this post was submitted on 12 Aug 2026
451 points (97.5% liked)
Fuck AI
7909 readers
1282 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
Because its not ai, it's lookahead autocomplete and those are the kinds of things that would follow the previous thing in the training data
There are models that use lookahead like BERT but ChatGPT explicitly doesn't.
Your autocomplete does use a simple neural network and trains on how you write.
I'm sorry, but you don't know what artificial intelligence is. ChatGPT is a generative pretrained transformer (GPT) is a type of transformer model is a deep learning architecture is a type of machine learning is a branch of artificial intelligence (where, for context, the other branch of AI is of a fixed programming, like an opponent in Video Chess for Atari).
You don't have to like that the field is called AI; you do have to acknowledge it's been called that since 1957 (just a year after Logic Theorist) and that LLMs objectively fit squarely within said field.
what the hell is this? nested recursive definitions? contaminated contexts? circular logic? ‘it is what it is because of what it’s called just like other things that have nothing to do with this being miss labeled for profit.’
It's called a tree. AI is the root of the tree. Thought that was made extremely obvious. Let me know if you still don't get it, and I can link you to some reading material.
but you are attempting to have a conversation with people. so you should probably prioritize coherency. and claiming that llm’s are AI just like video game characters defeats your point. you are ignoring the main reason people have issue with naming such things AI. is that it’s a marketing term used to falsely imply a capability and quality llm’s fo not and can not have, for profit. and if you are going to argue that point you will need to define the word “literal”
I did. You specifically apparently just didn't understand.
They are within the same tree. You again just don't understand the structure of the field and its subfields. I'm going to repeat myself with some additional context, and if you still can't understand, that's on you:
Work your way back up the tree, and you'll see that GPTs fit squarely into AI. (In this case, an LLM can be a lot of different kinds of models, but in 2026, these are typically transformer models like GPTs and another I didn't mention called BERTs. "LLM" just means any machine learning model trained on a huge text corpus.)
your argument is still curricular, ‘it’s is what it is because it is called that’ you have complete ignored the context of the complaint and the reason.
by retreating to an overbroad definition of AI coined by deceptive marketing to attach computer chess to the science fiction concept of a thinking machine that achieves agency and conception, for money, you have forfeited the point that llm’s are glorified auto complete.
i would not be so crass as to argue that llms are just glorified auto correct as a truth, but an LLM is not capable of agency, and is not capable of true understanding, it is a clever design of tokenizing data and assigning connections of these data points. but that is not agency nor understanding, and it does generate its responses using similar tools to auto complete governed by these connections.
you took a posters flippant disdain for LLM’s and ignored his actual point in its misclassification as a means to trick people into thinking it’s akin to magic personhood. and when called out you resorted to claiming you are technically right because an active flow chart can be called an AI.
you can’t have it both ways
I'm sorry, but "I'm sorry" is not how you start a sentence, unless you want everyone to know that you are not sorry (which you should be for starting a sentence like that).
I'm sorry, but I don't care.
Okay, but it's still just a fancy autocomplete, predicting which text would come next if this was in the training data.