To my knowledge, GPT-3 doesn’t store information about it’s “thought” process, so if GPT-3 is able to explain it’s own puns, it would necessarily be able to explain similar puns made by people.
Yeah, but similar puns aren’t made by people—GPT-3 is that bad at puns. So I wonder how well it would explain puns that are made by people (ie, actual puns).
To my knowledge, GPT-3 doesn’t store information about it’s “thought” process, so if GPT-3 is able to explain it’s own puns, it would necessarily be able to explain similar puns made by people.
Yeah, but similar puns aren’t made by people—GPT-3 is that bad at puns. So I wonder how well it would explain puns that are made by people (ie, actual puns).
Ah I see. Misinterpreted what you were saying in that last Note.