Get your own Free Home Page
KENNETH LIM @ ART PAGES
APPENDIX B
The Artificial Intelligence argument never ceases. One of the leading experts in AI technology, Marvin Minsky, continually pursues his support for AI research as featured in the following series of arguments he posted through the INTERNET system.
From: minsky@media.mit.edu (Marvin Minsky)
Subject: Re: Why want AI to fail?
Message-ID: <1993Aug26.014124.12695@news.media.mit.edu>
Sender: news@news.media.mit.edu (USENET News System)
Cc: minsky
Organization: MIT Media Laboratory
References: <1993Aug25.131352.25316@peavax.mlo.dec.com> <CCBp72.FC1@cs.bham.ac.u
k> <CCC5I5.7AK@fc.hp.com>
Date: Thu, 26 Aug 1993 01:41:24 GMT
In article <CCC5I5.7AK@fc.hp.com> allsop@fc.hp.com (Brent Allsop) writes:
>
>> And here's a question for you. Why do some people think that AI simply
>> must succeed?
>
> It must succeed and there are many obvious reasons why! I
>want to know what I am. I want to know how to improve myself. I want
>to be immortal. I want to be able to communicate actual experiences,
>i.e. I want to experience what you experience when you see red. I
>want to experience, real time, what my spouse is experiencing when we
>make love. I want to have a perfect memory. I want to work on 100
>things at once. I want to make 1000 copies of myself and have all my
>minds tied together consciously. I want to experience more than I do.
>I want to be able to buy a PhD in computer Science module at the store
>and install it in my brain. I want to make trips like Arnold did in
>Total Recall. I want to have true free will, in other words I want to
>be able to change what I want... and on and on and on...
[...]
> It's obvious. When we can do many of these things AI will
>have succeeded. Saying that AI will fail is saying that we will never
>be able to do any of these types of things and to me this is the only
>thing to fear.>
> Brent Allsop
Yay! I really share your attitude. You've put the finger on an entire civilization of gloomy sour-grapers. They pretend to think that humans are so infintely wonderful that nothing could possibly compare -- and in doing so, they discourage young people from working on the only possibility of improving us. We could even argue that if ]it weren't for that attitude, encouraged by most religions, science might now be a thousand years agead of where it is now -- and our 'mind-children' would already all have various forms of great -- and still growing --immortal brains.
The joke is that Pascal had precisely the wrong side of his famous wager: "What harm could come," he rhetorically asked, "from believing in God. If it exists I'd have a better chance for an afterlife, and if that's wrong, no harm is done." But consider that this very choice might have denied us precisely that reward.
From: minsky@media.mit.edu (Marvin Minsky)
Subject: Re: Why want AI to fail?
Message-ID: <1993Aug30.175304.10167@news.media.mit.edu>
Sender: news@news.media.mit.edu (USENET News System)
Organization: MIT Media Laboratory
References: <1993Aug26.014124.12695@news.media.mit.edu> <nagleCCDnCB.6yL@netcom.
com> <CCFAG0.K3y@psych.toronto.edu>
Date: Mon, 30 Aug 1993 17:53:04 GMT
Lines: 27
In article <CCFAG0.K3y@psych.toronto.edu> christo@psych.toronto.edu (Christopher
Green) writes:
>In article <nagleCCDnCB.6yL@netcom.com> nagle@netcom.com (John Nagle) writes:
>>minsky@media.mit.edu (Marvin Minsky) writes:
>>> ... You've [Brent Allsop] put the finger on an
>>>entire civilization of gloomy sour-grapers. They pretend to think
>>>that humans are so infintely wonderful that nothing could possibly
>>>compare -- and in doing so, they discourage young people from working
>on the only possibility of improving us. ...
>>
>What a brilliantly implemented false dichotomy, Marvin.
>I hear this from AI-ists all the time (Geoff Hinton, for one), and it is
>simply incoherent.
>The choice is not between AI and mysticism. There are (and have
>been) lots of alternatives to computationalism ...
Uh, huh. Name 3 that offer plausible such pathways to immortality. No fair mentioning
...breathing the breath of life into...
...discovering suitable non-mechanical causal powers,
...find the quantum-mechanical basis of consciousness
In other words, name three that a young person could start working on now.
From: gomez@barros.cs.ucf.edu (Fernando Gomez)
Subject: Re: Why want AI to fail?
Message-ID: <1993Aug31.073357.21086@cs.ucf.edu>
Sender: news@cs.ucf.edu (News system)
Nntp-Posting-Host: barros
Organization: University of Central Florida
References: <snodgras.746694268@crash.cts.com> <CCLAMs.IBq.2@cs.cmu.edu>
Date: Tue, 31 Aug 1993 07:33:57 GMT
Lines: 60
In article <CCLAMs.IBq.2@cs.cmu.edu> hpm@cs.cmu.edu writes:
>
>John Snodgrass writes
>> If someone said manned space travel was impossible in the 50s,
>>they were ninnies or fools or they simply meant it was impossible at the
>>time -- in which case they could be viewed as right.
>
> There were still some left in the 1950s, but many more
>earlier. They didn't mean it was impossible then, they meant it was
>impossible in principle. It's difficult for us to realize that
>proposals for space travel disturbed some people in past just as much
>as AI disturbs some now. You can get a flavor for it in chapter 5 of
>Willey Ley's old book "Rockets, Missles and Men in Space" From chapter
>5, summarising responses to Hermann Oberth's 1923 book "The Rocket
>into Interplanetary Space" (originally in German):
>
[Many things deleted here]
> -- Hans Moravec CMU Robotics
I have been reading this exchange with interest. I think that the important point in this discussion is "impossible in principle." That is what Searle and Co. need to prove. For those people who are interested in proves in principle, I recommend them to read Goedel. I am not as optimistic about achieving AI as Hans Moravec seem to be. Probably, I have read too much Kant and Wittgenstein, and my brain, like that of Don Quixote, has dried up. However, I have been always motivated and inspired by the optimism of Newell, Minsky and Simon, whose attitude I consider to be a very sane one when you are dealing with very difficult scientific issues. But, "cranking the cognition problem" is becoming a hard bone to chew. Nothing surprising about this. As G. Miller said long time ago, AI has made very little progress because Psychology had very little to offer to it. The problems we face in AI are, in my opinion, not engineering problems, but scientific ones. And,
against those people who think that at the root of this controversy it is funding money, they are wrong. There is very little money for strong AI, unless you are approaching it from a logicist point of view; and that money is scarce these days too, I think. It is hard for me to believe that, if you are interested in cognition, you could ignore AI (Models of Thought, SoM, SOAR, etc.). After all, Dreyfuss and Co have made a career by criticizing AI. And criticizing AI is still a good professional investment, unlike being a Marxist, say, which is is really down. But, I would recommened Marxism, especially in Europe, for the long-term intellectual investor.
-- Fernando Gomez
Chapter 1: Attention for Sale: Capitalism and Interactive Computers Chapter 2: Defining Human-Computer Interaction Chapter 3: Representing our Worlds: Digital Translation Chapter 4: Digital Intelligence: Parallel flow of Multiples? Chapter 5: Japanese Philosophy & Artificial Intelligence Research Soundwaves Conclusion Appendix A, B, C, D, E, F, G Bibliography
This page hosted by Get your own Free Home Page