macleod@drivax.UUCP (MacLeod) (11/06/87)
In article <799@sbcs.sunysb.edu> dji@sbcs (David J Iannucci) writes: > Has anyone >else even read _EOC_? I can't believe I haven't seen any discussion of it >in this group. When I first found out about the net, I wrote a short message describing some details of nanotechnology, then a long magazine-length article about nanotechnology and Buckminster Fuller's concept of "ephemeralization". I posted these to sci.physics and sci.space and possibly one other group. There were alomst no responses. I was baffled, given the obvious intelligence of the net.readers, that nobody wanted to talk about the fantastic claims and possibilities that Drexler comes up with. Perhaps we've heard so much bad news that we cannot believe anything good can happen. Or perhaps it all sounds like some Omni-magazine pseudoscience fairytale. I don't know; I stopped trying to stir up interest. Every now and then somebody will bring up Drexler in some context, like the fellow in sci.space who mentioned that with nanotechnology we could send nanobuilders ahead of the colonization ships and have cities ready when the colonists arrive.
dji@sbcs (David J Iannucci) (11/11/87)
In article <2698@drivax.UUCP> macleod@drivax.UUCP (MacLeod) writes: >I posted these to sci.physics and sci.space and possibly one other group. >There were alomst no responses. I was baffled, given the obvious >intelligence of the net.readers, that nobody wanted to talk about the >fantastic claims and possibilities that Drexler comes up with. I have recently been pointed to comp.society.futures for the discussions on nanotechnology. There's not a lot there, but there's something. I guess there just aren't as many people as I thought who have read _EoC_, and certainly not many who want to tackle something so highly controversial on a scientific level. Understandable. The sociological aspects are easier to deal with at this point. Still, I've never known the net.readership to back off from a juicy argument before 8^) -- Dave "it's a physical example of infinity" Iannucci SUNY at Stony Brook UUCP: {allegra, philabs, pyramid, research}!sbcs!dji CSnet: dji@suny-sb.edu ARPA-Internet: dji%suny-sb@relay.cs.net ICBM: 40 55' N 73 08' W
glg@sfsup.UUCP (G.Gleason) (11/12/87)
In article <2698@drivax.UUCP> macleod@drivax.UUCP (MacLeod) writes: >In article <799@sbcs.sunysb.edu> dji@sbcs (David J Iannucci) writes: >> Has anyone >>else even read _EOC_? I can't believe I haven't seen any discussion of it >>in this group. >When I first found out about the net, I wrote a short message describing >some details of nanotechnology, then a long magazine-length article about >nanotechnology and Buckminster Fuller's concept of "ephemeralization". I don't know why there was not much response before, maybe not many people were familiar with his ideas yet. Interesting that you should bring up Fuller, since I just finished reading _Critical_Path_ (I'm still working on _EoC_). I see many parallels between what the two of them are saying. _EoC_ clearly extends some of Fuller's projections much further into the future, as it should since Drexler grew up in the world that Fuller projected when he was younger. I didn't see your articles, if it has been quite some time you might consider reposting. There have also been several articles on _EoC_ in comp.society.futures that you might want to check out. Gerry Gleason
peter@sugar.UUCP (Peter da Silva) (11/15/87)
Even Drexler is uneasy about the "Grey Goo". -- -- Peter da Silva `-_-' ...!hoptoad!academ!uhnix1!sugar!peter -- Disclaimer: These U aren't mere opinions... these are *values*.
