This we know: machine learning is not intelligence, and so the LLM chat services are not artificial intelligence. Those are just autocomplete. #AI is a misnomer.
But the other day, a trade organisation for translators published a document referring to machine #translation as "AI". That's really another step in the wrong direction for what those two letters should be applied to. Nobody thinks translating a sentence automatically is a sign of artificial intelligence.
Whatever #intelligence is, no scientist or scholar has ever suggested that it's when you look at enormous amounts of text and use it to calculate what the next word should be.
@mrundkvist I agree with you.
The tech bros are arguing that intelligence is the same as cognition. If you can associate parts into a whole then that is the path to intelligence. An LLM is just that, a neural network of tokens and their connections. That is what they mean when they say they have achieved intelligence.
Of course there is no science backing it up. It’s all just opinions.
@mrundkvist which leads to stupid shit like claiming animals have no intelligence because they don’t deal with words like an LLM.
Dude, if you’ve even owned a cat, you wouldn’t draw such conclusions. 🤣
@trenchworms
I agree with you about the environmental and conflict-related aspect of the current craze. But that doesn't really play into what the word "intelligence" means. We might imagine a situation where Chat GPT *had actually* achieved intelligence *and* was a severe environmental problem. But it's never going to. It's just all so pointless.
@CTD
“AI” was always a wankeeteering catchphrase.
At least since ~1956.
https://en.wikipedia.org/wiki/History_of_artificial_intelligence?wprov=sfla1
https://www.sense-of-rebellion.com/
Martin Rundkvist
Mikael Lundin 🍀🥦♻️
CTD
trenchworms
Dźwiedziu
Fantômas
Paul Giulan