My father died of a heart attack 20 days before his 42nd birthday. It’s 16 days until my 42nd birthday. That VAR check nearly took me out.

im fully convinced that LLMs are not an actual net productivity boost (today) they remove the barrier to get started, but they create increasingly complex software which does not appear to be maintainable so far, in my situations, they appear to slow down long term velocity

GEPA for skills is here! Introducing gskill, an automated pipeline to learn agent skills with @gepa_ai. With learned skills, we boost Claude Code’s repository task resolution rate to near-perfect levels, while making it 47% faster. Here's how we did it:

We've been working on very long-running coding agents. In a recent week-long run, our system peaked at over 1,000 commits per hour across hundreds of agents. We're sharing our findings and an early research preview inside Cursor.

it is pretty telling that when you ride in a waymo you can’t give instructions to gemini to play a song or change destination or drive differently. when one of the great gilded tech monopolies of the world does not yet have a cohesive ai picture, what hope has the broader economy

I think GUIs are great (when done well)! But the bar where terminal applications are frustrating for the developer or the user is way, way too low. A world exists where new software is written in the terminal because it’s genuinely an exciting capable platform and that’s my goal.

IME really good software ppl who deeply care find the least use from LLM coding and are often ideologically opposed to it because they like to exert editorial control over every line. slop research coders such as myself don’t care as much and have much larger gains

Sam Altman says the perfect AI is “a very tiny model with superhuman reasoning, 1 trillion tokens of context, and access to every tool you can imagine.” It doesn't need to contain the knowledge - just the ability to think, search, simulate, and solve anything.