kdmoen@watcgl.UUCP (11/19/87)
peter@sugar.UUCP (Peter da Silva) writes: >Even Drexler is uneasy about the "Grey Goo". Rather than being apprehensive about Grey Goo, I am looking forward to the time when I can design a suite of nanomachines that will rebuild my body from the inside out, enhancing my mental abilities by a factor of 10**9 and my physical abilities by a lesser factor. Once a sufficient number of people have done this, I'm sure we will have no problems finding ways to deal with Grey Goo. The point of this is that, if Drexlers predictions pan out, then just about anything becomes possible. I must confess that I am still a trifle skeptical about the feasibility of all of this. I would like to *see* a fully worked out design for Grey Goo or even for a submicron supercomputer that can be, if not empirically tested, then at least scrutinized by the scientific community and run through a computer simulation. If it *does* turn out to be possible to build Grey Goo, then by the time fabrication technology catches up, perhaps we can have a wide spectrum of Goo killing techniques already available. Does anybody know how close the Foresight Institute/MIT Nanotechnology Group is to completing the designs for interesting nanomachines? -- Doug Moen University of Waterloo Computer Graphics Lab UUCP: {ihnp4,watmath}!watcgl!kdmoen INTERNET: kdmoen@cgl.waterloo.edu
russ@m-net.UUCP (Russ Cage) (11/26/87)
In <2411@watcgl.waterloo.edu> kdmoen@watcgl.waterloo.edu (Doug Moen) writes: >peter@sugar.UUCP (Peter da Silva) writes: >>Even Drexler is uneasy about the "Grey Goo". >[...] If it *does* turn out to be possible to build Grey Goo, >then by the time fabrication technology catches up, perhaps we can have >a wide spectrum of Goo killing techniques already available. You have to find it first. The difficulty is that, in order to decide if a particular bit of nanomachinery is Gray Goo (or a part thereof), you have to analyze its program to see if it ever quits reproducing. This is exactly equivalent to the halting problem, which is insoluble. -- The above are the official opinions and figures of Robust Software, Inc. Will you come quietly, or must I use earplugs? Russ Cage, Robust Software Inc. (313) 662-4147 {sw1a7,umix}!m-net!rsi
pokey@well.UUCP (11/29/87)
In the referenced message, russ@m-net.UUCP (Russ Cage) wrote: }In <2411@watcgl.waterloo.edu> kdmoen@watcgl.waterloo.edu (Doug Moen) writes: }>[...] If it *does* turn out to be possible to build Grey Goo, }>then by the time fabrication technology catches up, perhaps we can have }>a wide spectrum of Goo killing techniques already available. } }You have to find it first. The difficulty is that, in order to decide }if a particular bit of nanomachinery is Gray Goo (or a part thereof), }you have to analyze its program to see if it ever quits reproducing. }This is exactly equivalent to the halting problem, which is insoluble. Congratulations, you've just proved that the human immune system is equivalent to solving the halting problem, and therefore impossible. --- Jef Jef Poskanzer jef@lbl-rtsg.arpa ...well!pokey "Greetings from the World's Biggest."
cipher@mmm.UUCP (Andre Guirard) (11/30/87)
In article <1445@m-net.UUCP> russ@m-net.UUCP (Russ Cage) writes: >In <2411@watcgl.waterloo.edu> kdmoen@watcgl.waterloo.edu (Doug Moen) writes: >>[...] If it *does* turn out to be possible to build Grey Goo, >>then by the time fabrication technology catches up, perhaps we can have >>a wide spectrum of Goo killing techniques already available. >You have to find it first. The difficulty is that, in order to decide >if a particular bit of nanomachinery is Gray Goo (or a part thereof), >you have to analyze its program to see if it ever quits reproducing. >This is exactly equivalent to the halting problem, which is insoluble. Fortunately solving the halting problem is not necessary to controlling Grey Goo. Better safe than sorry: we kill any nanomachinery that we can't prove will stop reproducing in a reasonable amount of time. After all, there is little practical difference between something that never stops reproducing and something that stops after 100 million years. -- o Andre Guirard < ' The race is not always to the swift... "Thai green beads" / > but it's a pretty safe bet. ihnp4!mmm!cipher '
lambert@cwi.nl (Lambert Meertens) (12/01/87)
In article <1445@m-net.UUCP> russ@m-net.UUCP (Russ Cage) writes:
) The difficulty is that, in order to decide
) if a particular bit of nanomachinery is Gray Goo (or a part thereof),
) you have to analyze its program to see if it ever quits reproducing.
) This is exactly equivalent to the halting problem, which is insoluble.
