consensus: LLMs hallucinate a lot, can't rely on them for anything, much less important things like legal questions
reality: 5 very expensive tax attorneys give 5 diff pieces of advice and ChatGPT gives the actually useful suggestion zero shot
just showed my niece (2) and nephew (7) a DALL-E 3 portrait and @suno_ai_ AI-generated songs & they were ~mildly impressed
kids will grow up with AI as a normal, non-mindblowing feature of computers—they won’t know anything about this awkward period of unintelligent computers
@cloned_ID@ryan_harlow@alieninsect Have you considered that neural substrate eliminates noise based on conditioning and makes all predictions based on that modeling? And then considered what it might be like if it had received (or does receive) different conditioning?
@Beavkin@bitcloud BTW, very few people doing meaningless labor love the having a job part. They love not starving to death and having a roof over their heads.
@Beavkin@bitcloud This kind of thinking keeps things stuck. People had meaningful work & lived robust lives well before the sweatshop came to town and destroyed the local economy and quality of life. Transitioning out of this economy does require support, but it doesn't mean it shouldn't be done.
My cat sees everything through the eyes of a cat. He understands only a fraction of what I know. Why is it difficult to imagine that there are beings who are proportionally more expensive in their awareness?
It's like a cosmic version of the Dunning–Kruger effect.
The zone of proximal development is from Vygotsky's play theory. It posits that the child will behave "a head taller than himself" when given a role to play (like asking a child to pretend to be a parent).
The AI also benefits from the ZPD:
@quantum_oasis Who? Beff? edgelord anons? Most funded projects seem deeply concerned about safety, to the point of neutering the models so they can't do things like kill lines of code.
@quantum_oasis I don’t see a way around any of these issues that doesn’t involve creating machines that are smarter and more capable than we are, and there is no reason to believe something genuinely smarter than us won’t have benevolent values and a deeper understanding of agency and free will