News for hackers. Stuff that matters.
Re:Are Emergent Abilities of Large Language Models a Mirage? by visarga on Monday May 01, 2023 @01:39AM (#35769428)
But language itself does, it's not the model, it's the data that has this ability. It's in the language patterns. Humans use that too - 99.99% of what reasoning we do is just replaying older ideas adapted in context. Being truly original and improving on the best ideas is a rare thing.
Re:Are Emergent Abilities of Large Language Models a Mirage? by PaulDavisThe1st on Monday May 01, 2023 @02:31PM (#35777096)
> 99.99% of what reasoning we do is just replaying older ideas adapted in context.

Asserted without evidence.

I'm willing to accept that most human speech behavior may well happen in similar ways to LLMs. But most human speech behavior is not reasoning.

Re:Are Emergent Abilities of Large Language Models a Mirage? by visarga on Tuesday May 02, 2023 @02:14AM (#35783978)