As long as the GG itself is not insoluble ... :-)
There are theoretical differences between the GG containment problem and
the TM (Turing Machine) halting problem. Suppose the nanoprogram has been
proven to stop in a finite time. Only, after the machinery is released, it
turns out that the stopping time is 10^79 generations. Not exactly what we
want, eh? On the other hand, it does not really matter if the stuff keeps
indefinitely "alive" as long as its growth is checked. I can imagine a
nanomachined salami sausage that will regrow if you cut slices off, but not
beyond one foot. (The problem whether a machine--together with its
offspring--will grow beyond a prior established space limit S is
algorithmically solvable in theory, but the time this takes is exponential
in S, so this is no big help.) Finally, one has to take the (real-time
physical) input stream to the machine into account, whereas for the TM case
the input is given and finite. Perhaps the growth of my salami sausage is
provably checked except if it is kept on a diet of pure caviar. That would
not worry me ... until someone markets caviar with checked growth except if
fed enough salami. Part of the physical environment is radiation, which
will cause random mutations. Most mutations may be indifferent or lethal,
but in the analysis we must consider the possibility of malign mutations.
Even theoretical analysis can only give probabilistic results here.
--
Lambert Meertens, CWI, Amsterdam; lambert@cwi.nlsethg@athena.mit.edu (Seth A. Gordon) (12/01/87)
In article <4560@well.UUCP> pokey@well.UUCP (Jef Poskanzer) writes: >In the referenced message, russ@m-net.UUCP (Russ Cage) wrote: >}You have to find it first. The difficulty is that, in order to decide >}if a particular bit of nanomachinery is Gray Goo (or a part thereof), >}you have to analyze its program to see if it ever quits reproducing. >}This is exactly equivalent to the halting problem, which is insoluble. > >Congratulations, you've just proved that the human immune system is >equivalent to solving the halting problem, and therefore impossible. >--- From what I know of both subjects (i.e., about what Hofstadter wrote in _Godel,_Escher,_Bach_ and _Metamagical_Themas_), a *PERFECT* immune system is equivalent to solving the halting problem, and therefore impossible. Our immune system can IMPERFECTLY control infections, just as there are ways to IMPERFECTLY recognize infinite loops. We could always try using a blowtorch. OK, so we wouldn't always succeed... >Jef > > Jef Poskanzer jef@lbl-rtsg.arpa ...well!pokey > "Greetings from the World's Biggest." Seth Gordon, 52 Massachusetts Avenue #102, Cambridge, MA 02139-4324, US of A. a.k.a. sethg@athena.mit.edu [Internet] |---------------------------------- a.k.a. sethg%athena.mit.edu@mit-eddie.uucp | By the time you read this message a.k.a. sethg%athena.mit.edu@mitvma.bitnet | it may not even be *my* opinion.
peter@sugar.UUCP (Peter da Silva) (12/03/87)
Russ Cage: > Doug Moen: > >Me: > >>Even Drexler is uneasy about the "Grey Goo". > >[...] If it *does* turn out to be possible to build Grey Goo... > You have to find it first.... you have to analyze its program to see > if it ever quits reproducing. > This is exactly equivalent to the halting problem, which is insoluble. Besides, Grey Goo is more likely to be born by accident, not made. If you have problems with alpha hits in microtechnology, what do you think radiation will do to nanotechnology? You're gonna knock off the control system for some gadget before it's even out of the lab. This stuff makes biowarfare research look like a Thomas Salter chemistry set. (You can't do anything REALLY dangerous with those kits. I know. I tried) -- -- Peter da Silva `-_-' ...!hoptoad!academ!uhnix1!sugar!peter -- Disclaimer: These U aren't mere opinions... these are *values*.
