What do think is the best argument against the Chinese Room thought experiment?
15
29
Ṁ184resolved Jun 28
1D
1W
1M
ALL
1.3%
Too complicated; no one can agree, so it's not a useful intuition pump.
54%
Modules of a system may not themselves have the characteristics they grant the larger system. Bob, who's inside the room working the levers, does not know chinese any more than a neuron in my brain "knows" chinese.
1.3%
Proves too much
0.3%
Wouldn't actually work
1.0%
Since it is only clumsily describing how a human brain works, you might as well just talk about brains, and not waste time on hypothetical rooms.
6%
For humans to be conscious and Chinese Rooms to all not be conscious, there would need to be some process in the brain that can't be represented by Chinese Room manipulation rules. By the Church–Turing thesis, this is impossible.
0.6%
The argument begs the question of the existence of consciousness in other humans; there is no good evidence for this.
7%
Proving that anything doesn't have consciousness is inherently impossible. No level of inaction or mis-action could prove such a thing; even if you grant the CR argument, it's only arguing against one pathway to consciousness.
12%
That which points to the one GLUT that talks about consciousness, out of all the vast space of possibilities, is now...the conscious person asking us to imagine this whole scenario.( https://www.lesswrong.com/posts/k6EPphHiBH4WWYFCj/gazp- )
0.7%
Insofar as anyone will accept that anyone has consciousness, ever and always, they do it empirically. The Chinese room does not address how to judge this empirically, so it is entirely beside the point.
0.7%
"I can access the internet, but I don't know anything about IPv4, so I don't 'really understand' what I'm doing. Therefore, I am not conscious" is a stupid argument; Q.E.D.
0.7%
Clearly, since no one has a good explanation or description of consciousness, the correct belief is that no one has consciousness; you might think you do, but you are wrong. Therefore, the CR argument is meaningless.
2%
The program would be very complex. We can't say whether the person following it will "learn" Chinese just by doing so.
I will choose the answer according to whichever answer destroys the Chinese Room experiment as an argument against conscious computers. It is possible I choose multiple answers if I really like them.
Jun 20, 11:52pm: the title was meant to say: what do **I** think is the best...
Get Ṁ200 play money
Related questions
🏅 Top traders
# | Name | Total profit |
---|---|---|
1 | Ṁ85 | |
2 | Ṁ36 | |
3 | Ṁ15 | |
4 | Ṁ8 | |
5 | Ṁ5 |
Sort by:
answered
For humans to be conscious and Chinese Rooms to all not be conscious, there would need to be some process in the brain that can't be represented by Chinese Room manipulation rules. By the Church–Turing thesis, this is impossible.
answered
For humans to be conscious and Chinese Rooms to all not be conscious, there would need to be some process in the brain that can't be represented by Chinese Room manipulation rules. By the Church–Turing thesis, this is impossible.
@ManifoldMarketsUser Perhaps I should rephrase this to "to prove humans aren't conscious as well rather than proving that only computers aren't"
answered
That which points to the one GLUT that talks about consciousness, out of all the vast space of possibilities, is now...the conscious person asking us to imagine this whole scenario.( https://www.lesswrong.com/posts/k6EPphHiBH4WWYFCj/gazp- )
@MartinRandall Yeah, I liked the idea of the problem itself being conscious, like it's too specific of a specification. Though I also agree with @MichaelWheatley 's explanation that it is asking the question at the wrong level
answered
That which points to the one GLUT that talks about consciousness, out of all the vast space of possibilities, is now...the conscious person asking us to imagine this whole scenario.( https://www.lesswrong.com/posts/k6EPphHiBH4WWYFCj/gazp- )
@JoyVoid the general concept of conscious thought coming from the creator of the thought experiment is under utilized in philosophy, thanks for raising it.
answered
Proving that anything doesn't have consciousness is inherently impossible. No level of inaction or mis-action could prove such a thing; even if you grant the CR argument, it's only arguing against one pathway to consciousness.
@ManifoldMarketsUser That's begging the question, tho. Consciousness is an unexplained phenomenon in the deepest sense -- we don't have any good evidence for any pathway. Maybe consciousness is a side effect of having electricity running through nerves/wires, and has nothing to do with procedural operations. Maybe the only way to get a conscious entity is through years of training up by other conscious entities. It could just as well be that it is gifted by God (or the programmer of the simulation) based on shape or origins or random roll of the die.
answered
Proving that anything doesn't have consciousness is inherently impossible. No level of inaction or mis-action could prove such a thing; even if you grant the CR argument, it's only arguing against one pathway to consciousness.
