From extropians-request@gnu.ai.mit.edu Sun May 30 16:07:57 1993 Return-Path: Received: from usc.edu by chaph.usc.edu (4.1/SMI-4.1+ucs-3.0) id AA21230; Sun, 30 May 93 16:07:55 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 AA12527; Sun, 30 May 93 16:07:51 PDT Errors-To: Extropians-Request@gnu.ai.mit.edu Received: by wookumz.gnu.ai.mit.edu (5.65/4.0) id ; Sun, 30 May 93 19:02:16 -0400 Message-Id: <9305302302.AA25323@wookumz.gnu.ai.mit.edu> To: ExI-Daily@gnu.ai.mit.edu Date: Sun, 30 May 93 19:01:53 -0400 X-Original-Message-Id: <9305302301.AA25316@wookumz.gnu.ai.mit.edu> X-Original-To: Extropians@gnu.ai.mit.edu From: Extropians-Request@gnu.ai.mit.edu Subject: Extropians Digest V93 #0290 X-Extropian-Date: Remailed on May 30, 373 P.N.O. [23:02:14 UTC] Reply-To: Extropians@gnu.ai.mit.edu Errors-To: Extropians-Request@gnu.ai.mit.edu Status: OR Extropians Digest Sun, 30 May 93 Volume 93 : Issue 0290 Today's Topics: Chaitin, Holland, & AI [1 msgs] ECON/SPACE: Monopolizing the Solar System [3 msgs] ECON/SPACE: Sell the Universe Whole [1 msgs] ECON/SPACE: The Loonie Treaty -- irrelevant? [1 msgs] FTL: Traversable Wormholes 1 of 1 [1 msgs] Forwarded article. [1 msgs] HEALTH:Kenneth Seaton & Albumen? [1 msgs] New Brain Test [1 msgs] New Brain Test: "Einstein einstein.... Penrose!" [4 msgs] Response to last message Cold Fusion Not! [3 msgs] TECH: Cold Fusion Not! [1 msgs] VR: [perceived] motion sickness [1 msgs] Whole Earth Review Crypto Article [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: 52108 bytes. ---------------------------------------------------------------------- Date: Sun, 30 May 93 06:40:08 GMT From: price@price.demon.co.uk (Michael Clive Price) Subject: ECON/SPACE: Sell the Universe Whole Freeman Craig Presson: > Find it before you sell it dep't: > > UPI on 5/26/93 repeats a report from a meeting of the American > Geophysical Union in Baltimore to the effect that the Voyager > spacecraft, now 4-5 billion miles from the Sun, have begun to detect > the predicted LF radio emissions from the heliopause, thus providing > our first direct observation of the "boundary" of the solar system. Do say more.... What is the heliopause? Is it where the solar wind dies out? And where does it lie with respect to the Oort belt? Is this the Voyager craft(s) with the infamous 'Sounds of Earth' on? Mike Price price@price.demon.co.uk AS member (21/3/93) ------------------------------ Date: Sun, 30 May 93 07:15:47 GMT From: price@price.demon.co.uk (Michael Clive Price) Subject: New Brain Test: "Einstein einstein.... Penrose!" First James A. Donald saith: > Since the amount of processing power required goes up > nonpolynomially fast with the scaling variable, (in the > case of "neural nets" the size of the neural net") After being refuted, he now saith: > I meant that the size of the neural net growns non > polynomially fast with respect to the scaling variable. So you've already started to slide and wriggle. Just now you said the scaling variable was "the size of the neural net". What do you mean? I suggest you define the 'scaling variable' as log(net size) then you can claim that net size grows exponentially with the scaling variable. Of course processing power required will relate to net size and not this rather peculiar measure, but I doubt this will worry you (despite flatly contradicting what you said earlier). > The fact that there are such things as category recognition > neurons is in itself close to being disproof of the conventional > model of a neuron. If I were a cruel man I'd ask you to explain this and watch you go through gyrations. I'm a cruel man :-) Explain. > James A. Donald Mike Price price@price.demon.co.uk AS member (21/3/93) ------------------------------ Date: Sun, 30 May 93 7:31:37 GMT From: starr@genie.slhs.udel.edu Subject: ECON/SPACE: Monopolizing the Solar System >From: hal@alumni.cco.caltech.edu (Hal Finney) >Subject: ECON/SPACE: Monopolizing the Universe > >One example I have thought of where lack of property rights led to over- >rapid homesteading was the exploration of the New World by Spanish and >Portuguese explorers in the 17th and 18th centuries. I thought of the same example, and think that Robin's proposal is about as absurd as the Pope dividing the New World. >My knowledge of >this is limited to elementary school lessons, but as I understand it >both parties were engaged in aggressive exploration/colonization efforts >in order to allocate as much as possible of the Americas to themselves >under a homesteading-like system. This was expensive and dangerous. Sort of. They were agents of States claiming territory, but most of it was already owned by indigenous people. The danger came mostly from these and from competing European invaders. The expense was MORE than mitigated by the gold mining done in the New World and the gold shipped back to the Old. >They got the Pope to solve the problem by dividing the territory along >a particular meridian, Spain to the west and Portugal to the east. Although >this was not done by an auction, the point is that the allocation of the >resources solved the ownership problem and enabled a more efficient and >practical rate of development of the new land. It's not always true that >faster exploitation is better. This merely took care of the threat of war between Spain and