Hi, I work on a variety of things, most of which I talk about more on my blog than on social media. Here, you’ll probably find me talking mostly talking about Free Culture works and sometimes technology.
I don’t know if I’d recommend it to anyone else, but I like the Cook “programming language”. They’re a little AI-happy, but that does mean that they run a service that converts online recipes; no OCR that I’ve found. It works nicely for me, because I can dump the recipes in a git repository to move them around, and since it exports Markdown, I can drop recipes into my blog when the situation calls for it.
Personally, I think that using LLMs is a terrible idea, especially for Free Software. It creates a reliance on proprietary software[1]. Generated code has uncertain provenance, for the same reason that researchers can provoke LLMs to “auto-complete” famous books, undermining the license. LLMs rewrite instead of fixing, since the latter requires intelligence, so it makes less-maintainable code. And studies show that using models slows work down, rather than speeding it up.
I’ve only found two arguments in favor: The technology is (allegedly) here to stay. And sure, the studies say that it slows down work, creates bugs in most of its changes, makes everything harder to maintain, alienates the team, and erodes skill, but don’t worry, because I have a system.
I wouldn’t risk it on any project of mine, and especially not for anything released under a public license where it needs to promise people that they can use the code for any purpose.
Yeah, some claim to have open source licenses, and the OSI is desperately trying to figure out how to approve AI-specific licenses. But since there’s no (useful) way to modify the model without the training data, it’s proprietary. ↩︎
It looks like they have something at https://cook.md/, but I have to admit that the marketing-speak gives me a headache with their paid tiers and AI-bot thing.
However, because it’s all text, it’s still pretty flexible. For example, if we find a recipe, say Bánh Mì (gotta love Wikibooks having random things like this…), if you stuff
https://cook.md/in front of the URL, their (LLM-driven) service parses it out, like so, with a download file. That’s a bit fussier/uglier than handing it a link (or clicking a browser add-on), but probably still in-bounds for casual users, if only barely…