From extropians-request@gnu.ai.mit.edu Mon Mar 15 13:00:47 1993 Return-Path: Received: from usc.edu by chaph.usc.edu (4.1/SMI-4.1+ucs-3.0) id AA17887; Mon, 15 Mar 93 13:00:41 PST Errors-To: Extropians-Request@gnu.ai.mit.edu Received: from churchy.gnu.ai.mit.edu by usc.edu (4.1/SMI-3.0DEV3-USC+3.1) id AA28662; Mon, 15 Mar 93 13:00:25 PST Errors-To: Extropians-Request@gnu.ai.mit.edu Received: by churchy.gnu.ai.mit.edu (5.65/4.0) id ; Mon, 15 Mar 93 15:38:16 -0500 Message-Id: <9303152038.AA12587@churchy.gnu.ai.mit.edu> To: ExI-Daily@gnu.ai.mit.edu Date: Mon, 15 Mar 93 15:37:51 -0500 X-Original-Message-Id: <9303152037.AA12577@churchy.gnu.ai.mit.edu> X-Original-To: Extropians@gnu.ai.mit.edu From: Extropians-Request@gnu.ai.mit.edu Subject: Extropians Digest V93 #0133 X-Extropian-Date: Remailed on March 15, 373 P.N.O. [20:38:15 UTC] Reply-To: Extropians@gnu.ai.mit.edu Errors-To: Extropians-Request@gnu.ai.mit.edu Status: OR Extropians Digest Mon, 15 Mar 93 Volume 93 : Issue 0133 Today's Topics: BRAIN: Creatures of infinite habit [1 msgs] Culture: Ex-names [1 msgs] Culture: My Extropian Name [1 msgs] ECON/LAW: Tragedy of the commons [3 msgs] Extropians at Cypherpunks Meeting [1 msgs] Info: Market Research on PC Hardware Purchase [1 msgs] META: Extropianism lost in the noise [3 msgs] Meta: BBS Uploading (was: BBS Announcement!!!!) [1 msgs] NANO/AI/GP/TECH: Nanosystems, etc. [2 msgs] PAGAN: Witchcraft Trial [1 msgs] PPL: The Assassination Bureau Ltd [1 msgs] SOC: Taboos [2 msgs] frozen assets: [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: 50834 bytes. ---------------------------------------------------------------------- Date: Mon, 15 Mar 1993 02:00:48 -0800 (PST) From: szabo@techbook.com (Nick Szabo) Subject: NANO/AI/GP/TECH: Nanosystems, etc. [A glossary is included below] In a quite good article, Derek Zahn postulates and explores running GP over molecular subparts to generate nanomachine designs. Paul BaClace on this list has looked into GA for protein folding, and a bunch of other folks around the world are working on that now. I've looked a bit into a similar idea of using GAs over a list of parts for mills, robots, 3D printers, etc. to design self-replicating macrofactories. Each design is [read: would be] represented by a flow-graph of parts and their associated processes. Quite similar to the "autocatalytic network" used by Bagley, Farmer, Kaufmann et. al. to model pre-genetic, self-perpetuating metabolisms, but I must also spatially compose the assemblies. I must also develop a method for composing new parts designs; one of the main design strategies for self-rep is to trade off performance for standardization & simplification of parts. One fitness function for self-rep is the minimization of the complexity or cost of outside parts used, ie the closure of the system; another is the flexibility of the resulting factory (it should have some capacity to generate real products as well as itself), a third is replication time. It is intriguing to think of implementing GP in Autocad LISP, getting a bunch of parts libraries on gigabyte hard disks (CD-ROM too slow), and letting this puppy crank for a few (days? months? years?). GP builds reasonably good solutions quickly, then spends the bulk of its time optimizing them. For this search space it is quite possible that good designs will be found quickly, since there has been no serious attempt by engineers to explore this space (at least none in the recorded literature, and no the NASA study wasn't even close to being a serious design). Implementing GP in Autocad LISP may be a productive task in its own right, even if used to tackle less ambitious tasks. Any LISP or Autocad hackers looking for a job at a startup? Some things to watch out for, both specifically for Derek's nanopart design problem and for design with GP/GA in general: * One problem sometimes run into with GP/GA is non-composable or "deceptive" solutions: subparts contributing to greater fitness in one combination do not increase their likelihood of contributing to fitness in other combinations. GP/GA works very poorly for this small but important minority of design problems. * The fitness function has to judge whether one imperfect design is better than another imperfect design. In some problems it is difficult or impossible to come up with such a function. * How to represent the specifications (functional? positional? physical?) They must be represented so that the "distance" of a design from the specs can be evaluated. * The bottleneck for GP is usually the fitness function, which must simulate the interaction between the subparts. Fitness functions must usually be run 10^5-10^6 times to yield a satisfying result, so it must complete in seconds, not minutes or hours common for molecular simulations. Here are some strategies for speeding up the fitness function in these kinds of design problems (much of this matches Drexler's advice in appendix A of _Nanosystems_ for human theoretical engineers): -- Occam's Razor: Sacrificing one or two orders of magnitude in design performance may be OK if it makes the design simpler, decreasing the GP run time by orders of magnitude. -- Greatly speed