X-Message-Number: 14721 Date: Tue, 17 Oct 2000 23:33:11 +0200 From: Henri Kluytmans <> Subject: The Mind and Identity : on how brains work -->Hi, Thomas Donaldson wrote : >First of all, our brains do not operate like any computer system yet >in existence. Nobody claims that computer systems already exist that can emulate our brain correctly. That is a misunderstanding, the standard device envisioned for uploading is not a von Neumann (step-by-step processing) computing machine (like almost all conventional computers are), but very likely a very parallel computing system. Probably an artificial (probably electronic) neural network, i.e. an artificial brain. To prevent this kind of misunderstanding one should avoid explaining uploading as "transfer of the thought processes to a computer" but phrase it as "transfer of the thought processes to an artificial brain". Indeed systems like the architecture Hugo de Garis is currently working with: neural networks emulated on top of cellular automata hardware. Maybe the conventional electronic circuits themselves will be replaced by quantum dot circuits in the future, or at least an other type of electronic circuit with nano-dimensions. ---- But at a more fundamental level, basically one could say : In principle the "not-running" (or de-animated) state of our mind is only information. ( I suppose that everybody here does agree with this ? ) It is then a logical induction that a "running" situation must be an information process. >Connections between neurons change, at different rates, >but even change in maturity. Most neurons that are connected are connected >not by one synapse but by many. This alone makes them quite unlike any >existing neural net; among other features, this changeable connectivity >makes the simple use of wires (as in computers) not at all like how >our brains work. In general terms, a lot of what our brains do looks >a bit like that done by computer neural nets, but their structure >differs a lot and therefore they will behave differently in detail. However currently research is going on to create "exact" emulations of the functioning of biological neurons. The meaning of the word "exact" in this context is : emulations capturing all aspects important for understanding the complete functioning of biological neurons. (Sorry, I do not have a reference at this moment.) In the artificial neural nets Hugo de Garis is using, the growth of new dendrites is also simulated... (See http://www.hip.atr.co.jp/~degaris/ ) >Second, the hormones going through our brain affect its activity. >Our brain also affects the output of hormones. I see hormones as a kind of broadcast system. Anyway it does not seem to be very computationally intensive. >It seems unlikely that we can simply say that we work >"like computers", even including our desires, values, aims etc. Because current physical knowledge implies - that the mind stored in a frozen brain is information - this seems to imply that a running mind is an information-process. The definition of computing system is : information-processor. Thus according to this definition our brain is a computing system. (Although it's architecture is different from any conventional PC.) >If nothing else, that claim deserves far more justification than >it has yet received. IMO, all opposite claims deserve more justification. >It's particularly important here that feelings play a large >part in virtually everything we do... even the feeling of >curiosity is a FEELING, not something automatically generated >by information alone. If the mind is an information process, feelings as part of it are information processes too. ( For example Marvin Minsky's agents model seems to fit nicely to explaining feelings. ) I dont understand why feeling cannot be an information process ??? (According to the dictionary feeling = inner perception ) >Furthermore, our brains include a great deal of unconscious parallelism. >We aren't aware of this because our CONSCIOUSNESS only remains aware >of issues produced by this parallelism (in combination with our feelings). Yes. >AT ROOT, the operation of our brain is not symbolic, among other >issues. Use of symbols is an operation of our brain, but it' >s founded on non-symbolic activity. Computing systems (even the ones based on symbolic operation) seem to be capable of doing none symbolic operations fine. Current PCs with a von Neumann architecture (i.e. step by step symbolic operation) can recognize patterns by simulating neural nets. And, there are also computing systems with their hardware architecture based directly on a neural net. ---- A footnote : If quantum-computing occurs somewhere in the brain then this will complicate matters quite a bit. However it seems very unlikely that something like this does take place. There seems to be no scientific reason to assume so. On the contrary, quantum computation seems to require physically very isolated systems to keep their coherence long enough to do any useful computations. ---- >I am happy to provide references for every point I have made here >about how our brains work. These points do not seem to imply that the brain is not an information processing system. Grtz, >Hkl Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=14721