Research paper : https://arxiv.org/pdf/2410.05229

I work for a consulting company and they’re truly going off the deep end pushing consultants to sell this miracle solution. They are now doing weekly product demos and all of them are absolutely useless hype grifts. It’s maddening.

Do we know how human brains reason? Not really… Do we have an abundance of long chains of reasoning we can use as training data?

…no.

So we don’t have the training data to get language models to talk through their reasoning then, especially not in novel or personable ways.

But also - even if we did, that wouldn’t produce ‘thought’ any more than a book about thought can produce thought.

Thinking is relational. It requires an internal self awareness. We can’t discuss that in text so much that a book is suddenly conscious.

This is the idea that"Sentience can’t come from semantics"… More is needed than that.

I still fail to see how people expect LLMs to reason. It’s like expecting a slice of pizza to reason. That’s just not what it does.

Although Porsche managed to make a car with the engine in the most idiotic place win literally everything on Earth, so I guess I’m leaving a little possibility that the slice of pizza will outreason GPT 4.