The Chinese Room Argument | Hacker News Reader

The Chinese Room Argument

plato.stanford.edu

8 pointsjonbaer5 comments

> "by following the program for manipulating symbols and numerals just as a computer does, he sends appropriate strings of Chinese characters back out under the door, and this leads those outside to mistakenly suppose there is a Chinese speaker in the room."

Clearly there is actually a Chinese speaker in the room. It's the system (computer) implemented by the program. I'm with the "system" critics here. Searle's argument is Reductio ad absurdum which doesn't work on readers who don't find it absurd that a sufficiently complicated system could understand Chinese.

Counter argument to system argument goes like this: If man memorizes all the rules and procedures he will be the system. Still doesn't know Chinese.

He is running a system that knows Chinese. The system just isn't him.

I think the problem is 'what is intelligence' it's all wrapped up with consciousness and our idea of it is based on how we think.

A Tesla self driving car can, in the right conditions be much safer than a human, but then it does stupid things where a human wouldn't. Is it more intelligent than a human? Less? Or just different?

Just following steps like a computer generally does, or the operator is in this example is not what i would define as intelligence. It's more akin to cargo culting than actual understanding. But then if you go the AI route, the computers make decisions we can't reason about. I think that has its own issues because it stops being a tool (something we ultimately control) to being something else.

I would say AI is intelligence though. Although I'm uneasy with that assessment, because I'm not sure if I'm reading to much in to a process I don't understand, that occasionally gets the answer wrong, or maybe fallibility is a hallmark of intelligence. A tool just does what it's told, it doesn't have the opportunity to make a mistake.

I suppose another useful distinction is the ability to learn. That computer program couldn't learn a new language based on knowing Chinese. The human operator could. A dictionary has this limitation, and you wouldn't describe that ask 'knowing' Chinese.

Right, so passing the Turing test is not sufficient to prove conscious. How can you prove that you yourself are conscious? What if we are all simply complex parrots?

Reply on news.ycombinator.com