X-Message-Number: 14904 Date: Mon, 13 Nov 2000 07:35:26 -0500 From: Thomas Donaldson <> Subject: computers and brains Hi everyone! Yes, what I said about whether or not brains could be imitated by computers (they would have to have peripherals and work in real time, in the first place) deserves more discussion. One major way our brains differ from computers is that at least in some brain areas, we not only get generation of new neurons in adult human beings, but also constant change in their connections. Given that we have millions of neurons, neither of these features would be easy to imitate by (say) a system which had all possible connections between our neurons. These features also lie behind my question as to whether or not we really constitute Turing machines. Moreover it's at least not obvious that OUR real-time features ie. ability to respond to the world with our various senses ... can simply be imitated by a fast computer chip. The problem is that in many cases we have a single neuron doing relatively small jobs; to replace many neurons by a single computer chip would require the chip to spend time on each job, and this creates a different situation than what we'd get if all jobs were worked on simultaneously. I would actually support the use of neural net computers to get rough ideas about how we work. However we should not confuse the computers with the neurons (or even collections of neurons) in our brains. They are partial models, only... valuable as such, but hardly a full explanation of how our brains work. I would also point out that I was discussing "computers as we know them today". If we can someday build devices which fully imitate neurons, then we would have at least raised the possibility of "computer" versions of human beings. If they (and we!) turn out not to match Turing machines, that doesn't mean such machines are impossible. It just tells us that Turing machines don't take in the full ambit of possible devices ... hardly a deep thought. Is your toaster a Turing machine? Your oven, your washing machine? I believe the features I discuss in this message give those which suggest that current computers cannot imitate human brains. I will also point out that these features cannot be imitated merely by making more "powerful" computers. They are a different direction of development, not just a difference in "power". Best and long long life to all, Thomas Donaldson Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=14904