Comment by breppp
3 days ago
I am still impressed by the fact we have found a way to feed data to matter and get human surpassing intelligence, that's a very unexpected advance
3 days ago
I am still impressed by the fact we have found a way to feed data to matter and get human surpassing intelligence, that's a very unexpected advance
It’s not human surpassing intelligence though in my opinion. If we had the same constraints on that task then we would also come up with the idea of creating our own viewer and if, like the LLM, the knowledge to do so was embedded in our brain we would do it, it would just take longer because we can’t type as fast. So instead we would buy or compile a viewer since the problem is already solved.
IDK the models definitely do write qualitatively better prose (and better code) than the majority of people. If your definition of "human intelligence" is baseline performance of 0.2% of humanity it's certainly something else.
Better code and better prose are not great indicators of intelligence. Can it help you survive in the middle of the jungle? Can it grow rice? Can it give you first aid? Intelligence is the ability to learn skills and adapt them to the situation. The importance of skills is contextual. Better prose and code is just education, not intelligence.
5 replies →
they don't, cite one example of well written LLM prose? they can definitely infer a lot more slop than people have patience to, and anyone writing or reading slop prose has realised this and replaced sitting around writing and reading slop wiith llm's writing and reading slop and then grug-brains paste 10k words of slop say "grok, pls summarize, no mistakes" and grok tells them "What a brilliantly written article! ... "
1 reply →