computation with compositional grammars over a small number of parts with dominance of well-known short-range interactions. Nucleic acids, amino acids etc. are often not simple and there are many complicating long-range interactions. Make design assumptions that eliminate compute-costly interactions, eg cross-links simplify the protein-folding problem. -- Since molecular simulations take so long, we need a molecular cellular automaton that embodies a conservative, efficient abstraction of molecular simulation models, perhaps by modeling short-range interactions and forbidding configurations with significant long-range interactions. -- When simplifying simulations highest priority is to eliminate false positives; second priority is to minimize false negatives. -- Satisficing: we are looking for orders-of-magnitude improvement, not small percentages. Run GP only at its early steepest slope and be satisfied with significantly suboptimal design. * To represent reactions we need a library of eutactic processes. In addition to representing the finished design of the subpart, the GP might also represent the tools needed to make the subpart. Hopefully all these tools will already be in the tools library, but if not the GP might construct new tools out of the parts library. This gives us a viable path to the design, which is as important as the design itself. * The evolved designs usually resemble the worst spaghetti code. We need design beautifiers to remove redundancies, pretty-print, etc. for human analysis and inclusion in libraries for further runs. A final idea to throw into the ring: Von Neumann original designed a 2D cellular automaton both Turing-equivalent and self-replicating, and someone whose name escapes me was able to make a simpler one. (Not Langston; his isn't Turing-equivalent). GA or GP should be able to search this space quite quickly on a modern supercomputer. As with the self-replicating factory it is more likely to find better solutions since this is a difficult space little explored by scientists. Once we've found a diverse set of satisfactory self-rep automata in the simple 2D cellular world, we can start search 3D worlds, and after that start adding molecule-like interactions, bringing the cellular world closer to molecular machine reality. Good ideas from simpler self-replicating automaton can be seeded into the GP developing self-rep for more complex cellular worlds. While Derek Zahn's system designs down building more complex parts in a realistic simulation, this system designs up, making self-replicating automata in ever more realistic cellular worlds. Eventually the two paths meet in a design for a self-replicating nanomachine in a realistic molecular simulation: a Drexler assembler. Glossary -------- cellular automaton: set of state machines in contiguous cells in cellular world GP: Genetic Programming; composes solutions by swapping parse subtrees GA: Genetic Algorithm; composes solutions by swapping substrings fitness function: in GA or GP, determines which solutions composed by recombination or mutation participate in forming the next generation, and which are eliminated. Combines decoding, simulation & evaluation of the solution with a statistical selection bias. function set: as used here, simulated processes, reactions, etc., represented as LISP functions terminal set: as used here, compositional elements: parts, molecules, etc. represented as LISP terminals Nick Szabo szabo@techbook.com ------------------------------ Date: Mon, 15 Mar 1993 07:15:24 -0500 (EST) From: Harry Shapiro Subject: Meta: BBS Uploading (was: BBS Announcement!!!!) a conscious being, Andrew S Hall wrote: > I was also wondering what the Extropian policy would be on my uploading > various items from the group and the essay list. > I planned: > 1) Nothing would be uploaded w/o express consent of the author > 2) Everything would have full attribution to the author and the > Extropians. Re, list rules, the only thing that matter is that the Author gives permission - Note however, if the main Author, let's say "Tim May" has in his post a quote from Perry, you should, for example, also get Perry's permission. [I present this as an example as Tim always allows his "material" to be reposted while Perry requests you get permission.] -- Harry Shapiro habs@panix.com List Administrator of the Extropy Institute Mailing List Private Communication for the Extropian Community since 1991 ------------------------------ Date: Mon, 15 Mar 1993 07:02:04 -0800 From: D Anton Sherwood Subject: Culture: Ex-names The Habsburg dynasty took its name from a county whose name was originally Habichtsburg, i.e. Hawk Fort. So "Habs Meme-Hawk" could be considered punny. Me, I like the name I've got, except that "Sherwood" is not easy to pronounce in other languages. What do you think of "Siruda"? By the way -- the Passport Office raised no eyebrow when I renewed and changed from "Anthony" to "Anton". Anton Sherwood dasher@well.sf.ca.us +1 415 267 0685 1800 Market St #207, San Francisco 94102 USA "Don't forget, your mind only *simulates* logic." -- Glen C. Perkins ------------------------------ Date: Mon, 15 Mar 93 10:41:58 EST From: Brian.Hawthorne@East.Sun.COM (Brian Holt Hawthorne - SunSelect Engineering) Subject: Culture: My Extropian Name Someone asked about methods for changing one's name. Amara Graps came back with the Court Petition method and the Usage method. There is one more, at least for last names in Massachusetts. When my wife (nee Carolyn Elizabeth Moore, no relationship to the Carol Moore on the list, although her family calls her Carol...) and I (nee Brian Robie Holt) got married, there were convenient spots on the "marriage license" for "groom's surname after marriage" and "bride's surname after marriage". I filled in "Holt Hawthorne", she filled in "Moore Hawthorne". This gave us a legal document to show to government folks to get our names changed (on passports, driver's licenses, etc.). Thus, as far as the USgov is concerned, I am 'Brian Robie Holt Hawthorne'. The space in the surname only confuses the IRS. The State Dept. and SSA have no problem with it. Of course, I usually just use Brian Hawthorne. And all of my friends call me Rowan... Rowan ------------------------------ Date: Mon, 15 Mar 93 11:12:02 -0500 From: "W. Scott Meeks" Subject: ECON/LAW: Tragedy of the commons >Date: Sun, 14 Mar 93 16:36:19 EST >From: drw@BOURBAKI.MIT.EDU > >Another aspect of this that I don't understand is that some >libertarians seem to be opposed to intellectual property, that is, >they want to prevent the development of a system of private property >that would stimulate the production of useful information! Dale, this is intellectually dishonest and inflamatory and you know it. Scott ------------------------------ Date: Mon, 15 Mar 93 11:22:24 -0500 From: "W. Scott Meeks" Subject: ECON/LAW: Tragedy of the commons My public apologies to Dale Worley. I accidently sent a private email intended to go only to him to the entire list. Rather than imposing an arbitrary punishment on myself (such as not posting for 24 hours) I will instead attempt to make restitution by repairing some of the damage to Dale's reputation. I have known Dale for several years and he is generally well read, intelligent, and a good critical thinker. Scott ------------------------------ Date: Mon, 15 Mar 1993 8:58:45 -0800 (PST) From: GRAPS@galileo.arc.nasa.gov Subject: PPL: The Assassination Bureau Ltd Nick Szabo and Perry Metzger refer to the Soldier of Fortune case: (Nick:) >Could the victim's relatives sue said Yellow Pages for damages? >_Soldier of Fortune_ was put of business by this phenomenon. I've been following this case, the below is an article from a couple of months ago on the latest developments. Soldier of Fortune is almost certainly out of business at this point, but I understand from a fellow at the local gun shop that the magazine will probably be resurrected under a new name, slightly different format etc. He told me that the magazine for the last few months has been conducting surveys of its readers trying to get a better idea of what their readers want. Amara Graps graps@gal.arc.nasa.gov ----------------------------------------------------------------------- >From San Jose Mercury News, dated Tuesday, Jan 12, 1993. In a setback for the publishing industry, the Supreme Court on Monday let stand a $4.3 million damage verdict against a magazine for printing a classified ad that led to the contract slaying of an Atlanta businessman. A U.S. appeals court ruled last year that Soldier of Fortune magazine can be held liable because its ad posed an "unreasonable risk of ... substantial harm to the public." Monday's action by the Supreme Court left that ruling intact. The decision is not only likely to put the 17-year-old magazine out of business, a company official said, but also puts publishers nationwide on notice that they must do more to screen the content of classified ads. Media lawyers said the case (Soldier of Fortune vs Braun No. 92-817) marked the first time that the high court had upheld a damage award against a publisher merely for printing an advertisement written by someone else. In 1985, an unemployed Vietnam vet named Michael Savage had sent in a classified ad to Soldier of Fortune titled "Gun for Hire." He described himself as a "37-year-old professional mercenary (who) is discreet and very private. Body guard, courier and other special skills. All jobs considered." His address and phone number followed. Savage later testified that he sought work as a security guard, but instead accepted an offer to kill Richard Braun. On Aug 26, 1985, Savage and another man shot and killed Braun in the driveway of his suburban Atlanta home and wounded his 16-year-old son. The killers were convicted and sentenced to prison, but the Braun family filed a wrongful death suit against Soldier of Fortune. A jury agreed that the magazine was liable and ultimately handed down the $4.3 million verdict, an amount the magazine's owners said would be "fatal" to its existence. Last year, a U.S. appeals court in Atlanta upheld the award. ------------------------------------------------------------------- ------------------------------ Date: Mon, 15 Mar 93 10:04:07 PST From: desilets@sj.ate.slb.com (Mark Desilets) Subject: META: Extropianism lost in the noise Regarding