merkle@beatnix.UUCP (Ralph Merkle) (12/06/87)
>In <2411@watcgl.waterloo.edu> kdmoen@watcgl.waterloo.edu (Doug Moen) writes: >>peter@sugar.UUCP (Peter da Silva) writes: >>>Even Drexler is uneasy about the "Grey Goo". >You have to find it first. The difficulty is that, in order to decide >if a particular bit of nanomachinery is Gray Goo (or a part thereof), >you have to analyze its program to see if it ever quits reproducing. >This is exactly equivalent to the halting problem, which is insoluble. >Russ Cage, Robust Software Inc. (313) 662-4147 {sw1a7,umix}!m-net!rsi Nope. Just destroy anything that is not identifiable as 'friend'. That's the principal the immune system uses, and it seems to work okay most of the time. I should mention that Drexler has proposed more detailed designs of nano computers than the popularizations in EOC. The more detailed designs look quite plausible and look like variations on Babbage's Analytical Engine -- only with smaller gears, levers, etc. The key trick in simplifying the design problem is to (1) propose a few intuitively plausible building blocks -- rods that can be slid back and forth, knobs that stick out of the rods, etc. Then (2) propose a specific chemical structure for each building block, and demonstrate that -- within reasonable engineering tolerances -- it can be treated like the intuitively simple building block (example -- a 'rod' can be made out of a series of carbon atoms, alternately single and triple bonded). Finally (3) put the basic building blocks together into some interesting structure -- such as a computer. The major limitation today is the manufacturing technology. Given an atomic machine shop even I could start making interesting nano-machines using the 'building block' approach. Current molecular-scale fabrication technology is limited to the construction of arbitrary sequences of amino acids (protein engineering). Understanding how to build a protein is rather difficult right now, though a great deal of funding is being poured into research in this area. Building a working machine composed of hundreds or thousands of proteins is beyond current design technology -- though NOT current MANUFACTURING technology. If someone (God, ET's or whoever) gave us the specific amino acid sequences of the needed proteins, we could easily (this is the scientists 'easily', not the engineers 'easily') produce the proper proteins in the proper amounts. Assembly of the proteins is simple: stir them together and watch them self-assemble. This is a standard assembly technique used in existing protein-based machines (us!). Now it's just a small matter of programming :-) to build a molecular level machine shop from proteins. Ralph C. Merkle ...!sun!elxsi!merkle
peter@sugar.UUCP (Peter da Silva) (12/27/87)
In article <636@elxsi.UUCP>, merkle@beatnix.UUCP (Ralph Merkle) writes: > peter@sugar.UUCP (Peter da Silva) writes: > >Even Drexler is uneasy about the "Grey Goo". > > Just destroy anything that is not identifiable as 'friend'. That's > the principal the immune system uses, and it seems to work okay most of > the time. Great job. Your grey goo detector sounds just like the sort of thing to run wild and turn you into grey goo. These tiny little machines are going to be real susceptible to radiation... so what happens when a carbon-14 in your little sliding rod state machine decays and your nanomachine "immune system" suddenly forgets that your red blood cells are "friends". -- -- Peter da Silva `-_-' ...!hoptoad!academ!uhnix1!sugar!peter -- Disclaimer: These U aren't mere opinions... these are *values*.
beede@hubcap.UUCP (Mike Beede) (01/01/88)
in article <1315@sugar.UUCP>, peter@sugar.UUCP (Peter da Silva) says: } } In article <636@elxsi.UUCP>, merkle@beatnix.UUCP (Ralph Merkle) writes: }> peter@sugar.UUCP (Peter da Silva) writes: }> >Even Drexler is uneasy about the "Grey Goo". }> }> Just destroy anything that is not identifiable as 'friend'. That's }> the principal the immune system uses, and it seems to work okay most of }> the time. } } Great job. Your grey goo detector sounds just like the sort of thing to run } wild and turn you into grey goo. These tiny little machines are going to be } real susceptible to radiation... so what happens when a carbon-14 in your } little sliding rod state machine decays and your nanomachine "immune system" } suddenly forgets that