[FRIAM] GPT-3 and the chinese room

[uǝlƃ ↙↙↙]

I doubt it. I remember it being in the context of someone asking for a *simple* explanation of Noether's theorem. There was no context of your boss or a philosopher or anything like that. I don't even remember you being there. The Chinese Room wasn't really even relevant to that conversation. So, whoever said "Chinese Room" then, missed the point entirely.

Of course, one *could* make a relevant point about *compression*. Sometimes an elegant re-stating of someone else's thoughts/expression might seem simple or simpler. But I'd argue elegant/compressed expressions are no simpler than their rambling brethren. They often rely on hierarchy, where in order to get a compact expression, you have to rely on jargon that actually hinders comprehension by outsiders. So, by some measure, the more rambling version would be the simpler one.

And that recalls this article I read the other day:
https://aeon.co/essays/the-intellectual-character-of-conspiracy-theorists

which dovetails nicely with the one SteveS posted. The idea is that, rather than focus on reasons (justification) for thoughts/behaviors, we should infer the *character* of a person (or machine) and work with that instead. On the one hand, I agree. But similar to my rhetoric about the longer explanation being the simpler one, I often think that if our *inference* method is biased, then whatever character we infer will be biased ... perhaps even amplifying that bias. I'm always a fan of sticking as close to the data as possible. Never trust your inferences. So, I'm pretty sure I disagree completely with Cassam.  I have the same intuition with foreign language translations of books. I think I may *prefer* Google translate's mangled output to the nuanced, heavily modeled, re-imagined output of humans who speak the language of the original text.

This video comes to mind:

The Alternative Facts Gospel
https://youtu.be/78bsM7RbK0A

On 7/21/20 1:29 PM, Frank Wimberly wrote:
> Re:  Chinese Room
> 
> I mentioned the Chinese Room thought experiment to my erstwhile boss, a bona fide philosopher.  His reaction, "Anything follows from a false premise.". I think he meant that having a room full of Chinese scholars who laboriously execute a complex algorithm they don't understand is preposterous.  Maybe something like that reasoning caused you to react disdainfully when you did.


-- 
↙↙↙ uǝlƃ