undefined | Better HN

0 pointsb40d-48b2-979e24d ago0 comments

LLMs don't "reason".

0 comments

Why is this a meaningful distinction to you? What does "reason" mean here? Can we construct a test that cleanly splits what humans do from what LLMs do?

grey-area24d ago

Sure, things like counting the ‘r’s in strawberry, for example (till they are retrained not to make that mistake).

thot_experiment24d ago

There are humans that can't do that but are clearly capable of reasoning. Not a meaningful categorical split.

1 more reply

RedNifre23d ago

They don't see the letters, so how could they possibly succeed at that? It's like asking a human how many infrared flowers they see.

1 more reply

bensyverson24d ago

Take it up with OpenAI's API designers—it's their term

b40d-48b2-979eOP21d ago

You are the one repeating their lies.

j / k navigate · click thread line to collapse

0 comments

thot_experiment24d ago

Why is this a meaningful distinction to you? What does "reason" mean here? Can we construct a test that cleanly splits what humans do from what LLMs do?

grey-area24d ago

Sure, things like counting the ‘r’s in strawberry, for example (till they are retrained not to make that mistake).

thot_experiment24d ago

There are humans that can't do that but are clearly capable of reasoning. Not a meaningful categorical split.

1 more reply

RedNifre23d ago

They don't see the letters, so how could they possibly succeed at that? It's like asking a human how many infrared flowers they see.

1 more reply

bensyverson24d ago

Take it up with OpenAI's API designers—it's their term

b40d-48b2-979eOP21d ago

You are the one repeating their lies.

j / k navigate · click thread line to collapse