Hypotheses

In 1950 Alan Turing published a paper titled "Computing Machinery and Intelligence" that described a method for testing the intelligence of a machine. Called either the Turing Test or the Imitation Game, this thought experiment has become one of the foci of the Mind-Body Problem as Searle and others describe it. The required materials are two human beings and one digital computer (Turing specified that the machine be a computer). One person is to sit isolated from the other person and the computer, then ask questions of the two. If the questioner could not distinguish between the answers given by the computer and those given by the human to a statistically significant degree, the machine was said to have passed the turing test. The flurry of subversive responses is obvious - the computer would never make a mistake, the syntactic quirks of the human wouldn't be seen in the computer, the list goes on. The rules allow for the computer to play trickster to whatever degree its programmers can think to let it, so most of the previously mentioned objections are ostensibly simple to overcome. The real question, however, is whether or not passing the Turing Test actually illustrates intelligence, and if so, does this result constitute consciousness.

A strict follower of Cartesian thought would undoubtedly take the passage of the test as indication of intelligence, however, in his paper, Turing made it clear that he was uncertain of this himself. He stated in "Computing Machinery and Intelligence" that the question, "Can machines think?" is too amorphous for such strict analytical work. Rather, Turing left the question open ended, using his test as a replacement for this question as opposed to an answer to it.

More recent philosophers, on the other hand, have often cited a particular syllogism in defense of the utility of the Turing test for satisfying the concept of Artificial Intelligence. This syllogism runs as follows:

1 Humans are intelligent

2 Verbal behavior, particularly language use, illustrates that intelligence

3 If behavior that illustrates intelligence can be executed in a manner indistinguishable from the behavior of the human, the object executing this behavior is intelligent

4 Any object that passes the Turing Test possesses linguistic capabilities indistinguishable from human capabilities
______
Conclusion: Any object passing the Turing Test is intelligent

The immediate problem that comes to mind reading this arguemnt might be called the Mirror Argument. The objection focuses on the fact that the Turing Test looks only at behavioral results instead of the underlying causes of these results. To repeat the syllogism in the manner of this objection, one might start with the premise that all apples cast apple-shaped reflections in a mirror.

1 Apples cast apple-shaped reflections in a mirror

2 The perceived reflection indicates that the object is, in fact, an apple

3 If behavior that illustrates apple-ness can be executed in a manner indistinguishable from the behavior of the apple, the object executing this behavior is an apple.

4 Any object that exhibits apple reflection-casting capabilities indistinguishable from apple capabilities does so
______
Conclusion: Any object casting a reflection indistinguishable from an apple's reflection must be an apple

Since any object that casts an apple-shaped reflection indistinguishable from the real apple must be an apple, a well taken photograph of an apple is actually an apple. This is great news for the starving masses, but ontologically, I'm a little disturbed.

The problem is that while a well-written program might seem to answer the inquiries of a human being in a way that reflects intelligence, it is still merely lines of code operating on a machine composed of nothing more than electromechanical bits and pieces. The intelligence in this instance is not that of the program or the machine, but that of the programmer.

The trouble with this breakdown is that I'm beginning to sound like a dualist, which is not a trap to which I wish to fall victim.