[laminatedsmore]

I had a chat with GPT about this and it came up with the term 'data grounded cognition' to describe an 'intelligence' that is derived purely from (and expressed through) statistical patterns in data.

I quite like the term, and it seems quite unique (perhaps cribbing from 'grounded cognition' though that's an entirely different idea AFAIK)

[JohnFen]

"Cognition" means understanding and knowing. As problematic as "intelligence" is when describing these systems, I think "cognition" is even worse. "Intelligence" is vague and "cognition" is specific, but "cognition" is also incorrect.

[laminatedsmore]

The googled definition says "the mental action or process of acquiring knowledge and understanding through thought, experience, and the senses.", which I think fits ok.

Do you have a word that you prefer?

[JohnFen]

Right. That's the definition I think doesn't fit. GPTs do not "understand", and certainly do not "experience" or have senses.

> Do you have a word that you prefer?

Nope. That's why I use "intelligence" despite the problems with it. "Intelligence" may be a blank slate on which you can write whatever meaning you wish, but at least it can stretch to mean something accurate.

[laminatedsmore]

I'd say that knowledge (stored as weights) is acquired through experience (starting from a mostly blank start, weights are formed through exposure to training data). For me that's a useful mental model which helps me think about what a LLM is and isn't.

'Understanding' has an ambiguous meaning, while thought and senses are certainly not applicable.

But it's a semantic discussion about a novel and not well understood topic so :shrug: :)

[JohnFen]

> But it's a semantic discussion about a novel and not well understood topic

I agree entirely.