OPPAI ORACLE

I’d like to know what’s the general opinion here on AI. For me it was inescapable. Right when I got into a high school that had IT courses integrated with normal high school stuff (common federally funded high schools in Brazil), the big AI boom started to pick up pace in 2020. I was pretty skeptical of it, and I didn’t even use it to cheat on tests that much; I’d rather copy from a friend than from an AI, my friend has proven good grades, after all.
Eventually though I started to use it more and more to solve issues or help me out with things on my Linux system. First I would just send it error messages, then I was asking AI to write me whole scripts to automate stuff on my machine. It became a complete replacement of scavenging old forum posts or reading manual pages over and over, or even YouTube videos, which I already considered laziness on my part, watching a video solving an issue instead of finding the answer by myself.
Now I was using a machine to do things for me that would take me double the time and double the effort. Honestly, it made me so lazy that I don’t think I can code without it. But it made ricing and tinkering with my system actually fun. Should I really spend the whole morning on a shell script? Why not ask the AI? Uni homework I need to code? AI! Hobby project of something I know I could do but would take too long? Well, I’ll read what the AI wrote, so it’s totally safe! It’s almost a vice now. I know it’s bad for me (my brain development and the planet I live in, which is a part of me), but I can’t really hit the brakes and go back to simple debugging and error correction. It’s too addictive to be “productive” with AI.

one day i asked GPT to help me make a simple virtual MIDI device and it absolutely failed the task


I can't imagine how annoying IT schools/jobs must me nowadays with the LLM craze

I refuse to surrender my agency to a facile reductive autocomplete on steroids which lacks qualia and is thus incapable of making real architectural decisions or value judgements.

I don't know, making programs with AI is like making art by hiring someone to do it for you -- probably works, but I'm not interested in that. Telling others what to do for you is for managers. I like programming like an artist likes painting, I'll be programming by hand even if there are machines who do it infinitely better and faster, I just enjoy it.

i'm afraid people are gonna use LLMs as an excuse to make their software even more complicated and hard to read, why make your code comprehensible when the clanker is gonna handle it just fine anyway ?


I've seen claims as retarded as "AI is the open source dream come true" because it would somewhat give noobs the power to change software while completely ignoring the fact that what they're referring to as AI is the most proprietary software imaginable accessible only through a constantly monitored paid API

I don't really enjoy programming tbh, and I've only ever written big projects in Python in the past (unless you count Game Maker Language too lol) so any other languages I would need to put in a lot of time to learn. Claude is better at programming than its ever been and will only get better so I don't see why not to use it unless programming is your hobby which I guess is for a lot of people here.

netdoll wrote: Thu Jun 04, 2026 10:36 pm

I refuse to surrender my agency to a facile reductive autocomplete on steroids which lacks qualia and is thus incapable of making real architectural decisions or value judgements.

I'd be interested in your argument on how experiencing qualia somehow helps with coding... shit I don't even experience much reading words on a screen unless it's a story I'm reading.