From extropians-request@gnu.ai.mit.edu Thu May 13 16:36:49 1993 Return-Path: Received: from usc.edu by chaph.usc.edu (4.1/SMI-4.1+ucs-3.0) id AA18689; Thu, 13 May 93 16:36:46 PDT Errors-To: Extropians-Request@gnu.ai.mit.edu Received: from wookumz.gnu.ai.mit.edu by usc.edu (4.1/SMI-3.0DEV3-USC+3.1) id AA14302; Thu, 13 May 93 16:36:40 PDT Errors-To: Extropians-Request@gnu.ai.mit.edu Received: by wookumz.gnu.ai.mit.edu (5.65/4.0) id ; Thu, 13 May 93 19:30:35 -0400 Message-Id: <9305132330.AA27407@wookumz.gnu.ai.mit.edu> To: ExI-Daily@gnu.ai.mit.edu Date: Thu, 13 May 93 19:30:05 -0400 X-Original-Message-Id: <9305132330.AA27400@wookumz.gnu.ai.mit.edu> X-Original-To: Extropians@gnu.ai.mit.edu From: Extropians-Request@gnu.ai.mit.edu Subject: Extropians Digest V93 #0259 X-Extropian-Date: Remailed on May 13, 373 P.N.O. [23:30:32 UTC] Reply-To: Extropians@gnu.ai.mit.edu Errors-To: Extropians-Request@gnu.ai.mit.edu Status: OR Extropians Digest Thu, 13 May 93 Volume 93 : Issue 0259 Today's Topics: AI: Synaptic Storage Requirements [3 msgs] AI:Human Equivalence [2 msgs] Drake Equation and Modelling of the Unknown [2 msgs] First CDs, then... [7 msgs] Magic Numbers [1 msgs] POLL: Expertise survey results [1 msgs] TECH/PHIL: (non)interactive entertainment [4 msgs] TECH/PHIL: (non)interactive entertainmentu [1 msgs] TECH: Numerical Speculation [1 msgs] Administrivia: This is the digested version of the Extropian mailing list. Please remember that this list is private; messages must not be forwarded without their author's permission. To send mail to the list/digest, address your posts to: extropians@gnu.ai.mit.edu To send add/drop requests for this digest, address your post to: exi-daily-request@gnu.ai.mit.edu To make a formal complaint or an administrative request, address your posts to: extropians-request@gnu.ai.mit.edu If your mail reader is operating correctly, replies to this message will be automatically addressed to the entire list [extropians@gnu.ai.mit.edu] - please avoid long quotes! The Extropian mailing list is brought to you by the Extropy Institute, through hardware, generously provided, by the Free Software Foundation - neither is responsible for its content. Forward, Onward, Outward - Harry Shapiro (habs) List Administrator. Approximate Size: 50806 bytes. ---------------------------------------------------------------------- Date: Thu, 13 May 1993 13:21:21 -0500 From: extr@jido.b30.ingr.com (Craig Presson) Subject: AI:Human Equivalence In <9305131609.AA22799@wookumz.gnu.ai.mit.edu>, "Kent Hastings" writes: |> Reply to: RE>AI:Human Equivalence#000# |> Someone who will remain nameless because I erased it wrote: F'bork. That was I. Freeman Craig the Uneraseable. |> > Most of us read at a few hundred bits per second, but we don't |> > memorize what we read, either. If we could read as fast as we |> > could form images of pages, we could beat Evelyn Wood all hollow! |> [...] |> Anyway, one of the presentations was about speed reading. The |> claim was made that a new speed reading method using cycled white |> noise in stereo headphones could guide readers past the |> "sub-vocalization" barrier permanently, after a month of use. [...] Hm, I could try that with my MasterMind DLS, just leave off the goggles. I went to an Evelyn Wood come-on session many years ago. I didn't read any faster by following my finger than not. ^ / ------/---- extropy@jido.b30.ingr.com (Freeman Craig Presson) / / ------------------------------ Date: Thu, 13 May 1993 14:20:58 -0400 From: "Perry E. Metzger" Subject: TECH/PHIL: (non)interactive entertainment X-Reposting-Policy: redistribute only with permission Marc.Ringuette@gs80.sp.cs.cmu.edu says: > I see what you mean, but my experience is very different. I think movies > are the greatest entertainment medium yet devised by man. I like them, > despite their lack of interactivity, because they're the closest thing > to immersion in a completely alien environment that I have access to. By > watching movies about diverse things and people, you extend your experience! > > Of course, I look forward to using a fully interactive immersion VR like the > holodeck from ST:TNG. I think it would be fantastic to really "be there" and > I wouldn't be too surprised if such VR's made reality obsolete. But even > after reality becomes obsolete, people will still want to watch movies! > (In the sense of being entertained linearly and non-interactively.) > Sometimes when a person wants to "veg out", they actually _need_ some > scripted entertainment. Movies are less stressful than life, and that's > a good thing sometimes. I think there is something else missed here -- as an artist, its much easier to produce a good story if you are telling it or showing it. I doubt VR environments will be the same sort of thing as a great book or movie at all. Perry ------------------------------ Date: Thu, 13 May 93 12:02:48 PDT From: hal@alumni.cco.caltech.edu (Hal Finney) Subject: AI: Synaptic Storage Requirements From: derek@cs.wisc.edu (Derek Zahn) > Anyway, the issue for us here is how difficult the modeling is likely > to be -- because it's convenient we say "how many bits to specify > each neuron/synapse/whatever?" Educated