Meeting notice: The 03.Sept 02 meeting will be held at 7:30 P.M. at the Royal East (782 Main St., Cambridge), a block down from the corner of Main St. and Mass Ave. If you're new and can't recognize us, ask the manager. He'll probably know where we are. More details below. <-><-><-><-><-><-><-><-><-><-><-><-><-><-><-><-><-><-><-><-><-> Suggested topic: testing the claims made for advanced simulators For some years Jeffrey Harrow has been writing a valuable nanotech- friendly (and free) tech letter called The Harrow Report. The current issue retrieves remarks by R. Stanley Williams, HP Fellow and director of Hewlett-Packard Co.'s Quantum Science Research Labs in Palo Alto from the May 1 issue of Industryweek.com. (Subscribe at http://www.TheHarrowGroup.com. Server seems to be down at the moment, but I assume that is temporary.) Williams argues that it should be possible to increase the power efficiency of computing by a factor of a billion, essentially extending Moore's Law for a second fifty years. "That would put the power of all of today's present computers in the palm of your hand," Williams says, concluding, "The age of computing really hasn't begun yet." He says that HP's strategy is to ride this opportunity by reinventing the integrated circuit with molecular components. Harrow also quotes Verner Vinge on the meaning of this extension: The best analogy that I see is with the evolutionary past: Animals can adapt to problems and make inventions, but often no faster than natural selection can do its work -- the world acts as its own simulator in the case of natural selection. We humans have the ability to internalize the world and conduct "what if's" in our heads; we can solve many problems thousands of times faster than natural selection. Now, by creating the means to execute those simulations at much higher speeds, we are entering a regime as radically different from our human past as we humans are from the lower animals. What Vinge might be saying here is that a sufficiently large amount of brute computational force will get us to the singularity even if we continue to find it hard to write an artificial Feynman. While smartness is always nice to have, you can also get pretty good results by speeding up, say, a stupid chess program, a million times over. If brute force is enough, getting to the Singularity can practically be guaranteed. But is it? the computational burden on a simulation goes up by a number that is the sum of the exponent of the number of independent elements plus the number of cycles. Will "only" nine or ten orders of magnitude be enough to allow a machine to be seriously intelligent about most of life's issues? Ten orders wouldn't be enough to solve the game of chess or even boost a stupid program to Master status. For instance, if you assume ten interesting moves in a position, then "looking ahead" one move (in a stupid way, by brute force alone) means evaluating 10^2 positions (two plies); three moves, 10^6, and so on. Current computers might be able to do four or five moves inside the limits set by conventional time controls. Special purpose machines might get to six. But I think it would be hard to beat a Master without looking ten moves deep, which means raking through 10^20 moves. And chess seems like a much simpler environment than many engineering domains. There may be no alternative to building genuine AI after all: AI that can create and learn Ur-patterns -- patterns that recognize different patterns as being instances of themselves. Alas it is not at all clear how faster processers will get us to that point. <-><-><-><-><-><-><-><-><-><-><-><-><-><-><-><-><-><-><-><-><-> In twenty years half the population of Europe will have visited the moon. -- Jules Verne, 1865 <-><-><-><-><-><-><-><-><-><-><-><-><-><-><-><-><-><-> Announcement Archive: http://www.pobox.com/~fhapgood/nsgpage.html. <-><-><-><-><-><-><-><-><-><-><-><-><-><-><-><-><-><-> Legend: "NSG" expands to Nanotechnology Study Group. The Group meets on the first and third Tuesdays of each month at the above address, which refers to a restaurant located in Cambridge, Massachusetts. The NSG mailing list carries announcements of these meetings and little else. If you wish to subscribe to this list (perhaps having received a sample via a forward) send the string 'subscribe nsg' to majordomo@world.std.com. Unsubs follow the same model. Discussion should be sent to nsg- d@world.std.com, which must be subscribed to separately. You must be subscribed to nsg-d to post to it and must post from the address from which you subscribed (An anti- spam thing). Comments, petitions, and suggestions re list management to: nsg@pobox.com.