Post

At the risk of starting the flame war to end all flame wars... Modern LLMs (GPT-5.1, Claude 4.5, Gemini 3) produce excellent code and can be a significant productivity boost to software engineers who take the time to learn how to effectively apply them - especially if used with coding agent tools

(A few days ago I naïvely thought that this wasn't a controversial position any more, but recent events on Bluesky have shown that I was very much incorrect about that!)

Here's a tiny tool I just vibe coded (via Artifacts and Claude Opus 4.5) for entering text and getting it back in New Yorker style, for anyone who wants to type "naïve" tools.simonwillison.net/new-yorker-s... Claude transcript here: tools.simonwillison.net/colophon#new...

Weirdly there seems to be no option to share claude artifacts privately unless you have an institutional claude account It's a shame because building internal tools is a strong use case

Does no react reduce functionality significantly, in your experience?

no, it just reduces bundle size

That's a really interesting question - the kind of tools I build I don't think it does but I tend to stick with things with only one or two features - my main collection of 150 tools I've built this way is tools.simonwillison.net