guesses of this let us > play scenarios about technological requirements, as well as let us > reflect on what we can actually expect from our current machines; > comments of "it's the software, stupid" are perhaps inappropriate if > reasonable neuron models can only simulate tiny networks on current > hardware. As I recall, the topic of discussion is whether it is useful to attempt to calculate the information capacity needed to store or simulate the brain. Paul's point was that we know so little that such calculations are not useful at the present time. I can't really agree with Derek's point that our current educated guesses, couched in "convenient" (for us) terms, tell us much about technological requirements. At one extreme we have calculations of the sort seen in Mind Children or Engines of Creation (neither of which I have handy, darn it): 10^10 neurons times 10^3 synapses per neuron times 10^2 bits per synapse, leading to 10^15 bits total (I just made these numbers up as an example). At the other extreme we can figure that if we record the position and type of every atom in the brain that ought to be enough (well, to be safe let's figure every molecule in the body since some learning takes place in the muscles and joints and nerves): About 10^27 atoms times enough bits for position information at Angstrom level in 3 dimensions, about 10^2, times 10 bits per atom for the type: 10^30 bits. This would be an upper bound. Ralph Merkle actually used a calculation something like this in a discussion of whether cryonics could work. The difference between 10^15 bits and 10^30 bits is so large that it demonstrates how little we know. > Also note that whenever a researcher sets out to develop a model of > neural function, s/he is implicitly doing exactly the same sort of > guesswork that we're talking about here -- by choosing the level of > description and complexity of the model that they propose and study. > All we're doing is expressing that guesswork in a different way. No, we're doing something different. The researcher knows that she may be a long way from reality, that her improved theory is just a tiny step towards that long-term goal. What we're doing is taking her theory and saying, if this is actually how the brain works, then it will take X number of bits to store the brain's whole state. But no one is proposing that this is actually how the brain works, least of all the neuroscience theorists. The extrapolation is based on a false premise. That's why the calculations which result are not likely to be useful. Hal Finney ------------------------------ Date: Thu, 13 May 1993 15:18-EDT From: Marc.Ringuette@GS80.SP.CS.CMU.EDU Subject: TECH/PHIL: (non)interactive entertainment Perry's on target when he points out the thorny authorship problems of interactive fiction. We'll need a whole new culture of authorship. For instance, an author's main job may become one of designing compelling characters with traits, motivations, and memories, rather than writing down a set of pre-packaged interactions which bring out those characteristics. We run the risk of moving towards fiction which is more interactive but shallower, until people figure out how to be outstanding authors in the new medium. And until enough geniuses come along and write interactive fiction, we'll still want to read Thomas Hardy. -- Marc Ringuette (mnr@cs.cmu.edu). Freely repost/archive any of my messages. ------------------------------ Date: Thu, 13 May 93 11:38:02 PDT From: hal@alumni.cco.caltech.edu (Hal Finney) Subject: TECH: Numerical Speculation Robin writes, responding to Paul Cisek: > > Why must speculation only be done in words and not with numbers? > While perhaps more work to construct, numerical speculations generally > make their assumptions more explicit, and thereby *help* us remember > what we do and don't know. What's wrong with numbers? I would not criticize numerical speculation in general, but rather speculation as it is often practiced, which gives a false sense of accuracy. Paul gave the example of the Drake equation for the number of civilizations in the galaxy. This did make its assumptions fairly explicit, and perhaps could be said to help us remember what we do and don't know. The problem was, we really know almost nothing about many terms in that equation. And the equation didn't make that clear, at least in the form in which it is often presented. What you really need to do with a numerical speculation is to identify a range of values for each parameter, along with some arguments for why that range is approrpiate. Applying this to the Drake equation would change the result from "1.2 civilizations per galaxy" (which is nonsense) to something like "more than 1E-40 and fewer than 1E+10 civilizations per galaxy, probably" (which makes clear how little we know). In discussions of the computing power needed to simulate the brain (in real time, a stipulation seldom mentioned), similar fudge factors should be made explicit. We don't know how many neurons there are; we can't be sure glial cells (which outnumber neurons by a factor of 10 or more) aren't involved; we don't really know what mechanisms are