X-Message-Number: 8135 Date: Thu, 24 Apr 97 21:14:25 From: Mike Perry <> Subject: CRYONICS Mechanical Ducks, Consciousness Bob Ettinger, #8124 (22 Apr 97) wrote: > If it looks like a duck, walks like a duck, and quacks like a duck, >it still might be something that German artisans produced a century ago, >a clockwork duck. In other words, the Turing Test is garbage. My reaction is that a clockwork duck would not pass a reasonable test of duckhood--for one thing it presumably would have no ability to respond differently to different external stimuli. It would simply go through a predetermined series of steps, no matter what, unless prevented in some way. If the Turing test is "garbage" (though I don't agree) it is because it is inadequate, not because it is wrong-headed entirely. In other words, there surely is a behavioral test of intelligence, such that passing it would reasonably qualify the taker as intelligent. And here I'll raise an issue that seems to be lurking in the background on much of the recent discussions of consciousness on this forum, though it doesn't seem to have been addressed directly (or not enough). Suppose we have an entity that behaves *as if* it is conscious. For example, maybe it is a mechanical human, only with far more sophistication than the clockwork ducks. In particular it has an onboard device, made with advanced technology, that serves as its "brain." It reacts very convincingly like a human--the main difference being that *it is not made of meat*, but something very different. Does it have feeling? This question is a little more complicated than it may seem. Let's consider two possibilities. In case 1, the "brain"--call it George-- emulates a human brain, neuron-by-neuron, and maybe down to much lower levels. You can establish a close correspondence between processes going on in a human brain, and what is going on in George. You could, if you wished, make a human brain that would function analogously to George, making allowance only for the fact that both are probabilistic devices (we'll assume) so that no two "runs" are likely to be exactly the same. Again, though, George is not made of meat. Maybe he is a fancy optical computer made of glass fiber and such, or a quantum computer, or something else we haven't thought of at all. But whatever he is, he is very different from what we are made of, yet he emulates a human brain at every level that is significant, as far as we can tell. Here it seems we either have to grant that George has feeling, or accept the vitalistic position that there is some unknowable "essence" we must attach to a meat machine, that cannot be captured at the information-alone level. This latter I simply would refuse to do, and instead give George the benefit of the doubt. In other words, I see no way to avoid the conclusion that feeling is reducible to information processing. For the second scenario, I imagine that again you have George, who directs behavior that seems entirely human, yet there is *no* close resemblance to what goes on in a real human brain. So again, does George have feeling? Here I would say that, if George would fail the test based on processing details, it's probably a sign that the definition of "system with feeling" is faulty and in need of generalization. So again, I'd probably give George the benefit of the doubt, though here I can see possible complications. "George" for example, could be a composite like a nation, made of agents who individually have feelings, thoughts, etc. bearing little resemblance to George's. But overall, the way a system behaves seems adequate to determine whether *that system*--considered as an agent in its own right--has feeling, and what its feelings are. An apparent contradiction of this principle would be found with a person who is perfectly motionless. No behavior would seem to necessitate no feeling, yet clearly such a person could have feeling. But I see two ways of resolving this paradox. One is, we could just cut the gordian knot and say that the motionless "person" is an entity with no feeling, and the active brain inside is not this statue, but another agent entirely. Or alternatively, we could observe what is going on inside and include that in our total assessment of behavior. In any case, though, what we regard as subjective states should be deducible from what we can, in principle, observe. Mike Perry Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=8135