"[AI coding tools like Cursor] increase my enjoyment, as a programmer" THIS is something we don't talk about enough. When a tool makes the job better - eg because it removes the toil that's a drag to do (e.g. refactoring, repetitive stuff) then it's already a big win!

My theory is that very large companies can’t run effective performance programs, so they must do large purges periodically. They know they have a bunch of remote employees working 3 jobs, poor performers that receive “good enough” performance reviews, C players riding the bloat.

A common mistake that AI companies make nowadays is to not give their engineers enough time and mental calm to do their best work. Constant deadlines, pressure and distractions from daily AI news are poison for writing good code and systems that scale well. That’s why most AI APIs and products have reliability issues. A good company culture that mixes excellence with focus and enough rest leads to faster and better results. The best example of how to do it well is the early Google culture from 1998 which resulted in one of the largest scale and most reliable services on the web in just a few short years. Founders should copy some of the strategies that Larry and Sergey used. They are still underrated IMO despite their huge reputation.

At a past company, the head of engineering and the principal engineers decided to break our Ruby on Rails application into a Go microservices mesh. They created very detailed design documents and architecture diagrams. They went all out and used Kubernetes, gRPC, service templates, the whole shebang. The whole senior engineering leadership came from Amazon, where they were used to each team owning a distinct service. They tried to apply that model directly. But our issues were with code ownership and poor domain modeling. The entire application could have run on just a handful of EC2 instances. What was the result? Five years later, 70% of the application is still running on the Ruby on Rails monolith. Never completed the migration. But now they have to maintain two systems. None of the original leadership works there anymore.

I don't know what labs are doing to these poor LLMs during RL but they are mortally terrified of exceptions, in any infinitesimally likely case. Exceptions are a normal part of life and healthy dev process. Sign my LLM welfare petition for improved rewards in cases of exceptions.

I sent this note to our entire team at @Opendoor earlier today.

Excited to share Vibes — a new feed in the Meta AI app for short-form, AI-generated videos.

Introducing study mode in ChatGPT — step by step guidance for students rather than quick answers: openai.com/index/chatgpt-stu…