important in determining how individual neurons will behave. How do you quantify this? You just have to put in huge fudge factors if you want to do it at all. Then, the whole analysis tends to ignore the need for learning. Synapses change; neurons change shape; as Paul pointed out, the whole dendritic tree may need to be stored because it will determine which neurons are close enough that new synapses may conveniently form. All this is ignored in calculations of the form X neurons times Y synapses per neuron times Z bits per synapse. A simple numerical analysis of this form (which we have seen many times on this list) does not make its assumptions explicit. You have to read them into it: "let's see, he's allowing 30 bits per synapse, so evidentally he's assuming detailed synaptic shape is unimportant; he doesn't have any other state information than this, so evidentally he's assuming that changes in synapse state can be deduced by a model which knows only activation levels and current synaptic weights; etc." I don't think Paul's point was that numerical speculations are inherently invalid. It is rather that when we deal with a domain which has so many more unknowns than knowns, properly done numerical analyses will produce results with such massive error bars as to be virtually useless. In this domain, verbal arguments are more appropriate because they remind us that we are only waving our hands. Hal ------------------------------ Date: Thu, 13 May 1993 15:21:29 -0400 (EDT) From: Harry Shapiro Subject: First CDs, then... > Richard Kennaway says: > > It seems to me that the fundamental limit currently is not the time to > > print it, but the ability to produce one-off copies of acceptable physical > > quality and price. a conscious being, Perry E. Metzger wrote: > I disagree. > Here in the U.S., the bulk of book sales are made from large chain > stores like Waldenbooks and Barnes and Noble -- and made from stock at > the stores. People are used to instant gratification of their book > purchasing desires. To make this into a successful commercial venture > would require that people be able to buy things on the spur of the > moment and walk home with them a few minutes later. Retail shops that > require weeks for delivery won't make enough money to survive -- they > will be outcompeted by superstores. I agree that getting books I already know I want via some point of purchase printing program/setup would be good, I have two reservations. 1) A vast book warehouse with a touch tone based ordering system (push in the ISBN # and you account # and they send it) would meet many of Richard's needs and not require any technological advancements. 2) I like buying books. I have spent $5,000 a year on books, although I am now probally spending only about $1,000. I buy these books by broswing through them. A print on demand system would not work for me as I wan tot see the book before I buy it. -- Harry Shapiro habs@panix.com List Administrator of the Extropy Institute Mailing List Private Communication for the Extropian Community since 1991 ------------------------------ Date: Thu, 13 May 93 14:01:25 MDT From: edlane@godot.Eng.Sandy.Novell.COM (Ed Lane) Subject: AI: Synaptic Storage Requirements A (timely and relevant?) product announcment from a blurb in a magazine just picked up today at the news stand: "Computers With Humanlike Brains?" No computer can match the processing power of the human brain, but Intel Corp. and Nestor Inc. believe they've come much closer to that human ideal with a more advanced "neural network" chip they recently delivered to the govenment's Defense Advannced Research Projects Agency, or DARPA. The new chip, called the Ni1000, has 1024 artificial neurons. That's puny compared with the more than 100 billion neurons typical of a human brain. [...] ... the Ni1000 could improve the character-recognition capabilities of computers by three orders of magnitude, from 10 to 10,000 characters per second. Later this year, Intel says the Ni1000 should be available on add-in boards for PCs that use its 486 or Pentium microprocessors. OK, now the reference: Popular Science, June 1993. Page 47, Blurb by Sandy Reed. Any one have access to the Data Sheet on this chip? -Ed ------------------------------ Date: Thu, 13 May 93 15:11:07 CDT From: derek@cs.wisc.edu (Derek Zahn) Subject: AI: Synaptic Storage Requirements Hal writes: > As I recall, the topic of discussion is whether it is useful to attempt > to calculate the information capacity needed to store or simulate the > brain. Paul's point was that we know so little that such calculations > are not useful at the present time. One last response, then I stop debating the point for now. As to "usefulness", I suppose it's useful in tempering expectations for currently-available hardware. Also, I find it useful for my own thinking -- just like the Drake equation (of which I will apparently be a lonely defender), it serves to break the issues into pieces -- even if the error bars are large and even if there isn't much confidence in a "best guess", any style of calculation that reveals as many issues as we've discussed is useful -- if for nothing else -- as a