Anton's idea of a two-ended silencer, I like the idea, but think there needs to be safe-guards. It would be exceedingly easy for me to create an account, which, for the sake of civility, we shall call anti-harelb@sj.ate.slb.com Now, anti-harelb can "take out" harelb continuously. This is an undesirable thing, since someone is equally likely to (for revenge, say) create an anti-perry, or an anti-timmay or (heaven forfend) an anti-mark. Since we essentially have no verification procedure for who signs onto the list, we cannot implement this plan as proposed. Any thoughts? Aunty Mark ------------------------------ Date: Mon, 15 Mar 93 14:15:04 GMT From: jrk@information-systems.east-anglia.ac.uk (Richard Kennaway) Subject: SOC: Taboos Alex / Sasha writes: > Torture doesn't seem to be a problem to anybody; however, I would be >interested in exploring the difference between vandalism and torture >as lying in the sentience of the system that is subjected to deliberate >harm. A difference of degree, IMHO. I see vandalism as wrong not just because of the hurt it does to the owners and users of the vandalised property, but wrong even if one is vandalising one's own property. (This is a judgement according to my current values, not an assertion of absolute morality.) > Body-piercing, tattoing, etc. I have nothing against theoretically, >but most of the practitioners *I* observed have been pretty dirty types. Have you not noticed that a substantial proportion, perhaps a majority, of women in Western culture have pierced ears? Whether by "practitioners" you meant the people doing the piercing or the people with the piercings, I find it difficult to believe that most of them are "pretty dirty types". >I have no idea what/who PeeWee Herman is. You are fortunate. He has a show on American TV, which I haven't seen. It has spun off a film "Peewee's Big Adventure", which I have seen a clip from. I found it pretty repellent, though not for reasons having to do with societal taboos. Just my personal taste. What other people are repelled by about him is more likely to be the minor scandal he was involved in a year or two ago, concerning his presence in a gay cinema and indulging in the activities which customarily go on between the customers in such places. This (his behaviour, not the absurd scandal arising from it) is something which to me seems no more remarkable than an interest in football. What is the difference between a taboo and a personal dislike, other than the number of people sharing it? Is this thread concerned with notions which are taboos in society at large, or notions which may be taboo among the members of this list? Do homosexuality and body modification belong on David Lubkin's list (to pick out two subjects which are not at all taboo for me)? Considering the example of ear piercings, perhaps the difference between a taboo and a personal dislike is not that a taboo is widely shared, but that it is based on a widely shared stereotype rather than on an objective reality, and that the holders of the taboo have actually made no personal observation of the underlying reality -- even when face to face with it. Taboos in that sense are, IMHO, an impediment to personal development. If one abhors something, one should know why, and judge whether one's reasons are sufficient. Alluding to another thread, I understand that some magickal practices are designed to break down unthinking adherence to taboos that one has never seriously examined. Rational debate in a "safe space" such as this is another method of pursuing this goal. -- ____ Richard Kennaway \ _/__ School of Information Systems Internet: jrk@sys.uea.ac.uk \X / University of East Anglia uucp: ...mcsun!ukc!uea-sys!jrk \/ Norwich NR4 7TJ, U.K. Hobbitsfoot soup: A popular dish in Orcish and other civilised countries. Recipe on request. ------------------------------ Date: Mon, 15 Mar 93 12:52:16 CST From: twb3@midway.uchicago.edu (Tom Morrow) Subject: META: Extropianism lost in the noise >Regarding Anton's idea of a two-ended silencer, I like the idea, but think >there >needs to be safe-guards. It would be exceedingly easy for me to create an >account, which, for the sake of civility, we shall call >anti-harelb@sj.ate.slb.com Now, anti-harelb can "take out" harelb >continuously. > Aunty Mark Yes, this is a problem. I've proposed that we give people both ASCII arms and medicine. This would allow people to create accounts that continuously counteract the effects of others' anti-accounts. But it would also allow people to give themselves huge doses of virtual medicine, thus counteracting honest attempts to sanction them with ASCII arms. T.O. Morrow -- twb3@midway.uchicago.edu Law & Politics Editor: EXTROPY -- Journal of Transhumanist Thought Associate Executive Director: ExI -- The Extropy Institute ------------------------------ Date: Mon, 15 Mar 1993 13:53:26 -0500 (EST) From: Harry Shapiro Subject: Info: Market Research on PC Hardware Purchase I am looking names of computer hardware MARKET research firms. One's you have used and can recommend or have heard of as being good, etc. Please let me know ASAP. Call collect if you wish 212 941-4438 /harry -- Harry