also have a JSON markdown extractor, which is Firecrawl. Or you can have a virtual browser like Browserbase, or you can have a virtual machine like Morph. And then there's also maybe, like, a virtual sort of code environment, like Code Interpreter. So, like, there's just, like, a bunch of different ways to tackle the problem of give a computer to an agent. And I'm just kind of wondering if you see, like, everyone's just, like, happily coexisting in their respective niches. And as a developer, I just go and pick, like, a shopping basket of one of each. Or do you think that you eventually, people will collide?Future of browser automation and market competitionPaul [00:47:18]: I think that currently it's not a zero-sum market. Like, I think we're talking about... I think we're talking about all of knowledge work that people do that can be automated online. All of these, like, trillions of hours that happen online where people are working. And I think that there's so much software to be built that, like, I tend not to think about how these companies will collide. I just try to solve the problem as best as I can and make this specific piece of infrastructure, which I think is an important primitive, the best I possibly can. And yeah. I think there's players that are actually going to like it. I think there's players that are going to launch, like, over-the-top, you know, platforms, like agent platforms that have all these tools built in, right? Like, who's building the rippling for agent tools that has the search tool, the browser tool, the operating system tool, right? There are some. There are some. There are some, right? And I think in the end, what I have seen as my time as a developer, and I look at all the favorite tools that I have, is that, like, for tools and primitives with sufficient levels of complexity, you need to have a solution that's really bespoke to that primitive, you know? And I am sufficiently convinced that the browser is complex enough to deserve a primitive. Obviously, I have to. I'm the founder of BrowserBase, right? I'm talking my book. But, like, I think maybe I can give you one spicy take against, like, maybe just whole OS running. I think that when I look at computer use when it first came out, I saw that the majority of use cases for computer use were controlling a browser. And do we really need to run an entire operating system just to control a browser? I don't think so. I don't think that's necessary. You know, BrowserBase can run browsers for way cheaper than you can if you're running a full-fledged OS with a GUI, you know, operating system. And I think that's just an advantage of the browser. It is, like, browsers are little OSs, and you can run them very efficiently if you orchestrate it well. And I think that allows us to offer 90% of the, you know, functionality in the platform needed at 10% of the cost of running a full OS. Yeah.Open Operator: Browserbase's Open-Source Alternativeswyx [00:49:16]: I definitely see the logic in that. There's a Mark Andreessen quote. I don't know if you know this one. Where he basically observed that the browser is turning the operating system into a poorly debugged set of device drivers, because most of the apps are moved from the OS to the browser. So you can just run browsers.Paul [00:49:31]: There's a place for OSs, too. Like, I think that there are some applications that only run on Windows operating systems. And Eric from pig.dev in this upcoming YC batch, or last YC batch, like, he's building all run tons of Windows operating systems for you to control with your agent. And like, there's some legacy EHR systems that only run on Internet-controlled systems. Yeah.Paul [00:49:54]: I think that's it. I think, like, there are use cases for specific operating systems for specific legacy software. And like, I'm excited to see what he does with that. I just wanted to give a shout out to the pig.dev website.swyx [00:50:06]: The pigs jump when you click on them. Yeah. That's great.Paul [00:50:08]: Eric, he's the former co-founder of banana.dev, too.swyx [00:50:11]: Oh, that Eric. Yeah. That Eric. Okay. Well, he abandoned bananas for pigs. I hope he doesn't start going around with pigs now.Alessio [00:50:18]: Like he was going around with bananas. A little toy pig. Yeah. Yeah. I love that. What else are we missing? I think we covered a lot of, like, the browser-based product history, but. What do you wish people asked you? Yeah.Paul [00:50:29]: I wish people asked me more about, like, what will the future of software look like? Because I think that's really where I've spent a lot of time about why do browser-based. Like, for me, starting a company is like a means of last resort. Like, you shouldn't start a company unless you absolutely have to. And I remain convinced that the future of software is software that you're going to click a button and it's going to do stuff on your behalf. Right now, software. You click a button and it maybe, like, calls it back an API and, like, computes some numbers. It, like, modifies some text, whatever. But the future of software is software using software. So, I may log into my accounting website for my business, click a button, and it's going to go load up my Gmail, search my emails, find the thing, upload the receipt, and then comment it for me. Right? And it may use it using APIs, maybe a browser. I don't know. I think it's a little bit of both. But that's completely different from how we've built software so far. And that's. I think that future of software has different infrastructure requirements. It's going to require different UIs. It's going to require different pieces of infrastructure. I think the browser infrastructure is one piece that fits into that, along with all the other categories you mentioned. So, I think that it's going to require developers to think differently about how they've built software for, you know, application level so far. And I am excited to kind of explore more what that means. And I think we've seen from, like, you know, the customers that use Browsway so far, some really innovative ways to, like, take software and really read it. And I think, like, re-imagine it for AI and build things that, like, have chat interfaces, build things that have human loop flows, build