X-Message-Number: 4387 Date: Thu, 11 May 1995 16:43:48 +0200 (MET DST) From: Eugen Leitl <> Subject: Re: CryoNet #4378 - #4384 On Thu, 11 May -1, CryoNet wrote: > ---------------------------------------------------------------------- > > Message #4379 > From: (Thomas Donaldson) > Subject: Re: CryoNet #4338 - #4350 > Date: Wed, 10 May 1995 10:18:44 -0700 (PDT) > > Hi! > > To Mr. Clark: > If you only want to store yourself so that you can be revived after destructive > events, you need only the simplest kind of computer: one with LOTS of memory. Why using a computer, if a bunch of DAT tape catridges will do? I did an estimate: the amount of some 1000 will be enough. Provided, if one uses optical tape a much smaller number will suffice. If storage, not emulation is demanded there is absolutely no need using semiconductor memories. They are much too bulky/volatile/power-hungry/costly for that. > Just how much memory depends, of course, on what there is about yourself that > you consider essential, and on how far down in the hierarchy of neural parts > you wish to go ie. atoms? molecules (remember lots of biochemicals are quite Single neurons/synapses (weighted-edge graph information) will do. Provided I/my friends do not notice the difference between my original deceased me and the simulation it is accurate. At least accurate enough for me. > large, with many conformations)? explicit memories? etc etc. This hypothetical > computer requires only the processing ability necessary to receive you in > coded form and store you. > > I had thought that "uploading" involved much more elaborate computing: some > variety of computer which would not only accept you but actually give you the > sense of being awake and conscious, and a similar or greater control over the > outside world than that you have now. It is a special maspar architecture all right. But it comes with no built-in magic. (Apart from its massive paralellity and special architecture, that is ;) [snip] > The question lying behind such experiments is this: we may store our memory > in the connectivity of our neurons. HOWEVER since connectivity changes even It is the synapse weight, essentially. Connectivity: much less so. Actually, this information can be translated into a special representation, to be run upon edge-of-chaos ANM (automata network machines). I am preparing a longish post on that, hang on. > with simple time, are there ADDITIONAL ways in which we store our memory > or is connectivity the only one? If connectivity is the only storage method, > then loss of connectivity ==> loss of memory. If not, not. As you can guess, > that is a crucial question when we consider the current methods for cryonic > suspension. And if connectivity is the only store, then vitrification may get > one more vote: it's apparently less likely to cause that kind of disruption. The artefacts introduced by freezing are essentially nonexistant for uploading personality scan. After raw data acquisition (voxel block) and noise suppression/ contrasting/edge detection an freeze artefact filter (e.g. a neural-net based one) run does eliminate all freeze damage. A NN filter has distinct advantages: we need not know how it works en detail. Simply several calibration (learning) runs on model tissue will suffice. By the way, sometimes in the past I said that nondestructive scan is probably impossible. I have to reduce that claim. A kind of technology based in SQUID interferometry (several arrays of integrated SQUDS hooked to broadband data acquisition recorders, e.g. RAID arrays) might produce the necessary resolution and information on dynamic behaviour both. The necessary math theory (Fourier transform, mostly) already exists (radio astronomy). Skull shielding might pose a problem here, though. > I also don't mean here to exclude external ways of inferring connectivity > after it has been broken. And perhaps (if salamanders do NOT lose their > memory after this kind of brain disruption) that is exactly what their > healing process does. Whether our brains could do the same is another Brain can tolerate severe damage only gradually losing performance. As long as roughly the same strange attractors can be regenerated from the residues everything is o.k. > question, but on the optimistic side current work on brain repair tends to > suggest that we cannot repair our brains because of chemical changes which > actively prevent regrowth, rather than because of a lack of information. Essentially, brain tissue has built-in reserves, which are increasingly activated when we grow older. Regeneration at life time introduces several difficulties (danger on emerging uncontrolled runaway growth) so evolution seemed not to have explored that path further. > (But this is only a suggestion, not a truth). > > Best and long long life, > > Thomas Donaldson > . > > ---------------------------------------------------------------------- Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=4387