X-Message-Number: 946 From: (David Krieger) Subject: Re: Luddites Date: Thu, 2 Jul 92 11:10:38 PDT > Message: #942 - Re: Luddites > Message-Subject: More Luddites > Date: Wed, 01 Jul 92 16:46:28 -0400 > From: > I think the point of disagreement between Tim Freeman and I is that he thinks human capabilities are fixed and immutable, and I think that human capabilities can be enhanced. He quotes: > Thomas Donaldson <> says: > >So long as human beings command these tools to be made, they won't > >make them so as to put themselves out of business. Or make THEMSELVES > >obsolete. > > This is a tragedy of the commons situation. There won't be anyone > saying "let's build this new machine to make humans obsolete". > Instead, they'll be saying "let's build this new machine to make a > little bit more profit to make our stockholders happier". Eventually > "this new machine" will be something that fulfills the role of CEO, > and most of the stockholders will be other corporations with > recently-installed artifical CEO's. What seems more likely to me is that the CEO will enhance himself technologically, proceeding from "human" to "transhuman". This does not mean necessarily a complete Moravecian upload, but will most likely be progress along a continuum. The human or transhuman stockholders will probably be enhancement consumers as well, "keeping up with the Joneses" to stay competitive. > The economy will march happily > onward without us, assuming we can design a non-anthropocentric legal > system. Assuming the legal system continues to be human-centered, the > humans will become figureheads. ("Potted plants" was the term I used > in an earlier post.) This scenario requires that we will need to build machines more intelligent than human beings. No one has demonstrated yet that this can be done. I think the capacity for enhancing the capabilities of human beings (which we already know to exist, and presume to be intelligent :-) is at least as plausible, and more parsimonious, than worrying about the hypothetical capabilities of AI's we haven't built yet. > And (David Krieger) says: > > >Countless times in the past, occupations and > >the detailed technical abilities associated with them were "obsoleted" > >... by technological developments. > > I think we're using different meanings for "abilities". I'm using a > model where a human uses a little bit of domain knowledge and a large > amount of skill common to all humans to get the job done, and your > model seems to put everything into the domain knowledge category. Well, my model is the one that's right :-). See below. > >The computer was supposed to put hordes of file clerks and bookkeepers > >out of work -- and it did, but it also created even more jobs, for > >programmers, operators, ... [etc.]. > > Using "dexterity" as an abbreviation for all the skills people easily > learn but don't need to be explicitly taught, let's go through the > scenario for the file clerk: File clerking is already an obsolete occupation. The actual *function* of the file clerk was the manipulation and storage of *information*. The dexterity you believe to be important had to do with moving around the bits of paper to which the information was coincidentally attached. Computers make better file clerks than human beings already (with the aid of *programmers*, which I'll get to shortly. So, I think the scenario you present is already invalidated. > The file clerk needs to know a little bit about how the file system > works, and a lot of "dexterity". I think you have the proportions here precisely reversed. The real value of the file clerk is that *he knows where the information is and where it goes*. When the file system and 20 or so clerks are replaced by a file server, there are no longer bits of paper to be moved around -- the function continues to be performed, but because the little bits of paper are no longer employed, dexterity is of null value. Meanwhile, the file server is very good at moving the information around internally, but needs to be directed *every step of the way* by human intelligence (in the form of its stored program, and the input of human operators). > Now the clerk needs to try to learn a skill that isn't mostly > "dexterity" before the machines get that skill too. The invention of > the PICK-EM-UP makes the problem of retraining a displaced worker much > more difficult than it ever has been in the past. Before this time, > "dexterity" and a little bit of training could make a person > employable. After this time, lots of training is necessary to make a > person employable. As the amount of training required increases, the > desirability of instead designing a machine to do the job increases. As but the plausibility of designing a machine to do the job *decreases*. This was my point in my earlier post, which evidently wasn't sufficiently clear. You maintain that most occupations involve a lot of dexterity and a little domain knowledge. On the contrary, most occupations involve a small amount of physical dexterity (which, as you point out, nearly everybody has), and a large body of domain knowledge. I'm sure you are at least as physically dexterous as the average 19th-century blacksmith... but you lack the domain knowledge to do what he does (which was what the example of the horseshoe was supposed to demonstrate). > At some point the training ceases to be worthwhile. There will be a[n] > increasing chunk of the population that just can't keep up. On the contrary, this is the point at which the training becomes indispensable. I reject your view that most people aren't trainable. The Japanese are kicking our butts economically not because the average Japanese person is inherently smarter than the average American -- that is a racist contention with no basis in biological fact -- but because (among other factors) our publically-funded education monopoly does such a crappy job. Japan, with only half our population, has *more* scientists and engineers than we do. If I accepted your untrainability argument, I would be forced to conclude that the Japanese have some overwhelming and inherent superiority over the Americans, which is patently false. in raw brain-power. > >The effect of technology is not to eliminate jobs, but to move the > >available jobs to a higher skill level. > > Right! And eventually the required skill level is not achievable by > mere humans. You have a dim and unflattering opinion of humanity, it seems. The animal that was once jaguar fodder now leaves footprints on the moon. I refuse to believe that our technological abilities, turned on ourselves, can't increase the capacities that blind evolution gave us. If we can't, we may as well chuck the whole immortality idea, because something better suited to represent the future of life in the universe is bound to come along. > Message: #943 - Re: Luddites > Date: Wed, 1 Jul 92 10:09:43 PDT > From: Bruce White 3807 <> > Message-Subject: Re: #940 > > > Message: #940 - Re: Luddites > > From: (David Krieger) > > Message-Subject: Re: Luddites > > Date: Tue, 30 Jun 92 9:51:18 PDT > > > The computer industry now has a negative unemployment rate, as jobs go > > advertised and unfilled for months or years. > > NOT! > > The computer industry as a whole is in a slump and as more > aerospace companies continue to cut back, alot of software engineers > will be out of work. Hughes announced a layoff of 9000 people yesterday. > The software consulting arena in California is also very soft. > > ---------------------------------------------------------------------- Ja, and IBM has just "rescinded its full-employment policy" as well (that's an even better euphemism than "downsizing"!). I sure picked a bad illustration there... or did I? Perhaps I should have said the *global* computer industry has a labor shortage? At any rate, the "myth of the firing squad" still doesn't apply. Those laid-off workers represent an enormous capital resource in training and domain knowledge (if they don't, they shouldn't have had their old jobs in the first place!). Many of them might start their own companies -- that is, if they could raise the necessary capital. (Remember that Autodesk was founded in 1982 with an initial net worth of only $50,000.) Borrowing could be a problem, because in 1980 the U.S. government raised the limit on FSLIC insurance from $40,000 per depositor to infinity -- directly causing the savings-and-loan collapse. And venture capital is hard come by nowadays, too, because the government is trying to "curb the abuses" of junk bonds. Hmm, the government doesn't seem to be on our side, does it? That slump in the domestic computer industry might last a while, at that. The relationships between the U.S. and Japanese governments and their respective business communities is interesting, and the differences in the two approaches quite enlightening. All futurists really should read _Bionomics_ by Michael Rothschild. It's an eye-opener. dk Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=946