The situation around AI is extremely depressing to me. It's not enough that it fucks my future, I would get some consolation if I could get nice projects out of it, but the moment I do I will get unholy amounts of hatred from the very people I help. I'm expected to keep checking function signatures by hand and tediously implement all the things that can be fully described in higher abstractions, when technology that does that very well 30x faster is right there. And it doesn't help that people primarily from the rightwing crypto-legacy counterculture are making LLM use have an even worse rep by pushing slop or outright steal people's WIPs. The shape of the LLM offer is also incredibly toxic, coming exclusively as a prepaid subscription with hard deadline temporal limits such that even if you're here for outcomes, you have this constant gacha like toxic alarm of "if you don't find a task to give to the AI, you'll be wasting money". So people are like, oh no, my weekly limits reset tomorrow, and I still have 40% of credits left, clearly I must clone some nintendo retro recomp and ask ai to implement dlss5 or some really useful shit. Just in general people are encouraged to use it much more than just focusing on the outcome would have you. But even if you have the discipline to not fall to that, accept that yes you're wasting your sub, and that's OK, and focus on the end result and get there, it doesn't matter because it's treated as an corrupting virus that nullifies any merit in building things no matter how much of the project is borne from your intelligence. And so I stand looking at a shelf of polished projects that I know publishing would get me a net negative reaction. At this point I think just hate everyone here, from the cryptobros tokenmaxxers to the "whether people are morally allowed to build new technologies depends on whether it negatively impacts me personally" crowd. I just wish I had a fatal accident already.