Portugal, elimi- nating one of the sources of danger (and the most significant one). The Europeans were then left free to invade, rape, pillage, conquer, and enslave the indigenous people as they pleased - they did have overwhelming force due to superior technology, after all, and diseases they'd developed immunities to that the indigenous people hadn't. That this "enabled a more efficient and practical development" remains to be shown. You've begged the question. Certainly the Spanish and Portugese States and their agents profited greatly, but they imposed costs on others. Relative rates of economic growth have to be taken into account, too. Nor did I claim that "faster exploitation is better." I claimed that the sooner resources are exploited, the better. That is, the sooner they're brought into the sphere of production, the better. This has nothing to do with the rate at which they're used. Assume two savings accounts with identical interest rates and initial real deposit amounts. One starts now, another ten years from now. Which one yields more benefits to the depositor? To borrowers? To lenders? Tim Starr - Renaissance Now! Assistant Editor: Freedom Network News, the newsletter of ISIL, The International Society for Individual Liberty, 1800 Market St., San Francisco, CA 94102 (415) 864-0952; FAX: (415) 864-7506; 71034.2711@compuserve.com Think Universally, Act Selfishly - starr@genie.slhs.udel.edu ------------------------------ Date: Sun, 30 May 93 10:16:59 GMT From: price@price.demon.co.uk (Michael Clive Price) Subject: FTL: Traversable Wormholes 1 of 1 A while back Ray Cromwell said: > I just had a thought. Wouldn't exotic matter allow us to > open up a closed contracting universe by "sending matter off into > the sunset" ;-) Yes, my thoughts too. Although it might only be possible to open up some parts of a contracting universe. Use the exotic matter to create a wormhole extension to basement/laboratory/baby universe and disappear down it if the mother universe goes kaput. > -- Ray Cromwell Mike Price price@price.demon.co.uk AS member (21/3/93) ------------------------------ Date: Saturday, 29 May 1993 07:06:31 PST8 From: "James A. Donald" Subject: New Brain Test: "Einstein einstein.... Penrose!" Michael Price deliberately misrepresented my post. In <9305300620.AA19242@wookumz.gnu.ai.mit.edu>, price@price.demon.co.uk (Michael Clive Price) wrote: > What > do you mean? I suggest you define the 'scaling variable' as As I said, and he deleted, the scaling variable is the number of degrees of freedom in the most general category represented by the category recognition neurons. The required size of the "neural net" grows nonpolynomially fast, in practice exponentially fast, with the size of the scaling variable. Thus enormous increases in the size of the "neural net" will yield only minor and trivial improvements in its flexibility. > > The existence of category recognition neurons is in itself almost a > > disproof of NN theory. > If I were a cruel man I'd ask you to explain this and watch you go > through gyrations. I'm a cruel man :-) Explain. Before category recognition neurons were found, NN people claimed that there could not be such a thing. I take it NN theory has been appropriately revised to accommodate painful reality. It is clear that a large proportion of neurons, probably most, have straight forward real world meanings. For example a feature detection neuron might represent the presence of an edge with a particular orientation in a particular part of the visual field. The output of a motor neuron might represent the desired tension in a particular muscle. A more abstract neuron might represent the intention of moving some part of your body in a particular real world direction. So far no problem with NN theory. The problem is that amongst these neurons are a great many neurons whose output represents a wide variety of high level categories, for example "face of adult male", or the face of a particular individual. NN theory assumes that high level categories are not generated directly, but rather are a state of the network as a whole or a large subnetwork. Yet they are generated directly. Hence my comment that the existence of category recognition neurons are almost a disproof of NN theory. There are particular neurons that recognize particular high level categories. The location of these neurons and their timing constraints indicates that there is nothing unusual about their inputs. They collect their inputs from the same net as their neighbors. This looks very much as if the magic is not happening in the network as a whole. Each particular separate magic is happening in a separate particular neuron, each separately processing a large number of particular inputs. Where are the intermediate level abstractions? What are the intermediate level abstractions? Nobody can define them, nobody can find them. Thus for example it is easy to find neurons early in the sheep's visual pathway that respond to the face of a horned sheep, but impossible to find neurons that respond to horns in isolation. High level categories magically appear from low level features without any large intermediate process. --------------------------------------------------------------------- | We have the right to defend ourselves and our James A. Donald | property, because