this post was submitted on 17 Jan 2026
889 points (99.4% liked)
Fuck AI
5262 readers
2074 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
I, too, am a developer of closer to 15 years than I'd like to admit to myself, though mostly embedded and/or back-end. And while I have no problem with AI in its broad sense (obviously machine learning/spicy statistics, computer vision, and natural language processing and whatnot have potential to be enormously useful), I am generally hostile to generative AI. I thinking using copyrighted material as training data without the copyright holders' permission should be banned. And while I would have no objection to ethically-trained models in a hypothetical future where we have abundant clean energy to run the data centres and also all the new desalination plants we would need, that also remains a problem, and so I have resisted using such tools at work, too.
I agree with everything you've said after this point.
This is the crux of my problem. I find it to be overly judgemental. If you're self-employed and you need a website for your business or whatever, then you could pay someone to do it for you, but then you only have so much money in the budget. You could also learn how to code and/or graphic design and do it yourself, but then you only have so many hours in the day. If vibe coding produced something viable for you in the quickest, cheapest way, then that is obviously the rational and sensible thing to do. You might even spend the time learning something else instead that is more relevant to your interests.
Using generative AI to do something doesn't (necessarily) mean that you don't value the knowledge or skills required to do it the hard way, it only means that you value it less than something else that you might otherwise be doing with your time, and I don't think that is a moral failing.
As an example, I occasionally like to a bit of shitposting. Were it not for all those other things that I don't like about generative AI, I would probably be generating AI slop memes with the best of the them. As it is I mostly just stick to text-based comments with bad puns and references to song lyrics no-one will remember. I could put in the hours to learn how to use GIMP so I could do it without AI, but quite frankly I have books on the go, I've got a couple of musical instruments to learn/practise, and I spent all day at my software job, where I think critically (or so I claim), so I'd rather being doing those things instead. I don't think I have neglected my cognitive growth; I've just chosen to focus it on something different to what you might have.
Ohhh. I think we’re both defending different hills! I’m not against the use of generative AI for purposeful creation. What I’m against is the delegation of critical thinking.
It’s the difference between:
(That last one is admittedly my own guilty contribution to the slop soup and favourite desktop background of at least a whole year)
Versus:
The difference is oversight and vision. The first two are asking AI to execute well-defined tasks with explicit parameters and rules, the first example in particular offers the LLM an out if it finds itself at an impasse.
The latter examples are asking a prediction engine to predict a vague concept. Don’t expect originality/innovation from something that was forcibly constrained to pick from a soup made of prior art then locked down, because that’s what gradient descent essentially does to the neural networks during training: reduce the error margin by restricting the possible solutions for any given problem to only what is possible within the training set, which is also known as plagiarism.
Edit: a slight elaboration on the last part:
Neural networks trained with gradient descent will do the absolute minimum to reach a solution. That’s the nature of the training process.
What this essentially means is that effort scales with prompt complexity! A simple/basic prompt begets you the most generic result possible. Because it allows the network to slide along the shortest path from the input token to a very predictable result.