Post

two views from Anthropic 1. Claude Skills are for collaboration 2. Skills are for continual learning

“spec driven development is broken…a spec isn’t just a more detailed prompt.”

Meta superintelligence researcher developing CWD (code world model) have big ambitions of problems that a CWD can solve by loosely simulating code execution in latent space - the halting problem - distributed systems behavior

btw yes i’m at a conference today bsky.app/profile/timk...

Will Brown (willccbb) of Prime Intellect talking about RL scaling, the other side. How do you scale up your workforce of AI researchers without actually paying more?: increase the pool of researchers

Will: RL environments are the web apps of AI research Cursor’s composer-1 and codemax are both trained in an RL env containing Cursor & codex respectively. Model & product intertwined Prime Intellect’s environment hub is like github for these environments

OpenAI is talking about their RL finetuning APIs they find that RFT is really good at teaching an agent how to call tools, when to do it in parallel, etc. Overall it’s good for squeezing out that last bit of efficiency

a theme that’s forming across sessions — RL environments are *extremely* sensitive, they have to look identical to your prod environment which ofc is why everyone has async cloud agents, that’s no mistake! Cursor, Codex, etc. cloud agents is most of the way to an RL environment

who is presenting?

this guy’s on fire

Software 2.0 relies on validation If your code base doesn’t have verification & controls that are as good or better than your senior dev, you’ll get slop

There's a little spin going on here. All these things are true for hiring regular human devs. The thrust of the argument is that vibe coding is equivalent to replacing all your junior engineers with worse ones. Presenting these challenges as new just helps fend off "the bots will get better" replies

this talk is really thought provoking building a code base that AI can work on easily is hard. But if you do it well, you can absolutely have highly parallel agents but then building these automated verifiers is then a full time job