of the kind of animals that we | are. True law derives from this right, not from jamesdon@infoserv.com | the arbitrary power of the omnipotent state. ------------------------------ Date: Sun, 30 May 93 12:09:58 EDT From: Xiao Zhou Subject: HEALTH:Kenneth Seaton & Albumen? I just saw Kenneth Seaton on Tony Brown's Journal making some very provocative claims about hygiene, health, and albumen. Does anyone know the story on this guy? It would be difficult for me to summarize his claims. Suffice it to say that he presents some interesting opportunities for extropians if he's not a crackpot. His claims toward the end began to sound rather extreme, and he didn't provide copious citations of where to find these results. I didn't see the beginning of the program, so I may have missed something important. They did mention some report of the American Heart Assn.(?) ------------------------------ Date: Sun, 30 May 93 10:25:37 PDT From: hal@alumni.cco.caltech.edu (Hal Finney) Subject: New Brain Test: "Einstein einstein.... Penrose!" From: "James A. Donald" > Thus for > example it is easy to find neurons early in the sheep's visual pathway that > respond to the face of a horned sheep, but impossible to find neurons that > respond to horns in isolation. High level categories magically appear from > low level features without any large intermediate process. While this is an interesting discovery, it is not that different from what happens with the simple simulated neural networks used in research or engineering. Suppose a NN is trained to recognize letters of the alphabet, a common task. We give it an R, say, and it recognizes that. Now we can decompose the R into features: a vertical bar on the left, a round loop at the top, and a short slanted bar on the lower right. Does this mean that since the net recognizes R it is guaranteed to have earlier-stage neurons that recognize these three features? I think not. While the intermediate neurons do important processing of the image, the features they look for may not correspond to our intuitive ideas of what features make up a letter. It is entirely possible that the R may be identified by having two little lines on the bottom, a horizontal line on the top, no verticals in the middle, and the conjunction of pixels at the top left, lower right, and middle. In general, there are many possible ways of analyzing an image by a NN. The failure of a given network to break an image down in the way which we find intuitively obvious does not mean that NN models fail. Even our own visual system, in the early stages, does not analyze an image in the way that we break it down afterwards. How many of us think of what we see in terms of moving edges and contrasting dots? Only a few optical illusions and the experiments of neuroscientists expose these details. So, although I am impressed that sheep-recognizing cells exist "early" in the sheep visual pathway, the failure to find horn-recognizing cells feeding into them says nothing about the types of processing done by neurons. Anyone who has run many NN models will have seen similar results. Hal Finney hal@alumni.caltech.edu ------------------------------ Date: Sun, 30 May 1993 10:30:53 -0700 (PDT) From: szabo@techbook.com (Nick Szabo) Subject: Chaitin, Holland, & AI I'd like to thank Derek Zahn for his earlier comments on my Chaitin speculations post. He poked so many holes I'm still busy patching them up. :-) Tim May informs me that stochastic automata have already been written up in the literature under the name "Probabilistic Turing Machines". Rather than being startingly original I've blundered into another Schelling point in k-space, (sigh). To review and extend those speculations which emerged largely unscathed, I maintain the following hypotheses: * Either the Church-Turing thesis or the diagonalization proof of Turing Machine computational incompletness is wrong, as demonstrated by Chaitin's program which compresses the general solution to the halting problem, something a TM is not supposed to be able to do. * Turing Machines with an infinite random-number stream input tape provide a more succinct or general computational model for certain classes of problems that traditional TMs. Computational models that insist on finding a complete solutions to a class of problems are much weaker for these problems than models which are satisfied with good but approximate solutions. * The fact that traditional algorithm-complexity theory shows a computation to be NP-complete for the exact solution doens't mean that a very good approximate solution cannot be found in a vastly shorter time, eg via genetic algorithms which eliminate O(k^n/n) possibilities per step. * The vast bulk of practical human intelligence involves finding approximate solutions. Some general thoughts on AI: I distinguish between two kinds AI: (1) functional AI, which tries to duplicate the desirable capabilities of intelligence in their most succinct or efficient Platonic form -- everything from the ability to do arithmetic and sort numbers to the ability to do symbolic regression of quartic polynomials and construct default heirarchies from noisy data. (2) Modeling AI, which tries to model animal and human intelligence, striving for generality and faithfulness to the original rather than utility or performance