things that are more asynchronous because AI is slower. And those are patterns that are still emerging. And I don't think we have all the best practices yet.Key Use Cases for Browserbase: Automation, Agents, and Scrapingswyx [00:52:03]: I don't have much feedback on that. Like, that's true. Paul's right. Paul's right. You heard it here first. Quoted by Swyx. Yeah. Amazing. I'm framing that. It is not specific enough to be wrong.Paul [00:52:12]: That means Paul's right to me still.swyx [00:52:14]: I don't know if I'm hearing that wrong. I always try to prompt people for falsifiable problems. I think I'm just trying to make sure that I'm not making false predictions. Because, like, you can predict that things will be better generically, but how? And, like, those are the things where you, like, put a little skin in the game where…Paul [00:52:28]: Yeah. I mean, I can predict that Browsways will be a billion dollar company one day. So let's check back in five years and, you know, if I'm a PM at Coinbase, then something went wrong. Oh, boy.swyx [00:52:40]: Yeah. Yeah. We picked out a couple of your tweets about Foundry. Yeah. I think you're a pretty building public kind of guy. Yeah. I try to be. I think the main thing that I want to highlight as well is, you emphasized this at the start of your intro, which is you're a solo founder. I think that there's a movement towards more solo founders in the Valley more generally, but people who are hearing this for the first time have no idea. They're like, what do you mean? YC forces me to get a co-founder. Like, what is this? So I've heard you talk about this before, but maybe you want to recap your spiel for folks that haven't heard about it. Yeah. Yeah.Being a solo founderPaul [00:53:11]: I mean, I've had co-founders in my past company. I love my co-founders. They're my wedding. I think if you want to move extremely fast as a company, one of the hard parts about having co-founders is that there's like, you have to do the co-founder alignment and then the company alignment. And then there's people on the team that probably tell things to one co-founder because they have a favorite. And then like that co-founders represent their interests. Matt Brasway is a benevolent dictatorship. You know, like if I want to make a change, I work with the team and we all decide together. We move quickly. We don't have an extra layer of buy-in within the co-founder layer. Yeah. And frankly, like I think, especially with DevTools companies, if you're able to talk about your product and talk with customers and you can build product, you don't need to have a business guy or a business side. You know, I'm a developer first and foremost. I was raised by two salespeople, so I guess that's why I can talk to customers or something. But at my core- What kind of sales? I love, they did semiconductor and pharmaceutical sales. My mom and dad. Oh, very different. Yeah. Very different.swyx [00:54:08]: But also very enterprise. Good. Yeah.Paul [00:54:10]: Yeah. Yeah. Yeah. Yeah. I mean, like, it rubbed off on me in some way. I was just trying to play WoW as a kid and they made me play sports. So I don't know how it worked out the way it did, but it does all come back to like, as a solo founder, you need to be willing to like go out there and, you know, talk about your product, go talk to customers, go convince people to work for you, but then also have core principles of like how you want to build this company and like what product you want to build. And thankfully, if you can do all of that, you can be a solo founder. You just have to hire fast and put the right team around you. Yeah. And that's kind of the team that we do that's surrounding me and kind of lifting the whole company up.Alessio [00:54:44]: So there's kind of like the decision making and then there's like the culture of a company. Obviously as a solo founder, you have huge influence on everybody. Apple is maybe the usual example of like, you know, you have the Jobs and Wozniak. None of like, you can have two co-founders that are like each polarizing.Unexpected Use Cases of Browser Automationswyx [00:55:01]: There was a third co-founder, by the way.Alessio [00:55:02]: Who was the third co-founder?swyx [00:55:03]: I don't know. He sold his chairs very early on. Nobody talks about him, but he's like, he always has a, has a bit of a regret.Alessio [00:55:10]: Okay. But anyway. Yeah. How have you thought about building the culture? You know, obviously startups are like super intense, but you're also going to just run yourself to the ground all the time. Any insight doing it solo? Yeah.Paul [00:55:21]: I mean, like I talked about like how it's easier for me to make decisions being a solo founder. The real cheat code is like having a great team that you give a lot of agency and ownership to. A lot of people make the little tiny decisions that go into everything that makes Biospace great. Like the website, for example, I, I had some, like some involvement with that, but like a lot of that was the team. Right. And then the product. I think the team really has ownership of all, a lot of these day-to-day decisions that add up to make a cohesive product experience culturally, like we're fully in person. Maybe that's one crazy take that we do, but we're also like not too in person. Like our first meetings at 10 AM, people leave around five or six. We work Monday to Friday in person and those like, that's the, the expectation, right? I think people have gone too far with in-person where they're like seven days a week in the office, 9 AM to 9 PM.swyx [00:56:10]: That's too much. Just an anecdote. Yeah. I just visited an office. I'll keep them anonymous for now, but to my face, we are 9, 9, 6. Yeah. For those who don't know, 9, 9, 6 is 9 AM to 9 PM, six days a week.Paul [00:56:20]: I think we've taken it a little too far and for some teams, I know another anonymous company that does something like 9, 9, 6 and they're like crushing it right now. Right. So like, and like, it does get results, but like, I think for our culture, we gather in person, we put pants on every day and go to the office so that we can all work together. Or shorts, I guess. Right. And then like, we all know we're going to work outside of, out of the office. We're going