rhetrorical device. How many of our conversations on this list are lots more "useful", anyway? Even if not "useful", perhaps "interesting". Presumably that's why such people as Moravec and Merkle and Schwartz and Waltz and Cherniak have found it worth their time to write about. > The researcher knows that she may > be a long way from reality, that her improved theory is just a tiny step > towards that long-term goal. What we're doing is taking her theory and > saying, if this is actually how the brain works, then it will take X > number of bits to store the brain's whole state. But no one is proposing > that this is actually how the brain works, least of all the neuroscience > theorists. Yes, I see that in papers all the time (er, I'm trying to be jovial while I whine here...): "This theory is not intended to even be close to reality; in fact, it's probably such an insignificant step toward the goal that the theory's complexity may well be fifteen orders of magnitude too simple to account for the phenomena we're interested in." And, as far as I saw, nobody was taking somebody's mechanical theory and translating it into information requirements -- the articles quoted in this discussion engage in "theoretical speculation" about the level of complexity that an adequate theory might need, based on trained intuitions about the structures under discussion. There are steps involved in this process, such as the "brain function reduces to cell function" theory I sketched in a previous note, as well as theories about synaptic capacity and so on (made by people with much more knowledge than I). But since such talk is useless, I'll stop polluting the list with it. derek ------------------------------ Date: Thu, 13 May 1993 14:10:47 -0400 From: "Perry E. Metzger" Subject: First CDs, then... X-Reposting-Policy: redistribute only with permission Richard Kennaway says: > When I look for a book that's in print, I usually have to order it, which > takes weeks. Being able to get out-of-print books in such a time frame > would still be an improvement on not being able to get them at all, or > making a bootleg photocopy of an interlibrary loan. > > It seems to me that the fundamental limit currently is not the time to > print it, but the ability to produce one-off copies of acceptable physical > quality and price. I disagree. Here in the U.S., the bulk of book sales are made from large chain stores like Waldenbooks and Barnes and Noble -- and made from stock at the stores. People are used to instant gratification of their book purchasing desires. To make this into a successful commercial venture would require that people be able to buy things on the spur of the moment and walk home with them a few minutes later. Retail shops that require weeks for delivery won't make enough money to survive -- they will be outcompeted by superstores. The real greatness of instant books is that you can avoid inventory and space problems. The Barnes and Noble superstore that just opened in New York has, in my estimation, about three million dollars tied up just in the inventory on the shelves (never mind in the basement), and being in Manhattan the price per square foot of floor space likely adds millions a year to overhead. If the store could be a tenth the size, offer more books and handle the same sales volume profit margins would skyrocket and even fairly expensive printing equipment and dispensery kiosks could be justified. Yes, the capacity to get ANY book that a person would want, including "out of print" titles, is a big plus -- but remember, "out of print" books tend to be out of print because they are too small a fraction of sales to justify the effort of stocking them. Some money will be made off of this capacity -- but not nearly as much as could be made by eliminating stock and excess space. Perry ------------------------------ Date: Thu, 13 May 1993 16:58:38 -0500 (EDT) From: rjc@gnu.ai.mit.edu Subject: TECH/PHIL: (non)interactive entertainmentu Opps, I forgot to mention that I take in a Japanese Anime once in a while. -- Ray Cromwell | Engineering is the implementation of science; -- -- EE/Math Student | politics is the implementation of faith. -- -- rjc@gnu.ai.mit.edu | - Zetetic Commentaries -- ------------------------------ Date: Thu, 13 May 93 11:42:55 PDT From: tcmay@netcom.com (Timothy C. May) Subject: First CDs, then... Two points on instant books: 1. How is browsing done? I suppose the bookstore could have *one* (or a few, for popular titles) copy on the shelf for browsing. But the "unlimited selection" model means this will not always be the case. I like to flip through my books before I buy them. In fact, I suspect most of us look through many books before buying one. Very seldom do I walk into a bookstore looking for a specific, hard to find book (which is the ideal case for "instant books"). I'm not saying here that the system won't work, only that such systems will likely fail if they don't cater to typical browsing tastes. (Perhaps high-res monitors can be used to "preview" the books...this'll cost some bucks. Perhaps "sample pages" can be printed for free (Ha!) or for some low fee (and I'd still miss the feel of flipping around the book to see what I was getting). Plenty of opportunities, plenty of problems. 