better than the original. Traditional AI tends to fall into category (1), neural nets into category (2), although both camps pursue both goals, since the distinction has seldom been made. Clearly functional AI has succeeded in many areas, especially if we include the broad sweep of computer science under this rubric (as I think we should; all of computer science is concerned with implementing and extending aspects of human intelligence on computers). Modeling AI has largely failed to model human intelligence in terms of traditional Platonic computer science or AI research; better results are being had with more biologically faithful models like neural nets. I see the two fields diverging. One reason is that human brains are not the zenith of intellectual capability. Computers, those idiot-savants, are vastly *more* intelligent than humans in specific areas: spelling, grammar, many kinds of practical mathematics, sorting, searching, processing and layout of text and graphics, etc. Functional AI's most productive efforts come from amplifying specific intellectual capabilities thousand- and million-fold, not in merely implementing the same human capability on a computer. Functional AI benefits from specialization. On the other hand, modeling of human intelligence greatly suffers from both specialization and the commercial requirement that intelligence implemented on computers be better than human intelligence to be commercially valuable. Modeling AI strives for completeness and accuracy; it is getting better but has a long way to go. Functional AI might skip straight to radically different forms of "thought" that might have little to do with everyday human intelligence, such as formal logic, mathematical laws of physics, genetic algorithms, etc. Functional AI has arguably implemented only a small fraction of the utility of human intelligence, but much of what it has implemented has greatly surpassed the capability of its human counterpart. I plead for researchers in these different niches to recognize and value their differences instead of carping at each other. We will get nowhere complaining about "the failure of AI", ad nauseum. On the "all crows are black" conundrum, saying "all X's are Y" is hardly ever completely true. As proof, I'll spraypaint a crow hot pink. :-) A less silly example is the statement "all birds fly". This is usually true and thus is a very useful rule, because in information-theoretic terms it highly compresses a very large number of observations. But it is an *approximate* compression. Penguins don't fly. A more complete, but less compressed structure is the default heirarchy: "in general all birds fly, except ostriches and penguins don't". A widely used form of default heirarchy is the class heirarchy in object-oriented programming. Korzybski famously declared, "the map is not the territory". I'd rephrase that, "the compression is not the expansion." The default heirarchy, or any other logical structure, is always going to be a Platonic, idealized, incomplete model of the real world. By adding exceptions we can get closer, but we can never be exact. We must trade off compression for completeness. The cutting edge of functional AI, that is to say the work that IMHO provides the most general model of posthuman intelligence, the greatest potential for compressing observations, is the "emergent default heirarchy" (EDH) of John Holland et. al. Holland is the inventor of genetic algorithms; IMHO Holland ranks up there with Turing and Von Neumann as the greatest computer scientists of the 20th century. EDH's encode default heirarchies of synchronic and diachronic logical rules into a genetic code, over which genetic operators (eg crossover) and fitness functions are defined, implementing genetic algorithms. Teaming up with Holland are a philosopher of science and two cognitive psychologists. Their preliminary ideas can be found in Holland et. al, _Induction: Processes of Inference, Learning, and Discovery_, MIT Press 1986. A streamlined, more general heirarchical GA has been implemented by John Koza's genetic programming, which evolves Lisp s-expressions, cf. Koza, _Genetic Programming_, MIT Press 1992. Koza's GPs could be used to implement Holland's EDH's; such a combination might for example evolve object-oriented programs, a more general model of problem-solving than the various terminal sets currently used to solve different kinds of problems under GP. Nick Szabo szabo@techbook.com ------------------------------ Date: Sun, 30 May 1993 10:43:58 -0700 (PDT) From: szabo@techbook.com (Nick Szabo) Subject: ECON/SPACE: Monopolizing the Solar System Kryz writes: > countries with greatest colonial expansion, such as Spain and Portugal > in sixteenth and seventeenth century, or Great Britain and Russia in > eighteenth and nineteenth century, went into quite an economic downturn > following the expansions. Granted, but why should we care what happens to the state? What survived these decisions were the memes and genes of the individuals involved. Ignoring genes for the moment, let's look at the memes: vastly more people today speak English, Spanish, Russian and Portugese than speak the languages of all the other European cultures combined. For example Germany is more economically powerful than Britain today in and of itself; but