Summary

In this podcast with Joe Reis, Wes McKinney describes his journey from existential dread about AI in early 2025 to becoming fully immersed in agentic software development. He details his elaborate multi-agent workflow using Claude Code, Codex, and Gemini, with a custom code review system called RoboRev that reviews every commit. He argues that code produced by current AI agents is fundamentally buggy and requires rigorous multi-agent review. He explains why Go has become his preferred language for agentic development due to fast build times, and introduces his concept of 'The Mythical Agent Month' — that coding agents excel at building software facades but struggle with the hard work of making products robust, scalable, and maintainable.

Key Insight

AI coding agents produce fundamentally buggy code that requires rigorous multi-agent adversarial review, and they excel at building software facades while struggling with the 9x harder work of turning prototypes into robust, maintainable products.

Spicy Quotes (click to share)

• 8

  If you're just committing and shipping the code that's coming out of Opus 4.6, that code is a bunch of hot garbage. It has to be really rigorously reviewed by other agents and different agent sessions to have any semblance of confidence that you're shipping something that isn't totally slop.

• 7

  Coding agents are very good at building the facade, the Potemkin's facade of the software. But when you actually go into the thorny issues associated with making it robust, making it scalable, making it maintainable by somebody other than its author — all these things can go wrong.

• 8

  The whole reason to use Python is that it's easy to read and write. So if I'm not reading or writing the code, what's the point?

• 9

  I put off learning Rust just long enough that I never have to learn it.

• 5

  I went from existential dread a year ago to being very stimulated and very locked in. I kind of prefer the being locked in because being in a state of existential dread and feeling like my career is over was not very fun.

• 3

  Coding agents don't care whether it's fun or not. They'll just do it for you. And if it doesn't work the first time, they'll keep trying until it's done.

• 7

  My sense is that Opus has been reinforcement-learned to have faster cycle times to give developers the feeling of forward momentum. Faster means worse because it's doing less reasoning. But it creates more engagement with the users.

• 4

  What I think is cool about coding agents now is that you can transform periods of divided or scattered attention into productive time.

Tone

opinionated, energized, pragmatic