to work at home sometimes. We might come in on a weekend. The weekends are for fun work and that's really where we get to let people work on stuff that's not on the roadmap. And that empowers them to build something and bring it back to the team on Monday and say, look what I built. This is cool. Culturally, we're a lot of like former YCCTOs and like ex-founders or future founders. And I've just found that those people tend to be just really great early hires for a company. They, they get it. And I think for them, especially kind of the ex-YCCTOs. I see people who maybe didn't find PMF coming in and being at a company with PMF, it's such a refreshing thing for them because they can just come in and execute. And there's just so many clear things we have to go build. And if you're a talented engineer, being able to go build and make an impact every single day is like super fulfilling.swyx [00:57:25]: My question on the other hand is you also talk a lot about recruiting, especially in the podcast that you talk about. How come there's no browser-based recruiting agent? That's a good question.Paul [00:57:34]: I think it's because I don't do that much outbound. I do message people. Yeah. But a lot of it's now through referral. It's very like targeted. Like if I see somebody working on something really cool, I just message them. So I don't want like something trawling the web and like messaging every Kubernetes firecracker expert. I try and like look for them in my passive web browsing. And when I find somebody, I just want to like take the time personally, like say, Hey, I love what you're doing. I think it's really cool. And let's have a conversation. Yeah.swyx [00:58:03]: Off of Hacker News and other stuff. Yeah.Paul [00:58:05]: I love to hire off of Hacker News. Yeah.swyx [00:58:07]: Let you plug at the end. My attempt at this failed, which is I really hate LinkedIn Sales Navigator. I think that it is just grifting on top of people doing data entry for LinkedIn. And I hope that browser-based will someday help to kill LinkedIn Sales Navigator at this point.Paul [00:58:21]: I don't know if we will directly, but one of our customers definitely is trying to do that. So I think there's a couple that are on it. These AISDR companies are crushing it. Yup.Alessio [00:58:30]: The 996 company was an AISDR company.swyx [00:58:33]: There we go.Alessio [00:58:34]: Yeah. Very classic. This was great. Anything? Yeah. You got the run clubs too. What other things do you mix in, like both in the company culture and like the community culture? I know you bring people together. Yeah.Paul [00:58:45]: I think like we, like we try and build in public and like, like you can see a lot of the browser based people on Twitter. Every Monday we have a run club. People go running together. We don't run very fast, but it's like a good way to spend time together. I just look back fondly on my time being in person at my first company. And we have people like with a mix of people like are just early in career. People have been in the business for a long time. They've been in, you know, the workforce for 20, 30 years. So it's not just like a young people company, like it's a huge mix. But when you make people make a polarizing decision of like, I will come to an office five days a week, people then end up making more decisions that are aligned with a culture. So it's almost like if you can make your culture binary or you're in or out, it becomes easier to assimilate and like keep a cohesive culture. And I think it starts with being an office for us, but for other people it could be like moving or like using discord versus slack or like other like. Yeah. The, the binary decisions that people may have to make.swyx [00:59:36]: One thing I like asking founders is, you know, you're famously not an AI company or, you know, you, you serve AI companies, but you're not yourself a LLM sort of consuming company. But if you were though, what company would you start? What's what's like obviously a good idea.The Competitive Landscape of AI-Powered Browsing and AutomationPaul [00:59:50]: Yeah, I, I had this tweet like forever ago, which is like, there's so much money to be made in taking like proprietary research and then turning that into like an automation, which is obviously like a very like browser based inspired one. Like. Like listening to all the city halls or town hall meetings in like little towns and then knowing when they're going to like approve a new Walmart or something and then like buying up real estate around the Walmart because that will go up when they install this thing. So it's like really interesting to think about like how can you find new channels for data that will allow you to make like high alpha decisions and benefit you financially. So I think it's like some interesting stuff there, like just a bunch of conversations that happen in real life that are recorded, that are online, that you can go find using, you know, a web browser, of course. And then like making some interesting like decisions off of that. So I don't know, like I like browser stuff, like it's on brand, right? Like I have to, I'm consistent at least.Alessio [01:00:45]: Do not look at it on your phone through a native app, only look at it through the browser.swyx [01:00:49]: My favorite part of one of his videos, they had these guys holding this bee behind them while they were doing the demo. So it was like a really Easter egg. Yeah, that was stagehand, right?Paul [01:00:58]: Yeah, the stagehand video. It's not, they're not holding it. They're actually wearing these bee boxes on their heads. And we shot it like five times and poor Sean and Samil are like bobbing their heads back and forth with these bee boxes on because we can't afford special effects, man. It's really serious.swyx [01:01:13]: Good detail. Good effort detail there. Yeah. Thank you so much. Congrats on all your success.Paul [01:01:17]: Thanks for having me, guys. It's been a really good time.swyx [01:01:20]: Yeah, I'm sure we'll have you back again.Paul [01:01:21]: Yeah, I'd love to come back. Get full access to Latent.Space at www.latent.space/subscribe

  --------  

  1:01:33

Más podcasts de Tecnología

Podcasts a la moda de Tecnología

Acerca de Latent Space: The AI Engineer Podcast