English speakers are vastly more numerous, and have vastly more economic influence across the planet than German speakers, because the latter culture took more opportunies to expand (and had more such opportunties as an accident of geography, of course). Nick Szabo szabo@techbook.com ------------------------------ Date: Sunday, 30 May 1993 10:10:06 PST8 From: "James A. Donald" Subject: New Brain Test: "Einstein einstein.... Penrose!" In <9305301519.AA21036@wookumz.gnu.ai.mit.edu>, jamesdon@infoserv.com (James A. Donald) wrote: > > Michael Price deliberately misrepresented my post. Oops. My previous post may have been unclear, so that Price misunderstood it rather than misrepresented it. I have been assuming that people reading this thread have some familiarity with the problem that AI encountered in our unsuccessful attempt to emulate perception. This is obviously not the case, since the NN people, when the they run into this same problem, keep using new and obscure neologisms for it, instead of using the old and obscure neologisms. The problem is that in order to deal with the fluid and irregular variations encountered in the real world it is necessary to recognize collections of low level data as instances of categories - for example pixels as objects, frequency distributions as sounds, and so forth. Typical real world categories have a large number of internal degrees of freedom. Top down recognition involves matching all possible objects in all possible arrangements against the data to recognized. Bottom up recognition does not work, for reasons illustrated by: T Kanade Artificial Intelligence 17 409-460 (81) (negative chair experiment) (One of the few lucid papers of that era.) The whole, the thing to be recognized, is not merely a collection of independent parts. Rather the parts only make sense in the context provided by the whole, hence bottom up fails. Both ends towards the middle, is in practice, simply top down with good feature extraction. The feature extraction that animals perform seems to be no different than the feature extraction that AI performed. Obviously top down perception is impossible, because of limits of computational power in conventional AI, and limits on the size of the "neural net" in NN. Obviously bottom up perception does not work. Conclusion: Perception is impossible. But obviously it is possible. Not only is it possible, but it is possible to find particular neurons that are plainly doing the impossible, and stick electrodes into them. This is why artificial awareness is not an engineering problem, - not a question of insufficient power - but a problem of fundamental research - what the hell is going on here? --------------------------------------------------------------------- | We have the right to defend ourselves and our James A. Donald | property, because of the kind of animals that we | are. True law derives from this right, not from jamesdon@infoserv.com | the arbitrary power of the omnipotent state. ------------------------------ Date: Sun, 30 May 93 11:26 EDT From: kwaldman@tanstaafl.extropy1.sai.com (Karl M Waldman) Subject: Response to last message Cold Fusion Not! Ok, Ok so may I don't come up with the best hourly posts on flying yogi's, Chaitin or how to unsubscribe from the list. But to have no feedback to that last one is amazing (perhaps I've made everyone's kill list) As one of the original subscribers I've seen list volume go up up up and down and topics vary, HOWEVER I have never seen you ignore such a potential area. I figured I'd get at least, the very least is BULLSHIT, or some other doubt. I was hopeing some of the people in physics would poke into sci.physics.fusion and watch the fireworks (and comment on that). Tune in, Dr. Farrell just posted a awesome (not slang, look it up) scientific flame saying that "Schrodinger and Heisenberg [sp] and the Copenhangen interpertation were wrong!" [I'll post the message when I get to work on Tuesday] Bobdamn it, this guy says he has a cell that has been working for over a fucking year, and I know someone of credibility who has seen it. Look I'm still doubtful, until I see it I won't believe it (I'm working on that) but to have you miss it because of discussions on how to remove your self from the list is unbelievable. Have a nice day, Karl ------------------------------ Date: Sun, 30 May 1993 12:14:10 -0500 (EST) From: KMOSTA01@ULKYVX.LOUISVILLE.EDU Subject: ECON/SPACE: Monopolizing the Solar System Tim Starr claims that Spanish and Portugese States benefited greatly from exploitation of colonies. I must respectfully disagree. Interestingly enough countries without any colonies, such as Sweden, Norway, and, to a degree Germany (after losing wars and losing colonies), enjoyed better economic outcomes than imperial powers. I have no claim of know-all here, but countries with greatest colonial expansion, such as Spain and Portugal in sixteenth and seventeenth century, or Great Britain and Russia in eighteenth and nineteenth century, went into quite an economic downturn following the expansions. Even more so ... the downturns were prolongued, not just some temporary things like the Great Depression. This relates to the main discoveries of Ayn Rand, IMHO her claim to greatness (to annoy Mike Price, although no annoyance intended), which are: * my