News for hackers. Stuff that matters.

A comment on Do large language models understand us?. Parent comment

Re:Do large language models understand us? by subroutine on Saturday December 18, 2021 @04:47AM (#29603233)
Searle goes into the room with the copy of the program instructions. He gets passed a note written in Chinese, cracks his knuckles, and then opens the manual.

He reads, "Step-1: prior to translating the note, we are going to train your neural networks on a Chinese corpus consisting of 410 billion tokens from Common Crawl, 19 billion tokens from WebText, 67 billion tokens from Books, and 3 billion tokens from Wikipedia. Training in 5, 4, 3..."

He shouts, "wait but if you do that wo shuō zhōngwén."

[ Parent | on HN ]
Re:Do large language models understand us? by ww_wpg on Saturday December 18, 2021 @10:32AM (#29605161)
The search result I'm finding for that phrase is "Wǒ bù huì shuō zhōngwén. I kept saying it over and over again, the phrase I had memorized to explain my inability to speak their language".

So I guess it means "I don't speak Mandarin"?

[ Parent | on HN ]
Re:Do large language models understand us? by elikoga on Saturday December 18, 2021 @02:08PM (#29607505)
The "Wǒ bù huì[...]" part means "I negate/not can[...]"

It means precisely the opposite.

[ Parent | on HN ]