your red blood cells are "friends". I suppose one solution would be for them to also destroy anything that was nearly like themselves, but different (e.g., if there are 100,000 components, then if 99,900 < N < 100,000 of them are identical, try to destroy the booger). If we have some estimates of mutation rate, we can derive a probability distribution for a harmful mutant's prospering. Surrounded by huge numbers of correctly-functioning detectors, it seems at first glance that the survival of a nasty strain could be made arbitrarily small. Of course, if the mutation involved just destroying everything, the ``good'' detectors might be too slow on the draw to do anything. Sounds like a Western... -- Mike Beede Computer Science Dept. UUCP: . . . !hubcap!beede Clemson University INET: beede@hubcap.clemson.edu Clemson SC 29634-1906 YOUR DIME: (803)656-{2845,3444}
pokey@well.UUCP (Jef Poskanzer) (01/02/88)
In the referenced message, peter@sugar.UUCP (Peter da Silva) wrote: } In article <636@elxsi.UUCP>, merkle@beatnix.UUCP (Ralph Merkle) writes: }> peter@sugar.UUCP (Peter da Silva) writes: }> >Even Drexler is uneasy about the "Grey Goo". }> }> Just destroy anything that is not identifiable as 'friend'. That's }> the principal the immune system uses, and it seems to work okay most of }> the time. } }Great job. Your grey goo detector sounds just like the sort of thing to run }wild and turn you into grey goo. These tiny little machines are going to be }real susceptible to radiation... so what happens when a carbon-14 in your }little sliding rod state machine decays and your nanomachine "immune system" }suddenly forgets that your red blood cells are "friends". Quoting from "Engines", page 178: ...systems can work even when their parts fail; the key is redundancy. Imagine a bridge suspended from cables that fail randomly, each breaking about once a year at an unpredictable time. If the bridge falls when a cable breaks, it will be too dangerous to use. Imagine, though, that a broken cable takes a day to fix (because skilled repair crews with spare cables are on call), and that, though it takes five cables to support the bridge, there are actually *six*. Now if one cable breaks, the bridge will still stand. By clearing traffic and then replacing the failed cable, the bridge operators can restore safety. To destroy this bridge, a second cable must break in the same day as the first. Supported by six cables, each having a one-in-365 daily chance of breaking, the bridge will likely last about ten years. While an improvement, this remains terrible. Yet a bridge with ten cables (five needed, five extra) will fail only if *six* cables break on the same day: the suspension system is likely to last over ten million years. With fifteen cables, the expected lifetime is over ten thousand times the age of the Earth. Redundancy can bring an exponential explosion of safety. So, what happens when a carbon-14 in the little sliding rod state machine decays is: the other 99 similar state machines performing the same calculation outvote it, notice the fault, and call in a repair machine. Peter: read the book. --- Jef Jef Poskanzer jef@lbl-rtsg.arpa ...well!pokey "The only way that Listerine kills germs on contact is if you take the bottle and smash them with it."
sewilco@datapg.DataPg.MN.ORG (Scot E. Wilcoxon) (01/03/88)
In article <827@hubcap.UUCP> beede@hubcap.UUCP (Mike Beede) writes: : >Surrounded by huge numbers of correctly-functioning detectors, it seems at >first glance that the survival of a nasty strain could be made arbitrarily >small. Of course, if the mutation involved just destroying everything, the >``good'' detectors might be too slow on the draw to do anything. Sounds like >a Western... If a mutation involves destroying "everything", they'll destroy themselves. If a mutation involves changing "everything" for replication (ie, rampant Grey Goo), the raw materials are probably wrong for replication and the mutation will in effect be consuming poison. Perhaps for a day, then the city cleanup crew will let it try to deal with lasers and heatlamps... -- Scot E. Wilcoxon sewilco@DataPg.MN.ORG ihnp4!meccts!datapg!sewilco Data Progress C and UNIX consulting +1 612-825-2607 "I'm the girl that makes the thing that drills the thing that makes the thing that oils the thingumabob that goes around and makes the engines roar."