selfishness benefits all parties transacting with me; * selfishness is good (this will annoy both Democrats and Rush Limbaugh, gosh, what a life ...) even more so * selfishness is noble. Now a question -- as a balance to the celebration of Columbus' discovery, which multiculturalists so despise, shouldn't we celebrate the heroic effort of Chief Lapu-Lapu, who defeated and killed Ferdinand Magellan in the Phillipines? Krzys' ------------------------------ Date: Sun, 30 May 93 12:51:07 MDT From: Mark_Muhlestein@Novell.COM (Mark Muhlestein) Subject: Response to last message Cold Fusion Not! > [ Karl complains about lack of response to his "cold fusion" post ] OK. I am only a physics dabbler, but living in Utah, I have been following the cold fusion saga perhaps closer than most. There has been a lot of interesting discussion on the Usenet group sci.physics.fusion regarding Mills' and Farrell's work, but not much detail yet. Most observers are dismissing Farrell's theory (basically he posits fractional sub-ground state atomic orbitals), but people are withholding judgement on their excess heat claims until those results are fully disclosed. Farrell said they would soon give the details on their definitive experiment. Steve Jones at BYU has been able to partially replicate M&F's results, and he claims everything can be explained by H2 O2 recombination in the cell. (Yes, M&F's cell uses *light* water.) Pons and Fleischmann are still going strong in their Japanese-funded lab in France. They recently published a paper in Phys. Rev. Lett. A which is causing quite a stir. It appears that from the results that P&F are claiming, the choices are becoming quite stark: either they are seeing greater power density in their experiments than in nuclear fission reactor rods, or they are perpetrating an enormous hoax. Gene Mallove, the MIT researcher who wrote the pro-cold fusion book _From Fire to Ice_ posted recently a claim that within a few months dramatic, unrefutable evidence of cold fusion would be presented. We'll see. I would of course be thrilled if a cheap, plentiful source of energy could be found, but so far as I have seen, there has not been a rock solid result yet. I would be convinced by an isolated cold fusion cell which generates its own energy for electrolysis, plus runs a light bulb for longer than any battery of equivalent mass. Supposedly, Pons and Fleischmann have scaled up their experiments to produce something like this, so things are still interesting. If there is enough interest from extropians list readers who are interested in cold fusion but who can't read Usenet, I would be willing to post occasional notes on current events in that arena. Let me know. Mark_Muhlestein@novell.com ------------------------------ Date: Sun, 30 May 93 18:57:12 GMT From: price@price.demon.co.uk (Michael Clive Price) Subject: TECH: Cold Fusion Not! Karl Waldman asks for feedback on Cold Fusion Sorry Karl, I don't know anything about the specific experiments you mention. I do know I feel once bitten, twice shy on the subject and can give my overall impressions. When I first heard of it I thought: Great! Since then it's been disappointing. I don't know if there is anything to cold fusion, but I do know that it's going to be very hard / impossible to exploit. If the effect is so small that there's no hard, generally accepted evidence after, what, 3-4 years then it bodes ill for commercialisation of the process in the near future. Pons or Fleischman (sp?) still are publishing equivocal data. No one can agree on the presence of He3, neutrons or whatever. Is the source of putative energy chemial or nuclear? What a mess. I don't dismiss the idea out of hand as some people do, but I need clear evidence before I can think anything. Hence my silence. It sounds as if you know a lot about one experiment, tell us more. I look forward to the Dr Farrell piece (I don't/can't easily subscribe to sci.physics). Mike Price price@price.demon.co.uk ------------------------------ Date: Sun, 30 May 93 13:11:31 -0700 From: tcmay@netcom.com (Timothy C. May) Subject: Response to last message Cold Fusion Not! Karl Waldman asks why none of us have commented on the latest cold fusion theories/reports/flames going on over in sci.physics.fusion. I can think of several good reasons not to comment. At least good reasons for me not to comment: 1. Having just spent the last hour and a half skimming through hundreds of articles in sci.physics.fusion (thanks to the prodding by Karl), it seems to me that nothing especially noteworthy is going on. The same old stuff: complaints about Pons and Fleischman's latest inscrutable paper (who their coworkers were, what the apparatus was, bad technique, etc. The Farrell work Karl cited is getting some strong reactions, especially from Steve Jones, who lampoons the crankish work about hydrogen atoms being LRC circuits. (I'm not taking sides here, just reporting that there is by no means a new consensus that we Extropians ought to be responding to.) 2. As someone in that newsgroup noted, reading a few published papers is vastly more informative than reading a year's worth of idle speculations from casual participants in sci.physics.fusion. The same applies here to Extropians. To wit, what do any of us have to add to the debate, except our "gut reactions"? To put it bluntly, would the List be happier if I, or Perry M., or Mike P., or anybody else, read a few of the postings and then pronounced "Sounds like urban legend to me."