Re:I played chess against ChatGPT-4 and lost
by YeGoblynQueenne on Sunday March 19, 2023 @06:36PM
(#35224431)
>> I can ask an LLM what 2+2 is and it can answer with 4. That's a discrete
result. So how is this different from human thinking? Where is your evidence
that this is not a similar mechanism?
A language model can match "2+2" with "4" because it's approximating the
distribution of token collocations in a large text corpus, not because it's
approximating integer addition.
We know this because we know that language models are trained on token
collocations (word embeddings) and not arithmetic functions. We know how
language models are trained because we know how they're made, because they're
made by humans and they're made following principles that are widely shared in
academic textbooks and scholarly articles all over the place.
If you wanted to train a neural net to approximate integer addition,
like, really the entire of internet addition, you'd have to train the net on
every matching pair of "x+y" and "z" where x+y=z. We know this because we
understand how neural nets work, and how internet arithmetic works, again
because we've made both, and we know how both work and what they can do.
For the record, people have trained neural nets to count or do arithmetic,
and they can only learn it within some small limits, with respect to their
training set. I got those refs somewhere, ask and you shall be given.
>> Humans infer systems from language all the time.
Humans are not neural nets, and neural nets are not humans. Does that suffice?
I don't know if I can do any better than that. Humans do human things, neural
nets do neural net things, and humans can do things that neural nets can't
even get close to. Like, dunno, inventing arithmetic? Or axiomatizing it? Or
proving that its axiomatization is incomplete. That sort of stuff. Things for
which there are no training examples, not of their instances, but of their
entire concept class.
>> But you hold "intelligence" in too high a regard
Where does that stuff come from, I wonder? Of course I hold intelligence in
high regard. What do you hold in high regard, stupidity?
A language model can match "2+2" with "4" because it's approximating the distribution of token collocations in a large text corpus, not because it's approximating integer addition.
We know this because we know that language models are trained on token collocations (word embeddings) and not arithmetic functions. We know how language models are trained because we know how they're made, because they're made by humans and they're made following principles that are widely shared in academic textbooks and scholarly articles all over the place.
If you wanted to train a neural net to approximate integer addition, like, really the entire of internet addition, you'd have to train the net on every matching pair of "x+y" and "z" where x+y=z. We know this because we understand how neural nets work, and how internet arithmetic works, again because we've made both, and we know how both work and what they can do.
For the record, people have trained neural nets to count or do arithmetic, and they can only learn it within some small limits, with respect to their training set. I got those refs somewhere, ask and you shall be given.
>> Humans infer systems from language all the time.
Humans are not neural nets, and neural nets are not humans. Does that suffice? I don't know if I can do any better than that. Humans do human things, neural nets do neural net things, and humans can do things that neural nets can't even get close to. Like, dunno, inventing arithmetic? Or axiomatizing it? Or proving that its axiomatization is incomplete. That sort of stuff. Things for which there are no training examples, not of their instances, but of their entire concept class.
>> But you hold "intelligence" in too high a regard
Where does that stuff come from, I wonder? Of course I hold intelligence in high regard. What do you hold in high regard, stupidity?