this post was submitted on 25 Feb 2026
652 points (99.4% liked)
Fuck AI
6061 readers
1986 users here now
"We did it, Patrick! We made a technological breakthrough!"
A place for all those who loathe AI to discuss things, post articles, and ridicule the AI hype. Proud supporter of working people. And proud booer of SXSW 2024.
AI, in this case, refers to LLMs, GPT technology, and anything listed as "AI" meant to increase market valuations.
founded 2 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
The integration of AI into regular, deterministic software bothers me a lot. As a developer, I want my IDE to be consistent so I can feel comfortable that autocompletion and other functions do what I expect. But recently, and at an accelerating rate, I find the IDE to make inconsistent autocompletions - often suggesting a lot more code, which often contains errors and non-existant references.
I could use AI every now and then for work, but I wish it only helped when I'm explicit about it. Currently, it kills workflow and sucks out energy both from me and the grid.
Yes, this I feel is the crux of the situation. The industry is trying to hammer a probabilistic solution into deterministic holes. I mean, it’s no wonder people keep fucking up their shit with openclaw. If have an agent, it should do exactly what I tell it.
Agreed. It's a damn shame, because there are a lot of interesting ideas at play, lost in the miasma of geopolitics and acceleration. If only we could slow down, build/scale probabilistic hardware, and carefully test and validate which problems are best suited to deterministic or probabilistic computing.
But nah, let's be rapacious, elevate "AI" into an awe-inspiring entity, and thrust it into all the orifices.
I’m reminded of flight school. To finally fly solo, it’s not enough for a student pilot to fly well and know the rules. My instructors make it clear - it’s the consistency that’s important. Getting into a rhythm and consistently executing maneuvers in the same way every time is key.
It’s only a matter of time until someone attempts to shove AI into airplanes. If it can’t replicate the same results every time, that’s a massive problem.
I'm really surprised there hasn't been a tech bro screeching that his AI can fly a 747 without human intervention.
Talking to AI is sort of like talking to a neurotypical person, in that they make up half of what they think I said.
If you talk to AI and feel like you're going insane...that's what talking to you without understanding your facial expressions is like. Half of what people say has nothing to do with what I said, and you expect me to just fill in the blanks with your eyebrow movements or something.
Edit: Non-autistic people communicate entirely through inference and body gestures. When I say one thing, non-autistic people think I'm secretly saying some other thing that they imagined. That's sort of how AI is, it will take a small comment and blow it our of proportion or infer meaning that isn't there. That's what talking to most of you is like.