Patrick Collison on Stripe’s Early Choices, Smalltalk, and What Comes After Coding - The a16z Show Recap

Podcast: The a16z Show

Published: 2026-02-20

Duration: 53 min

Summary

In this episode, Patrick Collison shares insights on the evolution of programming paradigms and the impact of early language choices on Stripe's success. He reflects on his experiences with Smalltalk and Ruby, discussing how these decisions shaped the company's development environment and overall trajectory.

What Happened

Patrick Collison opens up about his early programming experiences, particularly with Smalltalk, which he describes as 'the best programming language.' He recalls how its development environment allowed him to fix errors mid-request and inspect stack frames, offering a level of interactivity that he found lacking in other mainstream languages. This environment significantly influenced his decision-making when founding Stripe, where he and his brother ultimately opted for Ruby and MongoDB. Despite the challenges of using less mainstream languages, Collison highlights that learning them wasn't as difficult as one might expect for new hires.

As Stripe approaches its 15-year mark, it is now undergoing substantial changes by introducing V2 APIs and rewriting core abstractions that were originally designed over a decade ago. Collison emphasizes the complexity of this process, likening it more to an 'instruction set migration' than a straightforward product launch. He discusses the importance of API design and abstraction, noting that these choices have far-reaching business ramifications, which continue to define Stripe’s operations today.

The conversation extends beyond programming languages to touch on broader topics such as AI's impact on productivity and Collison's current work on foundational models for biology at ARC. He expresses a desire to return to a more interactive development environment reminiscent of his earlier experiences, suggesting that innovation in programming paradigms has stagnated over the past two decades. This episode ultimately blends nostalgia with forward-looking insights on the future of coding and technology.

Key Insights