Interested to see Anthropic do this sort of move, but it's very useful to see.

I see it on myself too. It feels too irresistible to start adding more features to software you develop with LLM agents. Everything feels like just a few prompts and will be done in half an hour. Why not add this too? Just another sentence in the prompt. Next thing you know you have more features than you remember and the AI starts to have a really hard job keeping it all functional.

Coding with AI requires immense restraint and strong scope limits.

Because they're not trying to build a leading coding agent - they're racing to AGI. That means cramming everything and anything, from every angle because general intelligence is happening next year and that'll connect all the dots, leading to hypergrowth and complete industry dominance, across all domains.

Seriously, if you listen to Dario talk, it's non-stop how there's a tsunami coming, people have no idea of what's ahead, how general intelligence or superintelligence is right around the corner. But also, this is super dangerous, and only Anthropic can save us from a doomsday scenario.

Is there a coding use-case for Claude that is not "spraying in all directions"?