Cheeky Pint

Bret Taylor, co-founder of Sierra and Chair of the OpenAI board, joins John for a pint to discuss the rapid shift toward an agentic future. In this episode, Bret explains why outcome-based pricing is the future of software business models, and why he believes the atomic unit of AI productivity is a process, not a person. They cover why big companies struggle to adopt AI because they are “shipping their org charts.” Bret also discusses a new type of hyper-generalist, reflects on his experience with the OpenAI and Twitter boards, and explains why he believes we might see the end of the smartphone era.

Timestamps
(00:00:26) Coding
(00:16:23) Sierra
(00:27:14) Agentic UX
(00:38:47) Building support agents
(00:45:43) Co-developing with the models
(00:50:08) SaaSpocalypse
(01:00:50) Stripe Sessions
(01:01:33) Outcome-based pricing
(01:09:14) Is Sierra short AGI?
(01:13:50) AI productivity
(01:23:47) How to structure a tech business
(01:30:25) Board drama
(01:38:24) AI predictions

Garrett Langley is the founder and CEO of Flock Safety, a public safety operating system that helps communities and law enforcement eliminate crime. He sits down with John to discuss why most crime is opportunistic, how Flock helps clear over one million crimes a year, and the engineering challenges of building solar-powered cameras and autonomous drones. They cover the shifting landscape of criminal technology, why hardware requires making "one-way door" decisions, and his vision for a future where technology prevents crime before it happens.
Timestamps
(00:00:19) Flock
(00:19:51) Safety vs privacy
(00:23:54) Crime and technology
(00:32:36) Crime rates
(00:43:56) Corporate security
(00:52:16) Stripe Radar
(00:52:54) Competitive landscape
(01:02:41) Drones
(01:09:01) The Flock business
(01:11:39) Building hardware
(01:20:01) Cameras
(01:25:17) PD procurement
(01:32:56) Building your own drones
(01:40:52) What’s next for Flock?
Books:
The Digital Silk Road: https://www.amazon.com/Digital-Silk-Road-Chinas-Future/dp/0063046288
Boyd: https://www.amazon.co.uk/Boyd-Fighter-Pilot-Who-Changed/dp/0316796883

Reiner Pope is the co-founder and CEO of MatX, designing specialized chips for Large Language Models. A former Google TPU architect, he joins John to discuss why the current generation of AI hardware is hitting a wall. They cover the "uncomfortable trade-off" between latency and throughput for current chips, why MatX is betting on combining HBM and SRAM to solve it, and the massive logistical challenge of manufacturing chips at scale with TSMC. Reiner also shares his predictions for AI in 2027, why he prefers Rust for hardware design, and why the best iteration loops happen in your head before writing a line of code.
Timestamps
(00:00:15) Google’s AI revival
(00:07:54) MatX
(00:17:11) AI supply chain
(00:21:48) Designing chips
(00:37:11) TSMC
(00:44:17) Token pricing
(00:44:55) RL-ing chip design
(00:49:26) Design to production
(00:56:05) MatX culture
(01:02:57) Rust
(01:05:21) Cuckoo hashing
(01:09:35) Unexplored model architectures

The internet economy accelerated in 2025. The fastest-growing companies got even faster, agentic commerce and stablecoin payments started to kick into gear, and total payment volume on Stripe grew by a third as our customers continued to prosper.
Our annual letter covers the trends that we think are worth paying close attention to as the pace of change accelerates.
Timestamps
(01:48) The sorting machine
(05:45) Global by default
(09:01) Stable progress
(12:14) Working capital’s working. Capital!
(15:09) Escaping from low revenue mode
(18:33) The five levels of agentic commerce
(23:52) A Republic of Permissions

Eric Glyman is the cofounder and CEO of Ramp, the finance automation platform that now powers over 2% of all corporate spend in the US. He sits down with John and co-host Alex Rampell to discuss how Ramp scaled to over $1 billion in revenue in just seven years, and why the future of fintech is "selling time, not money." They cover the "SaaS apocalypse" (and why lines of code are becoming a liability), how Ramp uses AI agents to review 100,000 expenses a day with 99% accuracy, and why their internal data suggests the US economy is much stronger than the Census Bureau reports.
Timestamps
(00:00:21) Ramp business today
(00:04:27) The *correct* expense policy
(00:11:07) Bill Pay
(00:16:52) AI and software
(00:32:23) Stablecoin-backed cards
(00:33:06) Ramp data
(00:36:25) How to cut your expenses
(00:41:13) Ramp strategy
(00:57:08) Capital One
(01:06:34) Treasury