? We already have far too much hip-shooting here. 3. For someone to really contribute to the S/N on this, they need to be reasonably active in fusion or CNF (cold nuclear fusion) work, or at least to be reading the actual papers. The Net is just a jumping-off point. If someone out there actually has the time and the interest to read the literature, I'd enjoy reading their views on the status of CNF. As it is, there are several books on the CNF controversy which I still haven't found the time to read, so I sure as hell can't add anything to the debate except second-hand reactions to the noise in sci.physics.fusion. (From the silence surrounding Karl's original post, I'd guess others feel the same way. On cold fusion, there needs to be some fairly *compelling* evidence--and more than just a few articles in a newsgroup--before the interest in cold fusion will pick up significantly. Crying wolf, and all that.) So that's my admittedly peripheral outlook on the latest CNF "results." If anybody has really strong reason to believe there's something to these latest results (beyond crank theories about hydrogen atoms, circuits, the Uncertainty Principle, the "ash" of cold matter, and so on), please tell us about it. -Tim May, still a skeptic -- 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 Note: I put time and money into writing this posting. I hope you enjoy it. ------------------------------ Date: Sun, 30 May 93 21:00:00 GMT From: price@price.demon.co.uk (Michael Clive Price) Subject: New Brain Test James A. Donald says: > As I said [..] the scaling variable is .. The original claim: >>> Since the amount of processing power required goes up >>> nonpolynomially fast with the scaling variable, (in the >>> case of "neural nets" the size of the neural net") is semantically identical to: the amount of processing power required goes up NP fast with the size of the neural net So the definition of the "scaling variable" is irrelevant. > Before category recognition neurons were found, NN people claimed > that there could not be such a thing Category recognition neurons / hidden units are a prediction of NNs. Face recognition cells were conjectured before they were discovered. > The required size of the "neural net" grows nonpolynomially fast, in > practice exponentially fast, with the size of the scaling variable. > Thus enormous increases in the size of the "neural net" will yield > only minor and trivial improvements in its flexibility. When a NN is too large it switches from categorical learning into rote learning. The required size is before rote learning takes over, so that it's forced to meaningfully compress its microworld model. What is remarkable is the complexity of the microworld than can be processed by simple NNs. > The problem is [...] a great many neurons whose output represents a > wide variety of high level categories, for example "face of adult > male", or the face of a particular individual. NN theory assumes that > high level categories are not generated directly, [..] Where are the > intermediate level abstractions? Absence of evidence is not evidence of absence. > What are the intermediate level abstractions? I refer you Hal's post about non-intuitive intermediate hidden units. > James A. Donald Mike Price price@price.demon.co.uk ------------------------------ Date: Fri, 28 May 93 23:43:07 GMT From: whitaker@eternity.demon.co.uk (Russell Earl Whitaker) Subject: Forwarded article. This article was forwarded to you by whitaker@eternity.demon.co.uk (Russell Earl Whitaker): --------------------------------- cut here ----------------------------- Path: eternity.demon.co.uk!demon!zaphod.axion.bt.co.uk!uknet!pipex!uunet! europa.eng.gtefsd.com!howland.reston.ans.net!usenet.ins.cwru.edu! cleveland.Freenet.Edu!de786 From: de786@cleveland.Freenet.Edu (Charles P. Zeps) Newsgroups: alt.privacy Subject: Whole Earth Review Crypto Article Date: 28 May 1993 21:18:37 GMT Organization: Case Western Reserve University, Cleveland, Ohio (USA) Lines: 5 Message-ID: <1u5vjd$lld@usenet.INS.CWRU.Edu> NNTP-Posting-Host: hela.ins.cwru.edu I just got the 25th anniversary issue of the WER in the mail and it has a 20 pg. excerpt from Kevin Kellys' 1994 (Addison-Wessely) yocoming book,tentatively titled "Out of Control".The article is called "Cypherpunks,E-Money and the Technologies of Disconnection",pages 40-59,WER No.79 Summer 1993. --------------------------------- cut here ----------------------------- ------------------------------ Date: Sun, 30 May 93 14:54:40 MDT From: Stanton McCandlish Subject: ECON/SPACE: The Loonie Treaty -- irrelevant? Quoth Dani Eder, verily I saith unto thee: > > In reply to Robin Hanson: > > Governments abandoned their claims to the Moon in 1967 when they > signed the Outer Space Treaty, wherein it says that the Moon > (and other celestial bodies) are not subject to national appropriation. I've been thinking this, all the time this ECON/SPACE discussion has been going on, but so far all I can really conclude is that in the face of profit, that Treaty will be cast to the winds. Either that, or it will be irrelevant. What I mean by that is that, if sci-fi (and