Shapiro habs@panix.com List Administrator of the Extropy Institute Mailing List Private Communication for the Extropian Community since 1991 ------------------------------ Date: Mon, 15 Mar 93 10:26:39 -0800 From: tcmay@netcom.com (Timothy C. May) Subject: ECON/LAW: Tragedy of the commons Sorry to butt in on a thread I've stayed out of, but this caught my eye: Scott Meeks writes: >>Date: Sun, 14 Mar 93 16:36:19 EST >>From: drw@BOURBAKI.MIT.EDU >> >>Another aspect of this that I don't understand is that some >>libertarians seem to be opposed to intellectual property, that is, >>they want to prevent the development of a system of private property >>that would stimulate the production of useful information! > >Dale, this is intellectually dishonest and inflamatory and you know it. > > Scott Well, I'm putatively a libertarian and I have argued on this list that intellectual property laws may be unenforceable, archaic, and wrong. Many libertarians I know are moving in this direction. So I find Dale's comments generally accurate. The last bit in his quote about "prevent the development....stimulate the production of useful information!" is, I think, a fair opinion for anyone worried about the "no intellectual property" faction of libertarianism. (I'm still not sure myself what the ramifications of less intellectual property will be.) -Tim May P.S. I just saw that Scott meant this as a private communication to Dale. Since this doesn't change the substance of what I just wrote, I'm sending this out anyway. -- 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: MailSafe and PGP available. ------------------------------ Date: Mon, 15 Mar 93 13:35:12 -0500 From: pavel@PARK.BU.EDU (Paul Cisek) Subject: BRAIN: Creatures of infinite habit Some interesting replies from Nick, Jay, and Perry... (I apologize for mentioning 'overpopulation' - how unextropian... ;-) Perry writes (about changing neural principles): >I find it very hard to believe that it will never be possible. Never >is a very long time, especially for immortals. Well put. However, a similar statement could be said for faster than light travel, which really may be impossible (attach appropriate disclaimer here). Development of the brain seems to proceed much like morphological development of an embryo, guided by gradients of tropic substances and lateral competition between homogeneous tissue (a theory). A brain is initially very undifferentiated and interconnected, neurons bloom during the start of a critical period and are pruned away by competition. The gradients provide biases to the competition. Nick writes about the use of nerve growth factor to return the brain to a plastic state. This certainly can be done, neurons can even sprout new axons and collaterals. However, after the glial cells and other guiding agents have been lost, the axon has no idea of where to connect to. Furthermore, since functional systems are already in place, it will have no chance in the competition for target tissue. Some studies have been done in guiding axons using tiny plastic tubes, but only on a small scale and only in the simplest of neural systems. Here is where nanotechnology will surely lend a hand... Above I talk about the biological brain, but similar issues will exist for an uploaded brain as long as it is a faithful copy. If the development process is crucial to the functionality of a neural-like system, then it will be a serious problem for the engineering of immortality. >We really know very little about the mind that we can be sure of at this >point other than a few basics like that it seems to be rooted in its >physical substrate. When we know enough to discuss it intelligently, >I'll start worrying. True, the mind is currently quite a mystery. Your analogy to Paleolithic blast furnace design is excellent - I've always been amused by discussions of consciousness when nobody really has an inkling of what it is. However, I enjoy such discussions. By the same argument all talk of uploading is largely speculative, yet still interesting and valuable. And speaking as someone who devotes a lot of time to study of the mind, let me say that the stability/plasticity dilemma is a real one. Development really does seem to proceed by pruning of connections and specialization of tissue. Jay mentions that older brains appear to have 'reduced synaptic complexity'. What I believe he refers to is the quantitative comparison between young and old brains - young brains have vastly more synapses. However, in this case, 'less' and 'more' do not map linearly to 'simple' and 'complex'. A 2-year-old's brain has more synapses than at any other age, but this does not mean that the 2-year-old is smarter than at age 20. It is the reduction of synaptic density that defines increased functionality as brain systems specialize and adapt to their environment. However, as pruning and neuron death continues into old age, functionality is reduced again. Again, this may not be a basic neural principle, but simply the inability of the body to stop its growth once the growth becomes detrimental. Evolution never selected out phenotypic effects that would be of benefit past reproductive age (at least such is Dawkins' theory on aging). Hopefully once we have a brain constructed out of more robust components over which we have more control, we'll be able to simply stop this deterioration. The problem only exists if the deterioration is a fundamental necessity. Nick writes (about creatures of habit): >This characteristic, along with lack of motivation & fit values, will >tend to be selected out in an upload environment. Yes, and if this is so then immortals will have to choose to either remain in a plastic state (accepting any potential side-effects) or continuously regress brain subsystems to more immature states. I think the only reasonable solution is what Nick mentions - to periodically refresh certain systems (motor systems often, cognitive systems less often, etc.). The brain is certainly extremely adaptive, perhaps enough so to deal with such transplantation. And when implemented in a non-biological substrate the issues of immunology are avoided. Finally, as Perry put it, someone who uploads before a theory of mind is developed can simply wait around until one is. This is similar to the cryogenic suspension scheme. Since I do believe that someday such a theory will be possible, I consider this an acceptable solution. (My remaining worry about the prospect of radically altering the brain I will withhold) -Paul ------------------------------ Date: Mon, 15 Mar 93 11:20:31 -0800 From: tcmay@netcom.com (Timothy C. May) Subject: PAGAN: Witchcraft Trial Just to let you folks know, I'm still reading the material Mr. Raymond forwarded. I tried to find Adler's "Drawing Down the Moon," but no bookstores had it. I did read about half of "Starhawk"'s "The Spiral Dance," and skimmed a dozen or so books on magick, Books of Shadows, how to cast spells, rune magic, astrology, voodoo, etc. I also tried to post a general call for help (for an admitted skeptic) to alt.pagan and alt.magick, which I've been reading fairly closely these past several days. I didn't mention the names of our list or Mr. Raymond's name. (He's welcome to, though.) Alas, some witchly force intervened and my posting went into the aether. (Actually, I think I screwed up when I told my news posting program--which is "tin" and is not Mac-friendly the way my newsreader, "Eudora," is--to post it to "alt.magick, alt.paganism". The little space between the comma and the next group is enough to blow the posting into irretrievable bits!) I think I can reconstruct the posting from a keystroke capture program I hope I left turned on. If there are any more comments on witchcraft and rationality, hasten them along. Judgment Day--the Vernal Equinox--is fast approaching. -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: MailSafe and PGP available. ------------------------------ Date: Mon, 15 Mar 93 11:20:23 -0800 From: tcmay@netcom.com (Timothy C. May) Subject: Extropians at Cypherpunks Meeting I just wanted to tell you folks that Extropians continue to be well-represented at the Cypherpunks meetings, including the highest-attendance meeting ever, this past weekend. Since the meeting followed the Computers, Freedom, and Privacy Conference, we has lots of new attendees. About 50 people, crammed into the open space at Cygnus Support, Mountain View, CA. Fortunately, people came and went, so only about 35 were present at any one time. Extropians that I can recall were: Dean Tribble, Keith Henson, Steve Witham, Sandy Sandfort, Dave Krieger, and me. Celebrity attendees included: Phil Zimmermann (author of PGP), Phiber Optik (Legion of Doom case), Whit Diffie (coinventor of public key crypto), and others. Topics: * Status of PGP. Zimmermann talked for a while about the difficulties of developing with many coauthors. Some sort of "CVS"-type code repository is likely to be adopted, possibly running on the "demon.uk" machine in London (to get it out of the U.S.). (At least one U.K. cypherpunk was there--linked to Russell Whitaker, I'm sure.) * A project to build a secure phone using only software to drive widely-available sound boards was described. This would thwart efforts to restrict crypto by making even phone conversations encrypted, (Sound Blaster board for PC clones, "CELP" standard, etc.) * FIDO Net work was described. * Dean Tribble talked about reputation filters and using them with mail readers. I brought up the Extropian interest in mail filters (the Cromell/Shapiro/Feynman proposals) and also the "genetic algorithm" newsreader filter written at Autodesk by former Extropians list member Paul Baclace. Is there any interest out there in a joint project with the Cypherpunks to develop such mail or news filter programs? * and the usual discussion of crypto privacy (Latest great quote: "In cypherspace, no one can read your screen." Patterned after "In space, no one can hear your scream," of course.) and digital money. The "Wired" magazine article with lots of stuff on "crypto anarchy" and digital privacy in general should be out late this month, I hear. And the rumor is still that yours truly will be on the cover (along with Eric Hughes and John Gilmore). A reporter for the "Village Voice," Julian Dibell, was at the meeting--and he'll be coming