@Duncan "even if you grant the CR argument, it's only arguing against one pathway to consciousness" I thought about this too, but the CR argument is very general. Any pathway to conscious computers can be represented as CR manipulation rules. You could just have the CR person do anything the hypothetical conscious computer does.
answered
For humans to be conscious and Chinese Rooms to all not be conscious, there would need to be some process in the brain that can't be represented by Chinese Room manipulation rules. By the Church–Turing thesis, this is impossible.
answered
The argument begs the question of the existence of consciousness in other humans; there is no good evidence for this.
answered
Insofar as anyone will accept that anyone has consciousness, ever and always, they do it empirically. The Chinese room does not address how to judge this empirically, so it is entirely beside the point.
answered
The argument begs the question of the existence of consciousness in other humans; there is no good evidence for this.
answered
That which points to the one GLUT that talks about consciousness, out of all the vast space of possibilities, is now...the conscious person asking us to imagine this whole scenario.( https://www.lesswrong.com/posts/k6EPphHiBH4WWYFCj/gazp- )
answered
That which points to the one GLUT that talks about consciousness, out of all the vast space of possibilities, is now...the conscious person asking us to imagine this whole scenario.( https://www.lesswrong.com/posts/k6EPphHiBH4WWYFCj/gazp- )
So to put it in another way, the consciousness lies within the problem itself, because we need a conscious mind to imagine it. It lies within the book because such a book is as impossible to implement as water freezing itself. Even if you say such a book wouldn't be conscious, an hypothetical one is because we can't actually read in it
answered
That which points to the one GLUT that talks about consciousness, out of all the vast space of possibilities, is now...the conscious person asking us to imagine this whole scenario.( https://www.lesswrong.com/posts/k6EPphHiBH4WWYFCj/gazp- )
answered
For humans to be conscious and Chinese Rooms to all not be conscious, there would need to be some process in the brain that can't be represented by Chinese Room manipulation rules. By the Church–Turing thesis, this is impossible.
@ManifoldMarketsUser While I raise this argument when people bring up the Chinese room, I've always thought it's more likely to prove humans aren't conscious than computers aren't.
answered
The argument begs the question of the existence of consciousness in other humans; there is no good evidence for this.
answered
Proving that anything doesn't have consciousness is inherently impossible. No level of inaction or mis-action could prove such a thing; even if you grant the CR argument, it's only arguing against one pathway to consciousness.
answered
For humans to be conscious and Chinese Rooms to all not be conscious, there would need to be some process in the brain that can't be represented by Chinese Room manipulation rules. By the Church–Turing thesis, this is impossible.
If the brain uses some hypothetical uncomputable property of quantum mechanics to be conscious like in Penrose's theory (very unlikely IMHO), then a computer could just be designed that uses that the same way neurons would. We already have quantum random number generators and quantum computing.
answered
Since it is only clumsily describing how a human brain works, you might as well just talk about brains, and not waste time on hypothetical rooms.
We use auditory nerves and a language center look up the meaning of pressure waves in the air, so that we can make other pressure waves. When these waves do not go through the look-up tables (e.g., if you just look at a picture of the sound waves), they are nonsense. Talking about books and libraries and Chinese made sense 500 years ago, but it's silly to waste time on that now.
answered
Modules of a system may not themselves have the characteristics they grant the larger system. Bob, who's inside the room working the levers, does not know chinese any more than a neuron in my brain "knows" chinese.
"Knowledge" is when information is inside your brain (as opposed to inside a book on your shelf). So of course Bob, whose fiddling with dictionaries, doesn't know Chinese. But an artificial intelligence with a dictionary-and-human-powered input-output module plausibly DOES have the knowledge inside its brain. Is that any different from the parts of our brain that convert soundwaves into messages? If you made a neuralink that replaced a brain-damaged patient's language center with an antenna sending and receiving data from an ultra-fast Chinese room, would that person's subjective experience be any different from our own?
I'm not trying to persuade you of my point of view here, just to point out that we're back to square one: people arguing based on their divergent intuitions about where consciousness comes from. Once you use the analogy properly (stop equivocating between a machine and a component of the machine) it stops being able to illuminate anything.
More related questions
Related questions
Who first builds an Artificial General Intelligence?
Will the top chatbot in 2025 "think" before responding to a difficult prompt?
60% chance
Will Microsoft shut down it's A.I Lab in China?
47% chance
If the top chatbot in 2025 "thinks" before responding to a difficult prompt, will its thoughts be human-interpretable?
35% chance
When will self-improving AI outperform human-developed AI?
2030
Will AI be able to surpass a human using theory of mind when tested via Hanabi?
33% chance
Did Alan Turing believe there was a non-negligible risk that AI would exterminate or enslave humanity?
57% chance
If intelligent aliens visit us, will AI-based technologies outperform analog linguistics in translating their language?
70% chance
What's the most believable conspiracy theory?
POLL
Has Text-to-Speech Passed the Turing Test?
POLL