AI + a16z Patrick Collison on Stripe’s Early Choices, Smalltalk, and What Comes After Coding
25 snips
Mar 24, 2026 
Patrick Collison, CEO of Stripe and lifelong programmer, reflects on his Smalltalk and Lisp roots. He recounts Stripe’s early Ruby and MongoDB choices and the long work of API design. He talks about using LLMs for research and coding. He explores AI’s role in productivity and ambitious efforts to train models that ‘program’ biology.
AI Snips
Chapters
Books
Transcript
Episode notes
Smalltalk Startup Taught Live Debugging Power
- Patrick Collison built his first startup in Smalltalk because its interactive environment let him edit code mid-request and resume execution.
- He found Smalltalk's live debugger and continuation-style web frameworks far superior to Rails for rapid iteration, though the company failed for other reasons.
MSN Bot Showed Early NLP Limits
- Patrick wrote an MSN Messenger bot using Lisp that used a Bayesian next-word predictor and sometimes held lengthy unsuspecting conversations.
- He experimented more with genetic algorithms and keyboard-layout optimization than neural nets back then, which influenced his early AI exposure.
Bring Back Runtime-Centric Development Environments
- Patrick argues development environments should be runtimes, not just text editors, reclaiming the Lisp-machine/Smalltalk model.
- He praises Mathematica and wants IDEs that combine runtime, editor, debugger, and profiling in one place.