2. Keith Henson, who apparently is not reading the List right now, tried to start a commercial project to do exactly this: high-speed on-site printing of books. He called it "Gutenberg Express." When he was doing this in earnest, circa 1985-6, he met with little interest/success. (A few years later the local Tower Records installed a "make your own cassette" system, from a company whose name escapes me, and this lasted for about 6 months....then the company folded.) -Tim -- .......................................................................... Timothy C. May | Crypto Anarchy: encryption, digital money, tcmay@netcom.com | anonymous networks, digital pseudonyms, zero 408-688-5409 | knowledge, reputations, information markets, W.A.S.T.E.: Aptos, CA | black markets, collapse of governments. Higher Power: 2^756839 | Public Key: PGP and MailSafe available. ------------------------------ Date: Thu, 13 May 1993 17:07:31 -0400 From: "Perry E. Metzger" Subject: First CDs, then... X-Reposting-Policy: redistribute only with permission Timothy C. May says: > Two points on instant books: > > 1. How is browsing done? I suppose the bookstore could have *one* (or > a few, for popular titles) copy on the shelf for browsing. But the > "unlimited selection" model means this will not always be the case. I'd assume you'd sit in front of a monitor, scan lists of titles or pictures of book covers, and if you wanted to look at something you'd select it and scan the pages on the monitor -- if you decided that you wanted the book, you'd stick your credit card into the machine, and out would spit a book. > I like to flip through my books before I buy them. In fact, I suspect > most of us look through many books before buying one. I guess you'd just browse electronically. > don't cater to typical browsing tastes. (Perhaps high-res monitors can > be used to "preview" the books...this'll cost some bucks. The printing system itself will likely cost $100k or more. The display system is cheap by comparison. > 2. Keith Henson, who apparently is not reading the List right now, > tried to start a commercial project to do exactly this: high-speed > on-site printing of books. He called it "Gutenberg Express." When he > was doing this in earnest, circa 1985-6, he met with little > interest/success. (A few years later the local Tower Records installed > a "make your own cassette" system, from a company whose name escapes > me, and this lasted for about 6 months....then the company folded.) The problem with the Personics system (which you are mentioning) was that it was a way to produce custom cassettes -- and it required significant work on the part of the patron, and most songs were unavailable in the system, and cassette sales in general have been dropping, and the cassettes you got weren't as good quality as normal ones. I suspect you can get around all these problems. Perry ------------------------------ Date: Thu, 13 May 93 14:04:09 PDT From: Robin Hanson Subject: First CDs, then... Harry Shapiro writes: >I like buying books. I have spent $5,000 a year on books, although >I am now probally spending only about $1,000. I buy these books >by broswing through them. A print on demand system would not work >for me as I wan tot see the book before I buy it. I buy by browsing too. Music and movies are bought less by browsing, so they are better candidates for other approaches. No accident that CDs are being done first. Robin ------------------------------ Date: Thu, 13 May 93 14:03:24 -0700 From: tcmay@netcom.com (Timothy C. May) Subject: Drake Equation and Modelling of the Unknown Derek Zahn writes: ... >As to "usefulness", I suppose it's useful in tempering expectations >for currently-available hardware. Also, I find it useful for my >own thinking -- just like the Drake equation (of which I will apparently >be a lonely defender), it serves to break the issues into >pieces -- even if the error bars are large and even if there isn't >much confidence in a "best guess", any style of calculation that >reveals as many issues as we've discussed is useful -- if for >nothing else -- as a rhetrorical device. > >How many of our conversations on this list are lots more >"useful", anyway? ..... >But since such talk is useless, I'll stop polluting the list with it. I'm a defender of the usefulness of the Drake equation, as a sounding board for debating the sizes of the terms, if not for the conclusions themselves. I see nothing wrong with trying to pin down some estimates of how many civilizations there may be in a galaxy, just as I see nothing wrong with trying to pin down some estimates of when general purpose assemblers may be built. Ditto for brain estimates, obviously (I stayed silent on this thread as I had nothing substantive to add). It seems obvious that making some estimates, which can then be critiqued and deconstructed, is better than answering "probably not many" or "a lot!" to the questions asked about numbers of civilizations, years to general assemblers, or bits in the brain. It makes sense to decompose estimates into smaller pieces which