4 ms·
> it has nothing to say about how machines can understand language or how people can communicate better. There's an entire subfield of computational linguistic
by michaelmior 4y ago
> it has nothing to say about how machines can understand language or how people can communicate better.
There's an entire subfield of computational linguistics that would disagree with you.
> pretend that you understood it
This seems to suggest that no one who claims to have understanding does, which seems like a large claim.
> one of those is a useful skill and the other one isn’t.
I strongly disagree that understanding language is any more useless than understanding quantum mechanics, but based on your stance as I understand it, I can't imagine either of us will have any success in convincing the other.
- PaulHoule 4y agoLanguage in the abstract is one thing, James Joyce is another. So far as computational linguistics is concerned there is a long term fight which is something like large language models 99 - Noam Chomsky 1. That is, the productive half of computational linguistics solidly rejects the dominant paradigm of the rest of the field.