I am a thinking thing. What makes me a thinking thing? What is constitutive of (sufficient for) my being a thinker? The answer of Strong AI is that a thing is a thinking thing in virtue of its instantiation or implementation of the right sort of computer program. If we think of the brain as the hardware, then the mind is the software running on the hardware. The idea is that my thinking in the broadest sense of the term is just my brain's implementation of a very complex program. Suppose I am watching some birds and hearing them chirp. The visual and auditory data is delivered to my brain (my CPU) via my optical and auditory transducers, and this data is processed in accordance with the set of instructions which is the program that my brain is running. This results in some behavioral output. The chirping of the birds may elicit some such piece of linguistic behavior as the utterance 'Damn those noisy birds!'
Accordingly, my thinking brain is just a syntactic engine: it is a machine that is running a program. When I play my chess computer, I am doing what it is doing, namely, running a program. It is just that it runs a better program on a faster machine and so wipes the floor with me.
Now I agree with John Searle that this cannot be right. My thinking cannot be my brain's implementation of a program. This is because thinking involves a semantic dimension: my thoughts mean or intend something. They possess intrinsic intentionality. But no physical object or state possesses intrinsic intentionality. At most, physical objects or states possess derivative intentionality, intentionality that derives from a being who possesses intrinsic intentionality.
The Chinese Room (a version of which is here) is a thought-experiment devised by John Searle to make this more intuitive. A man with no understanding of the Chinese language is enclosed in a windowless room. He has a manual that allows him to match up strings of Chinese characters. Through slit A comes a piece of paper with squiggle-squiggle on it. Looking in his manual, the man sees (with his eyes, but with no understanding of meaning) that squiggle-squiggle correlates with squoggle-squoggle. So he passes out through slit B a piece of paper with squoggle-squoggle on it. Unbeknownst to the man, it is questions formulated in Chinese that are coming in through slit A and it is answers to those questions that he is passing out through slit B.
So what we have here is the implementation (instantiation, realization, running) of a certain program or set of instructions. Now the question is this: Does the man in the room, or the whole room including the man and his manual, etc., understand anything in virtue of instantiating the program?
The man may be thinking about sex while he does his boring job, but he is not thinking ABOUT anything just in virtue of his implementing the program. He has no understanding of the characters he takes in or passes out. And yet, viewed fromn outside, the room behaves as if it understands Chinese. But that's all it is: AS IF behavior. What we have is mere simulation of understanding. There is no artificial intelligence here; there is an artificial SIMULATION of intelligence.
It's a smart room, but it is smart the way my thermostat is smart: smart enough to 'know' when to turn on the air conditioner. What my thermostat does is 'sense' that the temperature is 82 degrees and 'remember' that I told it to turn on the AC when the temp. gets up to 82. Since it 'desires' not to end up on the junk heap, it does what it is told.
Now obviously my thermostat, 'smart' as it is, does not know, sense, remember or desire diddlyjack. The same goes for the Chinese room and any contraption you rig up no matter how complicated. Thinking is not the instantiating of programs.
Now let's consider part of what Ray Kurzweil has to say:
Searle is best known for his “Chinese Room” analogy and has presented various formulations of it over twenty years (see below). His descriptions illustrate a failure to understand the essence of either brain processes or the nonbiological processes that could replicate them. Searle starts with the assumption that the “man” in the room doesn’t understand anything because, after all, “he is just a computer,” thereby illuminating Searle’s own bias. Searle then concludes—no surprise—that the computer doesn’t understand. Searle combines this tautology with a basic contradiction: The computer doesn’t understand Chinese, yet (according to Searle) can convincingly answer questions in Chinese. But if an entity—biological or otherwise—really doesn’t understand human language, it will quickly be unmasked by a competent interlocutor. In addition, for the program to convincingly respond, it would have to be as complex as a human brain. The observers would long be dead while the man in the room spends millions of years following a program billions of pages long.
I would say that Kurzweil just doesn't get it. The point is that programs are entirely syntactical, while mental states have content, or as Searle likes to say 'semantics.' Since syntax is not sufficient for sematics, implementing a program is not sufficient for understanding.
Most importantly, the man is acting only as the central processing unit, a small part of a system. While the man may not see it, the understanding is distributed across the entire pattern of the program itself and the billions of notes he would have to make to follow the program. I understand English, but none of my neurons do. My understanding is represented in vast patterns of neurotransmitter strengths, synaptic clefts, and interneuronal connections. Searle appears not to understand the significance of distributed patterns of information and their emergent properties.
It doesn't matter whether we take the man or the whole system. There is no understanding of Chinese in the man even though he plays the role of the CPU in implementing the program, and there is no understanding of Chinese in the room as a whole. To speak of understanding as distributed shows a failure to grasp the importance of the unity of consciousness, a topic I will return to in a later post.