can be analyzed separately. In fact, I supported an idea that came up between Derek Zahn, Nick Szabo, and myself (I'm not sure if any of us discussed it on the list) some time ago to have a project to make more sophisticated estimates of some of these things....with the right kind of error bars, indications of nearly total uncertainty, critical path nodes, footnotes attached to the estimates, etc. The idea being to take something important to us, like the timetable for general assemblers, or the bit capacity of the brain, and come up with some models that could be critiqued, added to, assumptions checked, and timetables revised automatically as inputs changed. For the case of bits in the brain, a "spreadsheet"-like model that contained various terms (the fudge factors mentioned in the thread) and experimental or theoretical support for the values of each would be useful. To me, this is what hypertext is mostly about (of course, simple cross-references would do nicely...one doesn't need full-blown Xanadu-style links to get the benefits). As for the Drake equation itself, Frank Drake is here at UC Santa Cruz, so I suppose I could go ask him about his current thinking. Personally, I suspect we're nearly the only civilization in the galaxy, for reasons discussed many times on this list (Fermi paradox: why aren't they here already?, anthropic cosmological principle, etc.). (Barrow and Tipler made their own "Drake equation"-style estimates, with some terms given only within several orders of magnitude, and having this estimate was much more satisfying than a simple verbal estimate of "not many" or "lots" would have been.) -Tim May -- Timothy C. May | Crypto Anarchy: encryption, digital money, tcmay@netcom.com | anonymous networks, digital pseudonyms, zero 408-688-5409 | knowledge, reputations, information markets, W.A.S.T.E.: Aptos, CA | black markets, collapse of governments. Higher Power: 2^756839 | Public Key: by arrangement ------------------------------ Date: Thu, 13 May 1993 17:17:48 -0400 From: "Perry E. Metzger" Subject: First CDs, then... X-Reposting-Policy: redistribute only with permission Harry Shapiro says: > 2) I like buying books. I have spent $5,000 a year on books, although > I am now probally spending only about $1,000. I buy these books > by broswing through them. A print on demand system would not work > for me as I wan tot see the book before I buy it. You could still browse on line -- one could even build electronic "shelves" that one would virtually scan through on line... .pm ------------------------------ Date: Thu, 13 May 1993 17:11:29 -0400 (EDT) From: esr@snark.thyrsus.com (Eric S. Raymond) Subject: AI:Human Equivalence > This one's a real result, not a floating fudge factor. George Miller, > "The magical number seven, plus or minus two: some limits on our > capacity for processing information". _Psychological Review_ 63:81-97 > (1956). Yow! Thanks! I've been looking for that cite to add to the {magic number} entry in the Jargon File. The second edition of The New Hacker's Dictionary is in the pipeline now, BTW. -- Eric S. Raymond ------------------------------ Date: Thu, 13 May 93 14:37:47 PDT From: tcmay@netcom.com (Timothy C. May) Subject: First CDs, then... Perry Metzger writes: > > I like to flip through my books before I buy them. In fact, I suspect > > most of us look through many books before buying one. > > I guess you'd just browse electronically. > > > don't cater to typical browsing tastes. (Perhaps high-res monitors can > > be used to "preview" the books...this'll cost some bucks. > > The printing system itself will likely cost $100k or more. The display > system is cheap by comparison. Yes, any *single* high-res workstation is cheap by comparison, but check out a crowded bookstore, with several dozen patrons browsing the shelves! I doubt they'll be happy waiting for a screen to be freed up. (Someday, in-home or in-office systems may be common, and this wil alleviate the problem I just mentioned, but for now I assume we are talking about conventional bookstores replacing some or much of their printed inventory with high-speed printers and some means of previewing and ordering books.) I remain skeptical that a phase transition is imminent. -Tim .......................................................................... Timothy C. May | Crypto Anarchy: encryption, digital money, tcmay@netcom.com | anonymous networks, digital pseudonyms, zero 408-688-5409 | knowledge, reputations, information markets, W.A.S.T.E.: Aptos, CA | black markets, collapse of governments. Higher Power: 2^756839 | Public Key: PGP and MailSafe available. ------------------------------ Date: Thu, 13 May 1993 17:26:40 -0400 From: "Perry E. Metzger" Subject: Drake Equation and Modelling of the Unknown X-Reposting-Policy: redistribute only with permission Timothy C. May says: > As for the Drake equation itself, Frank Drake is here at UC Santa Cruz, so > I suppose I could go ask him about his current thinking. Personally, I > suspect we're nearly the only civilization in the galaxy, for reasons > discussed many times on this list (Fermi paradox: why aren't they here > already?, anthropic cosmological