serious socio-politico-economic) predictions are on the mark, govt. itself will be irrelvant before long. Corporations will have taken the place of govts., and well, govt. treaties are likely to be ignored by corporations, as things that do not apply. "People" (if you can call a slimemold-thing like a corporation "people") will quite cheerful stake claims and defend them with all the arms and armour they can muster when space becomes profitable. [Nostradamus mode off] any comments? -- When marriage is outlawed only outlaws will be inlaws! Stanton McCandlish, SysOp: Noise in the Void DataCenter Library BBS Internet anton@hydra.unm.edu IndraNet: 369:1/1 FidoNet: 1:301/2 Snail: 1811-B Coal Pl. SE, Albuquerque, New Mexico 87108 USA Data phone: +1-505-246-8515 (24hr, 1200-14400 v32bis, N-8-1) Vox phone: +1-505-247-3402 (bps rate varies, depends on if you woke me up...:) ------------------------------ Date: Sun, 30 May 93 16:23:29 MDT From: Stanton McCandlish Subject: VR: [perceived] motion sickness Quoth Russell E. Whitaker, verily I saith unto thee: > 2.) Legend Quest > A D&D (Dungeons & Dragons) style fantasy role playing > VR. Interestingly, like traditional D&D (gads, am I old > enough to call it "traditional"? ;-), player characters Heh, I'd say YOUNG enough. Most of the "fogies" think of all such games as "newfangled", and most of the kiddos think of D&D as "antiquated" (that's a direct quote from a RPG'er I was chatting with last nite). Maybe that makes you feel better. :) ANYWAY: > 3.) Harrier Jump Jet > This one is what it sounds like: a VTOL fighter > simulator. It has a distressing effect on me: simulator > sickness. I say distressing, as I have no problem > flying in the worst turbulence in a light aircraft: I'm > a private pilot. Like a lot of pilots, though, the > mismatch between proprioception (kinesthetic sense, > etc.) and visual perception makes me slightly nauseous. > Of the "Virtuality (TM)" games listed here, this is the > least satisfying. Glad this was brought up! I get this nausea playing other "VR" games (to stretch the term a lot). Anything with really good 3-D (i.e. Battle Zone doesn't qualify!) tends to make me want to do the technicolor yawn; Castle Wolfenstein 3D in particular has this nasty effect. I can play about 20 min- utes, and then I have to go lie down and groan else I feel like hell. I do notice that after going thru this cycle several times, I can then play for many hours straight (usually after I get more intensely involved). I was thinking, watching the "VR game" sequence in _Lawmower_Man_, that the partici- pants would have gotten violently ill in all likelihood. I never had considered the idea that it was the eye-perceived motion coupled with a LACK of body-felt motion that would cause this nausea. However this does seem to fit. In general I don't get sea-sick, especially if I can see the water, and I don't get motion sickness in cars, etc. Nor did I feel ill when playing a custom version of the old Star Wars arcade game that was mounted on a gyro-sort-of-thing (you played it sitting down, and the whole apparatus moved very realistically, though jerkily, as you moved the joystick; quite a blast). Anyway, I'd be interested in responses to the following, just to compare notes, from anyone: 1) Do you get similar responses to "VR"? 2) Does the nausea go away after more "immersion" (either better equipment that adds the motion in, or more intense involvement in the game) or more exposure? 3) If so, is the tolerance lasting, does it only apply to one game/experience, does it fade away totally after [x amount of time], etc? 4) Are VR designers taking this effect into account and trying to compensate for it adequately, by adding motion that matches the perception of motion? 5) Do you think there is a certain "strain" of people more fit for "virtual survival" as it were, people who do NOT get this effect? If so do you think it is a "mind" property or a "meat" property? A combination? Or is tolerance strictly learned? 6) Do you think such a matter could be of importance to future generations? Will REAL survival depend on such a trait? [scenario: the year is 2134, and most people work, play, etc in cyberspace. Those that cannot hack it, and toss their soylent green, or have other negative effects, such as vertigo, soon lose their jobs, and are homeless and starve, etc etc.] 7) Do you think the nausea has more to do with "meat" perceptions, or mentality? Is it just a matter of suspending disbelief, mind over body, etc? That's all I can think of at the moment. > Oh, about *Black Ice*: it's a British, *Wired* style magazine. > I'm still waiting for issue 2 to hit the stands sometime soon. > The contact address is: Would you happen to know the sub price in YankeeCash? (surface and air)? I'd like to get this mag ASAP. -- When outlawing is outlawed only outlaws will be lawmakers! Stanton McCandlish, SysOp: Noise in the Void DataCenter Library BBS Internet anton@hydra.unm.edu IndraNet: 369:1/1 FidoNet: 1:301/2 Snail: 1811-B Coal Pl. SE, Albuquerque, New Mexico 87108 USA Data phone: +1-505-246-8515 (24hr, 1200-14400 v32bis, N-8-1) Vox phone: +1-505-247-3402 (bps rate varies, depends on if you woke me up...:) ------------------------------ End of Extropians Digest V93 Issue #0290 ****************************************