[Up: prehistory] [Map] [No prior] [Robot Wisdom home page]
Newsgroups: rec.arts.books,comp.ai,sci.cognitive,sci.psychology.theory Subject: NEW: Does AI make philosophy obsolete? Date: 6 Sep 1995 08:21:06 -0500 Summary: the emerging consensus in science
After asserting that most philosophy has been made obsolete by this century's discoveries in biology and computer science, I was (predictably) challenged to prove this...
But alas, I was ill-prepared for this 'net.inevitable' reply... and it even seemed to me that to reply, I'd have to include a balanced summary of the full range of philosophical inquiry over the last three millennia!
But then I was reminded how Adler's survey of 'Great Ideas' claims to be such an overview, so I dug out my copy (two dense volumes) to see how hard it would be to extract a summary from it...
Adler has written 100 essays on 100 topics (the 100 Great Ideas), and then broken each idea into several dozen subtopics, and scrupulously collated references to each of these, from his canon of Great Books...
But even the 100 topics by themselves provide a pretty good overview of the historical range of philosophical topics, allowing me to work up a rough analysis of the different ways AI (especially) impacts philosophy. (I don't think I left out any categories, but I did include some twice. Many more could have been multiply classified, but to do them justice the analysis ought to be taken far beyond the basic 100. ...Any takers???)
The perspective I want to emphasize is that this century has brought a level of consensus on many of these topics, as to the sort of arguments that make sense, but this consensus hasn't really worked its way down (up?) far enough to overthrow the old ways of teaching philosophy...
Some of Adler's topics have been pretty solidly (ie, unanimously) grounded in quantitative science: Astronomy, Element, Matter, Mechanics, Nature, Physics, World.
Others have been put in a vivid new light by biology: Animal, Evolution, Immortality, Life and Death, Man, Medicine, Nature, Progress, Sense, Soul.
Some require a level of psychology that's very simple (but still mostly out of reach): Desire, Emotion, Experience, Habit, Happiness, Mind, Pleasure and Pain. (At this level, psychology-jargon becomes very dangerous, falsely reifying some very primitive abstractions.)
Some of Adler's 'Ideas' are concerned with how we know, and these are very directly addressed by AI: Definition, Dialectic, Form, Hypothesis, Idea, Induction, Knowledge, Logic, Memory and Imagination, Opinion, Reasoning, Science, Truth, Wisdom.
Many concern the sort of basic ontological categories that McCarthy was especially concerned with in his essay (and Lenat/Cyc, too, for sure): Being, Cause, Change, Element, Eternity, Evolution, Form, History, Infinity, Mathematics, One and Many, Opposition, Principle, Quality, Quantity, Relation, Same and Other, Space, Time, Universal and Particular.
Information theory and the first, crude model of the neuron challenge our old models of these: Art, Beauty, Language, Poetry, Rhetoric, Sign and Symbol. (Art, Beauty, and Poetry deserve more respect from AI, imho! They tell us something fundamental about cognition, but there's a sort of taboo against acknowledging this...)
Seeing social relations in a sociobiological context puts a new light on all of these (with AI supplying some insight into how they might have been 'implemented'): Courage, Custom and Convention, Duty, Education, Family, Good and Evil, Honor, Judgment, Labor, Love, Principle, Prudence, Sin, Temperance, Virtue and Vice, Will.
(John Maynard Smith and others have generated a set of arguments on these topics that are just vastly beyond the 'Greats', imho, and must be confronted-- though the consequences are not as bleak as some paint them. Their arguments are very subtle, and require a great breadth of self-knowledge to critique-- the ivory tower doesn't cut it...)
The same (I think) goes for these topics in political science: Aristocracy, Citizen, Constitution, Democracy, Government, History, Justice, Law, Liberty, Monarchy, Oligarchy, Punishment, Revolution, Slavery, State, Tyranny, War and Peace, Wealth.
And these categories may be teetering on the brink of complete obsolescence, unless somebody comes up with a miracle ;^/ Metaphysics, Philosophy, Religion, Theology, Angel, Chance, Fate, God, Necessity and Contingency, Prophecy. (I'm expecting something of a miracle, myself!)
Having advocated these views on netnews over the past few years, I've gotten detailed feedback (heh!) about what people find objectionable:
Scientists fear that science will be diminished by attention to art, and artists fear that art will be diminished by scientific analysis.
The anti-sociobiology contingent fears eugenics, and/or diminished human dignity.
The anti- symbolic-ai contingent assumes Cyc must fail (or protests, more simply, 'I don't wanna think about it').
The pro- social-science-jargon contingent turn a blind eye to the urgency of simulation.
And now, the anti- literary-progress (classicist) contingent is appalled that anyone would dream of retiring (eg) Plato...
j the robot wisdom pages jorn@mcs.com http://www.mcs.net../
From: jmc@Steam.stanford.edu (John McCarthy) Subject: Re: NEW: Does AI make philosophy obsolete? Date: 06 Sep 1995 15:01:57 GMT Organization: Computer Science Department, Stanford University
Congratulations on this post. I don't agree with all of it, but I think it is a brave attempt and partly successful, which is much better than par for the course.
-- John McCarthy, Computer Science Department, Stanford, CA 94305 * He who refuses to do arithmetic is doomed to talk nonsense. http://www-formal.stanford.edu/jmc/progress/