principle, etc.). Indeed, the real problem with the Drake equation is its assumption that the variables are independant, when they could easily be argued to be very dependant on each other. Perry ------------------------------ Date: Thu, 13 May 93 14:48:38 PDT From: tcmay@netcom.com (Timothy C. May) Subject: TECH/PHIL: (non)interactive entertainment Craig Presson writes, describing his entertainment tastes: > Haven't seen UteotW yet :-( Compression is great, but I have no idea what "UteotW" means! I am guessing this means "Until the End of the World," but it could mean other things. Could I suggest, politely, that people be less cryptic when asking hundreds of subscribers to do a macro expansion like this? TijaPC! -TFIWTJ -- .......................................................................... Timothy C. May | Crypto Anarchy: encryption, digital money, tcmay@netcom.com | anonymous networks, digital pseudonyms, zero 408-688-5409 | knowledge, reputations, information markets, W.A.S.T.E.: Aptos, CA | black markets, collapse of governments. Higher Power: 2^756839 | Public Key: PGP and MailSafe available. ------------------------------ Date: Thu, 13 May 93 17:07:47 CDT From: derek@cs.wisc.edu (Derek Zahn) Subject: Magic Numbers Eric Raymond exclaims: > Yow! Thanks! I've been looking for that cite to add to the {magic number} > entry in the Jargon File. That other "magic number" (just a couple of dribbly bits per second into LTM) comes (I think) from: Landauer, Thomas K. "How Much do People Remember? Some Estimates of te Quantity of Learned Information in Long-Term Memory". Cognitive Science 10 (1986). Pp. 177-493. derek ------------------------------ Date: Thu, 13 May 1993 16:54:12 -0500 (EDT) From: rjc@gnu.ai.mit.edu Subject: TECH/PHIL: (non)interactive entertainment For those extropians who watch tv or go to movies, what kind do you watch? I'm interested to see if extropians restrict themselves to "high art" or if they take in action/sf/comedy movies too. Here's what a usually watch on TV during the week: Saturdays: Star trek: The Next generation, Dr. Who Wednesday: Star Trek: Deep Space Nine. Tuesdays: Nova, if it's interesting Every Night if possible: Rush Limbaugh Sometimes: Entertainment Tonight Sometimes: C-SPAN Sometimes: Batman: The Animated Series It's about 5-6 hours a week, ~1 hr day. Last few movies I saw: Point of No Return Loaded Weapon 1 (big mistake) Unforgiven Movies I plan on seeing: Dave (a few have recommended it to me) Jurassic Park Arnie's ``The Last Action Hero'' I plan on seeing Jurassic Park for the same reason I wasted my time watching Lawnmower Man, to see the computer graphics. I plan on seeing Arnie's new movie because it is the most expensive movie ever made (>$100 million, more than Terminator 2) -- I'm curious as to why it costs so much. (it might be the advertising! Supposedly, the marketers are lauching a 1-mile long space billboard into low earth orbit) I'm a sucker for almost any SF movie no matter how bad it is. -Ray p.s. I would love to see a good screen adaptation of The Peace War, the Ungoverned, or Marooned in Real-time. -- Ray Cromwell | Engineering is the implementation of science; -- -- EE/Math Student | politics is the implementation of faith. -- -- rjc@gnu.ai.mit.edu | - Zetetic Commentaries -- ------------------------------ Date: Thu, 13 May 93 18:43:42 -0400 From: pavel@PARK.BU.EDU (Paul Cisek) Subject: POLL: Expertise survey results Here are the results of my survey on expertise. I thank all the people who responded - and apologize for the somewhat shoddy handling of this (I should have made the purpose more clear, provided a better template, allowed for personal info...). However, I think the result is somewhat interesting, and perhaps useful for people who have specific questions and don't know who to ask... The first part is some stats on broad areas of expertise, the second a listing by field, and the third a listing by name. As expected, the expertise of list members covers a pretty good range of fields, though it is somewhat heavily slanted toward computer related fields (again, not unexpectedly). If there are any mistakes in how I represented your information here, please let me know and I'll correct it. Also, if enough people request to be added to this I'll be happy to compose an update. I also welcome suggestions on how this could be improved. Thanks again to all who participated... Part 1) Here's a quick breakdown by broad areas of expertise. The numbers mean how many people expressed knowledge in a particular area. Computers: 24 Physics/Mathematics: 7 Economics/Finance/Law: 6 History/Philosophy: 6 Space Technology: 3 Part 2) Here's a more precise list of who expressed knowledge about a particular field. I've included only those fields that were mentioned more than once. Computer science: Bruce Baugh, Robert Brooks, Dave Burns, Peter Cappello, Alexander Chislenko, Paul Cisek, Anthony J. Gatlin, Dave Gordon, Amara Graps, Elias Israel, Tim May, Stanton McCandlish, Peter McCluskey, Harvey Newstrom, Ray Peck, Freeman Craig Presson, Mike Price, Eric S. Raymond, Marc Ringuette, Nick Szabo, Rens Troost Data encryption/security: Anthony J. Gatlin, Tim