merkle@beatnix.UUCP (Ralph Merkle) (01/05/88)
in article <1315@sugar.UUCP>, peter@sugar.UUCP (Peter da Silva) says: } } In article <636@elxsi.UUCP>, merkle@beatnix.UUCP (Ralph Merkle) writes: }> peter@sugar.UUCP (Peter da Silva) writes: }> >Even Drexler is uneasy about the "Grey Goo". }> }> Just destroy anything that is not identifiable as 'friend'. That's }> the principal the immune system uses, and it seems to work okay most of }> the time. } } Great job. Your grey goo detector sounds just like the sort of thing to run } wild and turn you into grey goo. These tiny little machines are going to be } real susceptible to radiation... so what happens when a carbon-14 in your } little sliding rod state machine decays and your nanomachine "immune system" } suddenly forgets that your red blood cells are "friends". The suggestion that you 'destroy any non-friend' was made in response to the erroneous statement that identifying grey goo was as hard as the halting problem. This is clearly not the case. It is not essential that you infallibly determine whether or not a particular nano-machine running a particular nano-program will or will not destroy the world before you destroy it. 'Beyond reasonable doubt' might be a better criteria. The simplest way of insuring that 'counter goo' doesn't get out of hand is to make counter goo NON replicating. A factory manufactures counter goo, the counter goo is shipped to the area under attack, and the counter goo does its thing. If the counter goo is equipped with 'tiny time pills' that make it disintegrate after some reasonable period of time, then it's even more rigidly limited. The suggestion that nano machines are subject to radiation damage is true, the inference that this implies it is impossible (or even difficult, once you are capable of manufacturing a nano-machine in the first place -)) to manufacture a nano machine that will self-destruct rather than run wild is not true. If each 'counter goo' machine had several self-destruct circuits, and each self destruct circuit would fire in the event of a detected fault, then the probability that the overall machine would continue to function following any fault can be made arbitrarily low. The detailed design is left as an exercise to the reader.....-) I am not overly concerned about accidental destruction of the planet -- I AM concerned about PLANNED destruction of large parts of the planet -- past history suggests several groups will make vigorous efforts to develop systems capable of this, and they might well USE them..... My concern becomes acute when said systems are pointed at ME -- so far, it hasn't happened -- and I'd like it to stay that way. Ralph C. Merkle ...!sun!elxsi!merkle
jlh@loral.UUCP (The Happy Hacker) (01/06/88)
I've been looking for this book since october-november, with no luck.
(Engines of Creation by Drexler, that is). Where can I get a copy?
Is it under science and technology, biology, or what? And finally,
is it softcover or hardcover? Any direct pointers to a San Diego
store would be appreciated.
Jim
--
Jim Harkins
Loral Instrumentation, San Diego
{ucbvax, ittvax!dcdwest, akgua, decvax, ihnp4}!sdcsvax!sdcc6!loral!jlhpeter@sugar.UUCP (Peter da Silva) (01/08/88)
In article <4879@well.UUCP>, pokey@well.UUCP (Jef Poskanzer) writes: > In the referenced message, peter@sugar.UUCP (Peter da Silva) wrote: > }real susceptible to radiation... so what happens when a carbon-14 in your > }little sliding rod state machine decays and your nanomachine "immune system" > }suddenly forgets that your red blood cells are "friends". > So, what happens when a carbon-14 in the little sliding rod state machine > decays is: the other 99 similar state machines performing the same > calculation outvote it, notice the fault, and call in a repair machine. I'd feel a lot safer if they self-destructed instead of trying to repair the damage. The machines have to be self-reproducing anyway. Even so... > Peter: read the book. I read a couple of his early papers one of his cohorts was handing out at an SF convention about 6 years ago. I seem to remember this argument... I'd like to see the system run for a few years in an isolated orbital station before I trusted them in my body. I'm probably not going to live so long, but I hope my children or grandchildren or great-grandchildren are just as cautious. -- -- Peter da Silva `-_-' ...!hoptoad!academ!uhnix1!sugar!peter -- Disclaimer: These U aren't mere opinions... these are *values*.
johng@ecrcvax.UUCP (John Gregor) (01/08/88)
In article <657@elxsi.UUCP> merkle@beatnix.UUCP (Ralph Merkle) writes: > > I am not overly concerned about accidental destruction of the planet -- >I AM concerned about PLANNED destruction of large parts of the planet -- >past history suggests several groups will make vigorous efforts to develop >systems capable of this, and they might well USE them..... My concern >becomes acute when said systems are pointed at ME -- so far, it hasn't >happened -- and I'd like it to stay that way. What about the machines themselves taking over? The computers Drexler proposed could supposedly pass a Turing test for being human. What is more human than taking over? I don't think the machine would see humanity as a threat, but more as a waste of resources (same for trees, mountains, CRAYs, etc.). If a nanocomputer the size of a Chevy Nova could out-think every computer and person currently in existence, what could a computer that covered the entire planet (several meters deep) do? I think the Chevy Nova sized machine would wonder about the same thing and try an experiment... It's not the GREY GOO we need to worry about it's the GREY MATTER !!! The possibilities of what a planet wide machine like that could do are incredible!!! It could launch a baseball sized nanomachine cluster at each of the other planets to start the same process (first analyze, then consume) if earth couldn't provide enough raw materials. Pity we wouldn't be around to marvel at the accomplishments. John