X-Message-Number: 7896 Date: Wed, 19 Mar 97 12:29:04 From: Mike Perry <> Subject: digital vs. analog, movies Here are more comments relating to uploading. The worry is expressed that we are straying from cryonics in such discussions, but I think they are relevant to our general theme of how to perpetuate our existence, which is what cryonics is about too. Here I'll address two issues that have been raised by Thomas Donaldson and others. One is the idea that "the world is not digital," which includes persons. A person simulated in a computer would therefore fail to really be that person in any reasonable sense; the range of experiences, etc. possible in this domain would necessarily be more limited than in the real world. However, as I've said before, based on physics, the world is digital after all. Based on the Beckenstein bound for quantum systems with finite spatial extent and energy content, Tipler arrives at around 10^45 bits for the size of the "message" that completely characterizes a human being at the quantum level. If we raise 2 to the power of this number, we come up with the maximum number of states a human being could ever be in (barring the future possibility of brain augmentation). It is finite, and the idea that we are analog devices capable of a continuum of different states is an illusion caused by the overwhelming number of different states involved. We really are digital--along with everything else in our world. (In fact it is highly unlikely that all the quantum states are significant for the human being; Tipler estimates the brain's real information capacity at around 10^15 bits, which is echoed more or less by others. This is still a big number but much closer to the capacity of our existing computers.) The other issue, which Thomas raises, is that a computer simulation of a person would be like a movie--no real consciousness involved. And I can see how this would come up. If the computer's actions were based simply on reading a disk file containing precomputed information, for instance, its actions would be very like a movie--in fact it would be a kind of movie. But the computer could also interact with the outside world, and the characters inside could communicate with you or me, who would supply unpredictable input which they would have to decide how to respond to. Under those conditions I see no reasons why their "consciousness"--if they were convincing about it--would have to be considered inauthentic or "not real." But let's look closer at that "movie." If I could "run" a simulated person entirely from precomputed data, could that simulation experience feeling? Well, in dreams people experience real feeling (though they are not fully conscious). The dream sequence must be generated by a part of the brain over which the subject has no or only limited control; events seem to be happening as they do in real life. Presumably there is some brain mechanism that supplies these "events." Such a device could just as well be a piece of fancy hardware we have made. I think too that the impulses that create the illusion of real events, though probably generated at random on the fly (partly at least) could just as well be read from a precomputed file. The person experiencing the dream would not know the difference (if not told beforehand). So in this case something very like a movie, with predictable, precomputed events every step of the way, would also be a process involving some consciousness and real feeling. Mike Perry http://www.alcor.org Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=7896