Build Date: Tue Apr 23 17:50:08 2024 UTC

I hate to be a one issue voter, but the first candidate that promises public booze fountains gets my vote.
-- Johnnie Royale

Report from Spiritual Machines

by Arkuat

2000-04-04 16:53:58

Arkuat gives you the inside scoop on the "Spiritual Machines" panel and conclave. Wacky excitement ensues!

[Last Saturday (4/1/2000) a group of people at Stanford University came together to discuss a recent article by Bill Joy in W I R E D magazine ("Why the Future Doesn't Need Us", April 2000). The article discusses some of the difficult ethical questions that new technology is bringing. And by "new technology," I don't mean computers and cell phones: they're talking about nanotech, AI, and all the other crazy sci-fi stuff that is bringing us to "The Singularity."

Pigdoggers Arkuat and Gene Gene the Dancing Machine headed to the conference, and Arkuat submits this excellent report. Beaujolais! --Mr. Bad]

Okay, apparently Doug Hofstadter set this whole thing up, and he did a great job moderating. He invited a bunch of prickly egomaniacs, and managed to extract some useful information from them in public without allowing any murder or mayhem to be committed.

I think Ray Kurzweil opened. I don't really understand Kurzweil very well, to tell you the truth, but he seemed to be trying to soothe people into comfortable acceptance of the notion of using prostheses that are your intellectual ("spiritual", even!) equals or superiors. "How to Keep and Hire Ueberdaemon Godlings for your Domestic Help without getting Eaten by Cthulhu" or something.

Hans Moravec got up and tried to soothe people likewise, mostly be telling them how in ten years there would be a robot to do their vacuuming, and one to load and unload the dishwasher just five years after that. Gene kept muttering "Heil Hans! Heil Hans!" in my ears, excitedly, throughout Moravec's presentation.

Bill Joy got up and explained how back when he was busy trying to keep tabs on the moves of the research departments of Oracle and Microsoft and IBM, he didn't worry about this stuff because he assumed Moore's Law would hit a ceiling at about three or four atoms per transistor. With computational performance stuck there, he didn't think biotech/nanotech/robotics would get out of control. But a physicist friend of his apparently informed him in detail of a technique (which I don't yet have even the vaguest understanding of, myself) that would extend the progress of Moore's law one-million-fold (yes, six orders of magnitude) beyond where Joy had pegged the ceiling. This, he claims, suddenly made the prospect of full-speed development of biotech/nanotech/robotics far too frightening leading him and his buddies to call for an application of the brakes. He even called for a "relinquishment" of nanotechnological research.

Ralph Merkle got up and pointed out that you could relinquish published research in democratic societies all you wanted to, but that wouldn't do much to prevent secret research, and particularly secret research in undemocratic nations, going forward full speed ahead. It started looking like a classic Teller vs. Oppenheimer moment (mostly because Joy had commented on Oppenheimer's career at length). Merkle also claimed that in order to make informed and effective public policy decisions, policymakers should be exposed to as many results of early research in nanotechnology as possible, to minimize the chances of their making disastrous mistakes. Merkle did a good job of presenting the Foresight Institute party line in this one... forewarned is forearmed, if this stuff is dangerous, then the more we learn about it in advance, the better, etc etc etc.

John Holland and his student John Koza, the initial developers of genetic-algorithms, or genetic-synaptic-weighting of artificial neural network "synapses" (which is probably how the golems will be grown, ultimately, yes) seemed to be the token conservatives. They kept emphasizing how remarkable and poorly understood and incredibly powerful animal nervous systems (including human brains) are, and tried to soothe the audience by denying that human-equivalent machines were going to be available anytime before 2100 at soonest.

Frank Drake showed some nice slides about the most economical designs for very-large ground-based radio telescopes. Don't even ask me what this has to do with spiritual robots, but I guess Hofstadter had the idea that the only thing that could compete with spiritual robots in the Materialist Weirdness front would be Ay-leenZ, and so they had to be worked in there somewhere for extra color or something.

Kevin Kelley also did an "upbeat" piece, asserting that "replacement" was just the wrong way to think about this stuff, and that we would rather be "symbiotic" with the pheared new challengers. Stewart Brand later fed him a few softball questions from the audience. It made me wonder why these people, who have done some brilliant work, are wasting their time on their current "Clock of the Long Now" masturbatory shenanigans. I suspect that the orbits of the Earth, Moon, Mars, etc will continue to be keeping good time long after their vaunted clock has fallen into disrepair, though I'm still not sure where they are planning to locate the thing. Even after we disassemble the various large bodies of the solar system for spare parts (and probably the Clock with them), the nearby pulsars will still provide excellent time for gigayears. But this is all beside the point, and has nothing to do with Spiritual Robots either. I just wanted to comment yet again on my continuing incredulity on the vagaries that age-related mental decline can induce in members of my species.

A couple of times someone in the audience tried to ask why there were no non-materialists on the panel (though they usually asked this much less succinctly, rambling about God and summarizing a few idealistic metaphysical theories instead or asking why the Dalai Lama or the Pope wasn't up there too). Hofstadter's reply (paraphrased rather brutally) was "Yeah, there are no non-materialists on this panel because I planned it that way. Next question?"

After Moravec and Kurzweil and Merkle had combined to make Bill Joy extremely uncomfortable and upset (Merkle seemed to think that Joy was dangerously insane and irresponsible, and Joy seemd to think that Merkle was dangerously insane and irresponsible), Hofstadter closed by posing a question to Moravec about whether it would be so terrible if human beings really were replaced by their memetic offspring, and then read a quote from a Moravec interview (I wish I had it handy, because it was beautifully phrased, but I don't... perhaps Gene can find it?) to the effect that getting replaced by the progeny of your mind would be much less of an evolutionary failure than stagnating as an unchanging, "stuck" species for billions of years would. The latter, he said, would be the real failure. Moravec just said that he still stood by his earlier reply, and the crowd broke out into vigourous applause while Bill Joy looked slightly puzzled and very indignant.

Over.  End of Story.  Go home now.

T O P   S T O R I E S

Spam Assassin - a mail filter to identify spam

C L A S S I C   P I G D O G