May, Marc Ringuette Economics: Dave Burns, Alexander Chislenko, David Friedman, Krzysztof Ostaszewski Electrical Engineering: Dan Davis, Mark DeSilets Finance/investment: Tim May, Krzysztof Ostaszewski, Nick Szabo History: Bruce Baugh, Blair Haworth Mathematics: Alexander Chislenko, David Friedman, Amara Graps, Krzysztof Ostaszewski Neural Networks / AI / Genetic Algorithms / Robotics: Paul Cisek, Dan Davis, Mark DeSilets, Peter McCluskey, Marc Ringuette, Nick Szabo Philosophy: Dan Davis, Stanton McCandlish, Eric S. Raymond, Rens Troost Physics: Dani Eder, David Friedman, Amara Graps, Tim May, Mike Price Science fiction/futurology: Bruce Baugh, Alexander Chislenko, Edward J. O'Connell Space: Dani Eder, Amara Graps, Nick Szabo Part 3) Here are the actual entries for all the list members who responded. (In an attempt at standardization, I needed to do some paraphrasing. I apologize if anyone's entry has been distorted - please let me know and I'll correct.) Bruce Baugh: Computer consultant/temp worker - history, science fiction, comic books - Bruce.Baugh@p23.f40.n105.z1.fidonet.org Robert Brooks: Computer engineer - VLSI design and simulation, multiprocessor computer architecture - rb@hprrb.rose.hp.com Dave Burns: Ph.D. student (Economics) - economics, computer programming - XIAOZHOU%PUCC@BUACCA.bu.edu Peter Cappello: Associate Professor of Computer Science, UC Santa Barbara - The relationship between algorithms and architectures with respect to scheduling concurrent computation and application-specific CAD tools. - cappello%cs@hub.ucsb.edu Alexander Chislenko: Software engineer - mathematics, CS, economics, futurology, global modeling, demographics, evolutionary theories - sasha@cs.umb.edu Paul Cisek: PhD candidate (Cognitive & Neural Systems) - computer science, neural networks, neuroscience - pavel@cns.bu.edu Dan Davis: PhD candidate (Electrical Engineering) - neural networks, learning theory, egoist philosophy (Rand and Stirner) - davisd@pierce.ee.washington.edu Mark DeSilets: Electrical Engineer - control systems, robotics, systems engineer - desilets@sj.ate.slb.com Dani Eder: Space Transport guru at Boeing - officer of Society for Creative Anachronism, physics - eder@hsvaic.boeing.com David Friedman: Academic economist - theoretical physics (PhD), economics, medieval cooking, medieval combat, law - ddfr@midway.uchicago.edu Anthony J. Gatlin: Cadet (Computer Eng.), US Military Academy - data and computer security, data encryption, applications programming, intelligence gathering - x62727g2@usma8.USMA.EDU Dave Gordon: President of Algorithmic Arts Inc. - computer graphics rendering, modeling and animation, software development and production - dgordon@crow.omni.co.jp Amara Graps: Scientific programmer - physics, numerical methods, planetary science - graps@argus.arc.nasa.gov Blair Haworth: PhD candidate (Military history/history of technology) - library operations, general bibliographic/reference technique - bhaworth@acpub.duke.edu Elias Israel: SunSelect Engineering - CS, X windows, MS-windows, OOP, language design, software management, SGML - eisrael@suneast.East.Sun.COM Tim May: Investor - physics, computer science, cryptology - tcmay@netcom.com Stanton McCandlish: BBS sysop, computer consultant - Celtic culture, smartdrugs, underground publication, fringe philosophy/subculture - anton@hydra.unm.edu Peter McCluskey: Grad student (Computer Science) - neural nets, C++ - pcm@cs.brown.edu Harvey Newstrom: A.S. Computer Science; B.A. Business Administration - CS, computer security, network engineering, network security, software engineering, nutrition, life extension, western occult tradition, and comparative religions - hnewstrom@hnewstrom.ess.harris.com Edward J. O'Connell: Science Fiction writer - Graphic Artist/DTP production slave/Imagesetter operator - ejo@world.std.com Krzysztof Ostaszewski (a.k.a. Krzys'): Chartered Financial Analyst - mathematics (PhD), actuarial science, mathematical economics, finance & investments, classical real analysis - KMOSTA01@ULKYVX.LOUISVILLE.EDU Ray Peck: Software guy - computer architecture/chip design, software - rpeck@pure.com Freeman Craig Presson: Software Scientist - OS, realtime systems, distributed systems, system management - extr@jido.b30.ingr.com Mike Price: Systems analysis/Software engineer - theoretical physics - price@price.demon.co.uk Eric S. Raymond: Systems programming - music, philosophy, martial arts - esr@snark.thyrsus.com Marc Ringuette: Grad student (Computer Science) - machine learning, information filtering, cryptography, robotics - Marc.Ringuette@GS80.SP.CS.CMU.EDU Anton Sherwood: Office drudge - heraldry, dabbler in linguistics, math - dasher@well.sf.ca.us Nick Szabo: Software engineer - CS, Unix, C, perl, genetic programming, finance, space industry, sociobiology - szabo@techbook.com Rens Troost: Systems Administration Manager - CS, Philosophy (18th/19th Century) - rens@stimpys.IMSI.COM Christopher Weeks: Undergraduate (technology education, ceramic art) - technology education, ceramic art - C576653@MIZZOU1.missouri.edu ------------------------------ End of Extropians Digest V93 Issue #0259 ****************************************