out here to Santa Cruz on Thursday to interview me for an article. Busy times. -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: MailSafe and PGP available. ------------------------------ Date: Mon, 15 Mar 93 13:14:14 CST From: derek@cs.wisc.edu (Derek Zahn) Subject: NANO/AI/GP/TECH: Nanosystems, etc. [A glossary is included below] -- Good Idea! Nick Szabo responds with very interesting character strings: > I've looked a bit into a similar idea of > using GAs over a list of parts for mills, robots, 3D printers, etc. > to design self-replicating macrofactories. That's interesting, especially because sometimes processes operate on parts to produce other parts, and sometimes they produce other processes! LISP can do that... I wonder if a (very strangely) typed lambda calculus might be an appropriate formalism to use for activities of this kind, since it can produce and operate on functions. > Quite similar to the "autocatalytic > network" used by Bagley, Farmer, Kaufmann et. al. Yes, catalytic networks are probably better for modelling the kinds of processes we're discussing, but they are more difficult to represent and operate on than LISP S-expressions. I do wonder if such networks can be transformed via graph grammars, and in fact whether catalytic network "recombination" operators could be defined as production rules in such a grammar. I have suspicions, based primarily on issues of smooth deformability and evolvability, that parallel networks will prove to be more appropriate information processing models for evolutionary development than procedural languages -- but that's a topic for another time (and my thesis proposal). > * One problem sometimes run into with GP/GA is non-composable > or "deceptive" solutions: subparts contributing to > greater fitness in one combination do not increase their > likelihood of contributing to fitness in other combinations. > GP/GA works very poorly for this small but important minority of design > problems. In fact, design problems may be more prone to deception than other types because of the way parts must work together. The fitness function can help if cleverly designed, but > * The fitness function has to judge whether one imperfect design > is better than another imperfect design. In some problems it > is difficult or impossible to come up with such a function. For this reason, I was kind of surprised that GP works at all! Still, most of the problems to which it has been applied are those that 1) have well-defined partial success definitions, and 2) those are partially successful in general because they contain some of the right pieces that can be used in better solutions. Clearly, a fitness function that measures progress toward the moon would look more favorably on a balloon than explosives, but that's what we call a "local maxima"! All we need to do is make sure that we cast our problems in terms that allow better fitness functions to be defined! [Other excellent points about practical difficulties of applying GP to design problems omitted because I don't know how to fix them yet] > -- Since molecular simulations take so long, we need a molecular > cellular automaton that embodies a conservative, efficient abstraction > of molecular simulation models, This is quite intriguing. In fact, just yesterday I was puzzling over an article called "Beyond the Turing Machine" by Brosl Hasslacher (from _The Universal Turing Machine: A Half-Century Survey_), in which Hasslacher claims that our usual computational paradigms are completely unsuited for such physical modeling, and that "skeletal CA games" are really the way to go. > * To represent reactions we need a library of eutactic processes. Yes, this and the related requirement that the catalog of reactions be COMPLETE with respect to the reagents we use are, I think, the most important obstacles to immediate productive implementation of the evolutionary approach to nano-component design. > the GP might construct new tools out of the parts ... > This gives us a viable path to the design, which is as > important as the design itself. Right, that was the whole point of using the GP paradigm to represent hierarchical construction processes. It highlights an important distinction: the difference between the requirements and the synthesis. The two-pronged approach that Nick submits for attacking this issue is very interesting, and in fact quite compelling: "designing down" -- searching for self-replication given a realistic environment "designing up" -- searching for realistic environment given self-replication Nick suggests that the two paths could meet. Here's how: We'll [well, *somebody* will!] define and redefine and reredefine a functional component-based interface that will be the actual goal of the "designing up" task -- something more akin to lego bricks or tinkertoys than chemicals. The nature of this functional interface will be based on (1) convenience for expressing the ideas that develop about self-replication and the *functional* motifs that comprise it, and (2) flexible molecular building block libraries that result from "designing down" toward more complex, robust parts. Note that Nick's scheme could