LLMs can’t reason — they just crib reasoning-like steps from their training data pivot-to-ai.com/2024/10/13/l...

LLMs can’t reason — they just crib reasoning-like steps from their training data

Researchers at Apple have come out with a new paper showing that large language models can’t reason — they’re just pattern-matching machines. [arXiv, PDF] This shouldn’t be news to anyone here. We …

Yes. Many good posts about this Apple paper lately garymarcus.substack.com/p/llms-dont-...

LLMs don’t do formal reasoning - and that is a HUGE problem

Important new study from Apple

As you say, this was always understood. But papers like this from big names will hopefully help some who still see the faces of reason and consciousness in the clouds. Curse our own pattern-matching inclinations!