X-Message-Number: 11773 From: Thomas Donaldson <> Subject: some answers to those who replied to me Date: Wed, 19 May 1999 00:05:32 +1000 (EST) Well! Various people have felt the need to reply to my postings. Perhaps this is flattering, perhaps not. One effect of this is that I really can't answer everything in a short time. So lets see: Mr (maybe Dr, I don't intend to denigrate him) Broderick seems to believe that we might be imitated by a sequential computer. As I have said before, we are very highly parallel, and this strongly effects the ability of nerve cells (which yes, are individually slow) to carry out the processing needed. I would like Mr Broderick to stop waving femto- seconds around and give actual calculations. And to do so, he must also consider how fast our brains can process information (it's important here that our consciousness of a change often occurs a relatively long time after our brains have already reacted). Can computers emulate any mental event? Here we get into a perhaps subtle distinction. Yes, computers can provide calculations which act --- as calculations --- just like anything you wish (if you know enough about it to put it on a computer in the first place). But that activity remains a purely symbolic event. If you put an ocean storm on a computer, and then go out sailing, your sailboat is not going to be upset by events on that computer, even if they emulate a hurricane by their calculations. Not only that, but is WE who interpret those calculations. A hurricane does not require interpretation; flight or protection, yes, but not interpretation. And to add to this problem, we do not fully understand the world which we might want to emulate means that our calculations, no matter how hard we work on them as calculations, simply cannot fully emulate the world. These thoughts tell me that virtual people in virtual worlds are no more real than Donald Duck. If, on the other hand, we actually tie a computer in with lots of peripherals, particularly those which allow it to act in the world rather than just emulate such action by calculations, then we do have a fair chance of making a conscious device. Its calculations, of course, would not be attempts to emulate the real world, but thoughts of what to do in response to various events in the real world. A hurricane is coming and so I better run into a good building to protect myself. I feel hungry so I should go find something to eat. My foot hurts, so I better look at it to see what is wrong. And because of the nature of these thoughts, I strongly suspect that the best choice of computer to make such a device would be several neural nets tied together, probably with a single sequential computer getting data from all those nets and giving some general guidance. And for the case of dreams, no, they are not virtual. The best current theory about dreams is that they are part of our brain's activity in understanding events which happened before that sleep. Just as other brain activity, they are reactions to events which took place. Of course even if they were a form of virtual reality, they'd say nothing about the possibility of virtual reality and a virtual person on a computer... our brains are NOT computers isolated from all other input. And finally, some comments about nanotechnology. Despite some apparent attempts to use that word only for one kind of nanotechnology, there are many technologies which work at that scale --- beginning with chemistry and leading on into biochemistry. And other forms of nonbiotechnical nanotechnology have been suggested by others. I have also not been impressed by the ideas about brains produced by Merkle or Drexler, or even Minsky. A lot of experimental work needs to be done, still, before we come to the kind of understanding we will need. However, I still believe that the full extent of damage to brains caused by cryopreservation --- not to mention periods of ischemia beforehand --- needs careful examination rather than just a few electron micrographs. Among other issues which need addressing is that of the distribution of various brain chemicals in cryopreserved tissues. It happens to be true that we can characterize some neurons completely by the presence of particular proteins; such information may ultimately allow us to work out just what the structure of the brain was before cryopreservation. I actually think that this kind of work would only really be done, now, by cryonicists --- though I personally would agree with a decision to put much more money and effort into much better cryopreservation of brains. We will someday need to revive those frozen by current methods, but while in storage they aren't going to change. And I would not agree that destruction of a particular brain was so great as to mean it could not ever be revived until we have studied these issues --- all the different methods by which we can infer the former state of a damaged brain --- and found that even with such knowledge that particular brain was irreparable. What does this say about nanotechnology? It says nothing at all about the variety espoused by Drexler and Merkle, but it's very clear that we WILL need means to manipulate cells at nanoscales if we want to repair damage of the kind which happens now. The other strategy, of working out how to completely avoid that damage, looks as if it will work if supported enough, and is clearly superior for those of us lucky enough to still remain alive. I do not pretend that this will satisfy all those who have popped up with questions. But it may help. And this is only my first reply, not my only reply. Best and long long life to all, Thomas Donaldson Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=11773