be used, not just for self-replication, but for other abstract requirements -- e.g. computation. derek 2000 GLOSSARY lambda calculus: Notation for studying functions, created by Alonzo Church in the 1930's. Often used as a bridge between theory of mathematics and computation. Equivalent in computational power to the Turing Machine. catalytic network: often a chemical reaction needs a "catalyst" which is an agent that is required but is not transformed by the reaction. Dependence on catalysts can be expressed as a graph, which is called a catalytic network. An auto-catalytic reaction is one in which a product species catalyzes its own production. Processes in cell nuclei are prototypical autocatalytic reactions. Important because they increase strong nonlinearities into normally linear chemical processes. graph grammar: rules in grammars replace structural components with other structural components. Normal grammars operate on strings, which is convenient since ordering is the only important structural relation. However, when more general structural relationships exist between components, those relationships must be updated when one or more components are replaced by one or more others. Grammars with this capability are often referred to as "graph grammars". eutactic process: one in which precise molecular positions are determined. ------------------------------ Date: Mon, 15 Mar 93 11:33:38 -0800 From: tcmay@netcom.com (Timothy C. May) Subject: META: Extropianism lost in the noise >>Regarding Anton's idea of a two-ended silencer, I like the idea, but think >>>there >>needs to be safe-guards. It would be exceedingly easy for me to create an >>account, which, for the sake of civility, we shall call >>anti-harelb@sj.ate.slb.com Now, anti-harelb can "take out" harelb >>>continuously. >> >> Aunty Mark > >Yes, this is a problem. I've proposed that we give people both ASCII arms >and medicine. This would allow people to create accounts that continuously >counteract the effects of others' anti-accounts. But it would also allow >people to give themselves huge doses of virtual medicine, thus >counteracting honest attempts to sanction them with ASCII arms. > >T.O. Morrow -- twb3@midway.uchicago.edu Why not simply use psychic disruption, voodoo, and sympathetic magick to disrupt or block the postings of our opponents? The effectiveness of these methods has recently been proved beyond any doubt on this list, and various religious tracts contain ample historical proof. This way, undesirable posts could just be nu ------------------------------ Date: Mon, 15 Mar 93 11:33:33 -0800 From: tcmay@netcom.com (Timothy C. May) Subject: SOC: Taboos Richard Kennaway, commenting on Pee Wee Herman: >What other people are repelled by about him is more likely to be the minor >scandal he was involved in a year or two ago, concerning his presence in a >gay cinema and indulging in the activities which customarily go on between >the customers in such places. This (his behaviour, not the absurd scandal >arising from it) is something which to me seems no more remarkable than an >interest in football. Sorry, but the movie in question was "Nurse Nancy," starring the spectacularly endowed Sandra Scream. It contained no gay scenes, and Paul Rubin's alleged activities were reportedly of a solitary nature. (This info from the various news reports. And, just to be sure of the facts, I myself rented this film. :-} ) -Tim ------------------------------ Date: Mon, 15 Mar 93 14:34 EST From: kqb@whscad1.att.com Subject: frozen assets: Steve (sneal@muskwa.ucs.ualberta.ca) said: >>D. Anton Sherwood: >>> What about taking it with you -- that is, setting aside a trust fund for >>> when you wake up? >> >> Nick Szabo: >>There are apparently rules against perpetuities (except in Luxembourg?), >>but I wonder if I could insure my reanimation along these lines: > > It's Lichtenstein, not Luxembourg, that allows perpetual >trusts. The relevant part of Spitz's Tax_Havens_Encyclopaedia sez: > ... Steve, Thanks for the information on Lichtenstein law. FYI: Cryonics message number 0004 has the following note on this topic: The Reanimation Foundation is set up to enable you to "take it with you" and provide financial support for your reanimation, reeducation, and reentry. It is based in Liechtenstein, which does not have a Rule Against Perpetuities, and thus allows financial assets to be owned by a person long after the person is declared legally dead. Reanimation Foundation c/o Saul Kent 16280 Whispering Spur Riverside, CA 92504 (800) 841-LIFE I cannot claim detailed knowledge about this plan, as I haven't put up the money to get into this particular game. As far as "taking it with you" goes, though, it is the most organized attempt I have yet seen. (Most of us who are signed up for cryonic suspension are so much more concerned about whether or not we will get suspended-and-eventually-revived that making the next step, taking monetary assets with us, has not received much attention.) Kevin Q. Brown INTERNET kqb@whscad1.att.com or kevin_q_brown@att.com ------------------------------ End of Extropians Digest V93 Issue #0133 ****************************************