X-Message-Number: 25481 From: Tim Freeman <> Date: Thu, 6 Jan 2005 08:10:29 -0700 Subject: The Singularity Is Not A Fantasy For That Reason, IMO References: <> In message 25473, says: >Here's why: No one has the foggiest idea how to command, control, or >even orient just a single robot in a physical environment anywhere near >as complex and demanding as that of an assembler. In the simplest case, the physical environment of an assembler will be completely controlled by the people running it. For example, in Merkle's hydrocarbon assembly paper at http://www.zyvex.com/nanotech/hydroCarbonMetabolism.html the environment external to the assembler is a very pure mixture of a handful of different molecules, and the environment internal to the assembler has every atom in a known place. In this situation the problem of controlling an assembler will be comparable to the problem of controlling a fully automated factory. Everything in there is in a known location, by design, so you don't have to use general-purpose algorithms to go find things. >The nanotech apologists' counter to this is that, with the advent of >assemblers, real AI is just around the corner, and this AI can easily >handle the command and control issues of an assembler. Or a pound of >assemblers together. No worries. Strawman argument. Cite a credible nanotech apologist who says this. The actual counter is that controlling an assembler is fairly easy, once you know how to build and operate one. Operating an assembler "in the wild" would be more difficult than operating one in a controlled environment. Maybe that's what you're talking about. However, an E. Coli has enough smarts to self-reproduce in the wild, so the claim that AI is required even in this case seems dubious. -- Tim Freeman http://www.fungible.com Programmer/consultant in the Sunnyvale, CA area. I'm presently available. Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=25481