hipstersmoothie.com

Shit i was literally in the middle of using it for perf work </3

Statement on the US government directive to suspend access to Fable 5 and Mythos 5 anthropic.com

I'm lazy so I automated implementing user feedback for @standard-reader.app @userinput.app -> @airglow.run -> GitHub Actions Script -> Claude Routine -> Pull Request -> Verify @railway.com + @neon.com Preview -> Merge and Deploy 🎉 Read about it here 👇

Atmospheric Automated Development hipstersmoothie.com

I think so?

Anthropic just allowed using your Claude sub in third party harnesses again. Today, it was planned that you would have to use extra usage credits to use Claude in third party harnesses and to use… | Anthony Kroeger linkedin.com

Guns for everyone ✅ LLMs for everyone ❌

I dont really get why anyone uses the cli agents rather than something like cursor Cursor itself can look like most AI tools, you can configure it to whatever you're doing or how you're working. You're gonna need the editor eventually The UX just feels better with a real UI

For @cocore.dev in the last day I - converted ALL of the backend code to idiomatic effect - added o11y and @honeycomb.io - squashed the top errors that were obvious Pre ai all that is a weeks long effort, some I’d even have to learn a lot (effect) Now it’s super high quality, no effort.

1. You must make no mistakes and have a perfect product 2. You can under no circumstances use AI to do it I dont make the rules!

I am oddly entertained by reading the llm output while it works through a problem. you can call bs quicker and you also might learn things (i know those are at odds lol)

In an AI first dev world I lean on these skills far more - Project planning and sequencing - Design and UX The only time i'm really dipping into anything dev related is mostly when i need to guide AI to an ideal solution or decisions around tooling

I like tanstack start but kinda feel there needs to be better divides between the client and server it can be easy to get into a spot where even bleeding edge ai has trouble not including a fn in the frontend

AI Harness definitely matters for UI work

@zzstoatzz.io can you update your pub search labeler to support "*" currently (according to ai): Their labeler returns labels for an exact DID but {"labels":[]} for *.

It’s funny that emdashes and eyebrows share the same shape and are both ai tells