X-Message-Number: 12728 Date: Sat, 6 Nov 1999 09:59:42 -0500 (EST) From: Charles Platt <> Subject: Reply to John Clark On Sat, 6 Nov 1999, CryoNet wrote: > That's not quite true, a lot of progress has been made. The trouble > is that as our programming tools get smarter our ambition gets larger > too and we write more complex programs, so it never gets any easier. John, I was guilty of overstatement. I should have said that I saw few if any substantial advances in programming techniques in the last 40 years (since the invention of BASIC, really) rather than the last 50 years. Object-oriented programming was some improvement, but fundamentally we're still stuck with the same old problem: trying to define tasks in code, in such a way that every possible eventuality is taken care of, and complex actions are reduced to a simple instruction set. Your points about atoms being easier to manipulate that sheet metal and automotive components is well take; but still, when contemplating the brain, we have a horribly complex three-dimensional problem, which no robot comes remotely close to addressing in the real world today, so far as I am aware. The block-stacking demonstrations I have seen are trivial by comparison (even though they use a limited range of objects). If I saw real progress toward machines that can manipulate three-dimensional structures in a truly versatile manner, I would be encouraged. But I have not seen any such thing. Of course, if we have true AI, the AI will be able to tackle this kind of challenge more easily than we will. But initially, human beings are going to have to develop the robots, or the AI that will develop the robots. Sometimes I wonder if we can really get there from here--within a matter of decades, anyway. It may be that the prerequisite to brain repair will be AI that evolves rather than being built from the ground up. This will require considerable time, initially at least; and by the time it has matured, who knows whether it will pay any attention to us anyway? I conclude that there are so many imponderables and unforeseeable problems and variables, no grounds for optimism exist. On the other hand, the same argument could be used to conclude that no grounds for pessimism exist. We simply do not, and cannot, know what will be possible or when. --CP Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=12728