I made a nice little CLI tool for testing the subjunctive in Italian. Claude code spun off a bunch of Claude API calls to generate example sentences with fill-in-the-blank spots for the correctly conjugated verb. Having an AI generate a prompt for and call out to other AIs was a bit surreal!
Nice article! The generated images make me so nostalgic for the early days of AI image generation. DeepDream and others had such uncanny, interesting generations.
> Several commenters suggested the original essay was written by an LLM. They were half right. Both that essay and this one were written with Claude as a drafting partner. I directed the argument; the LLM helped with prose. I mention this not as confession but as demonstration: the human brought the utility function, the machine brought the compute. If that division of labour bothers you, I’d suggest the discomfort says more about the Bitter Lesson than about my writing process.
Very cool idea. Interested to see how this progresses.
One question: how worried are you about over-training on this particular dataset? i.e. instead of generalizing you lean more toward memorization? Obviously you leave out a validation set but since you're meta-optimizing the model itself by its performance on the validation dataset you're still at risk of over-fitting.
I got a cheap Chinese one (no camera, wifi) in 2024 and it's been a game changer. Yeah it's kind of dumb but it runs every day and picks up an unholy amount of dust, cat hair, and the like. Maybe if you were already vacuuming every day they're pretty useless but for me it's been night and day. As another commenter said, they're also surprisingly repairable, and I bought a ton of spare parts before the tariffs went in.
Really nice article! I have some OpenGL familiarity and tried out Vulkan but bounced off of it due to all of the up-front complexity just getting something running. Might give it another shot now!
https://youtu.be/Is8N7B9b0GQ