Hacker News Reader

Comment by brkebdocbdl | Hacker News Reader

thanks for exemplifying the problem.

intelligence is what allows one to understand phrases and then construct meaning from it. e.g. the paper is yellow. AI will need to have a concept of paper and yellow. and the to be verb. LLMs just mash samples and form a basic map of what can be throw in one bucket or another with no concept of anything or understanding.

basically, AI is someone capable of minimal criticism. LLMs are someone who just sit in front of the tv and have knee jerk reactions without an ounce of analytical though. qed.

View on HN

> intelligences is what allows one to understand phrases and then construct meaning from it. e.g. the paper is yellow

That doesn't clarify anything, you've ever only shuffled the confusion around, moved it to 'understand' and 'meaning'. What does it mean to understand yellow? An LLM or another person could tell you things like "Yellow? Why, that's the color of lemons" or give you a dictionary definition, but does that demonstrate 'understanding', whatever that is?

It's all a philosophical quagmire, made all the worse because for some people its a matter of faith that human minds are fundamentally different from anything soulless machines can possibly do. But these aren't important questions anyway for the same reason. Whether or not the machine 'understands' what it means for paper to be yellow, it can still perform tasks that relate to the yellowness of paper. You could ask an LLM to write a coherent poem about yellow paper and it easily can. Whether or not it 'understands' has no real relevance to practical engineering matters.

LLMs absolutely have a concept of ‘yellow’ and ‘paper’ and the verb ‘to be’. They are nothing BUT a collection of mappings around language concepts. And their connotative and denotative meanings, their cultural associations, the contexts in which they arise and the things they can and cannot do. It knows that paper’s normally white and that post-it notes are often yellow; it knows that paper can be destroyed by burning or shredding or dissolving in water; it knows paper can be marked and drawn and written on and torn and used to write letters or folded to make origami cranes.

What kind of ‘understanding’ are you looking for?

Is what you're describing simply not what people are using the term AGI to loosely describe? An LLM is an AI model is it not? No, it isn't an AGI, no, I don't think LLMs are a path to an AGI, but it's certainly ML, which is objectively a sub-field of AI.

intelligence is what allows one to understand phrases and then construct meaning from it. e.g. the paper is yellow.

That's one, out of many, definitions of "intelligence". But there's no particular reason to insist that that is the definition of intelligence in any universal, objective sense. Especially in terms of talking about "artificial intelligence" where plenty of people involved in the field will allow that the goal is not necessarily to exactly replicate human intelligence, but rather simply to achieve behavior that matches "intelligent behavior" regardless of the mechanism behind it.

> basically, AI is someone capable of minimal criticism

That's not the definition of AI or intelligence

You're letting your understanding of how LLMs work bias you. They may be at their core a token autocompleter but they have emergent intelligence

https://en.m.wikipedia.org/wiki/Emergence

Reply on news.ycombinator.com