josh@planchet.rutgers.edu (06/29/89)
[Well, this is the first Update I've sent with the new software. Let's see if it makes it intact. As always, this is patched together from a MAC binary file, but they've switched from Ready-Set-Go to Pagemaker, which has a nasty habit of knocking hunks out of words. How'd *you* like to have to guess "meristematic" from "mer^%#@atic"? (I kid you not...) Anyway, any typos or other idiocy herein is almost surely mine. --JoSH] +---------------------------------------------------------------------+ | The following material is reprinted *with permission* from the | | Foresight Update No 5, 01 Mar 89. | | Copyright (c) 1988 The Foresight Institute. All rights reserved. | +---------------------------------------------------------------------+ FORESIGHT UPDATE No. 5 A publication of the Foresight Institute Preparing for future technologies Board of Advisors Stewart Brand Gerald Feinberg Arthur Kantrowitz Marvin Minsky Board of Directors K. Eric Drexler, President Chris Peterson, Secretary-Treasurer James C. Bennett Editor Chris Peterson Publisher Fred A. Stitt Assembler Russell Mills Publication date 1 Mar 89 Copyright 1989 The Foresight Institute. All Rights Reserved. If you find information and clippings of relevance to FI's goal of preparing for future technologies, please forward them to us for possible coverage in FI Update. Letters and opinion pieces will also be considered; submissions may be edited. Write to the Foresight Institute, Box 61058, Palo Alto, CA 94306; Telephone 415-948-5830. **************************************************************** NANOTECHNOLOGY POLICY A team of graduate students and faculty at the Lyndon B. Johnson School of Public Affairs at the University of Texas at Austin has been asked to conduct a study of the political and economic ramifications of nanotechnology. Headed by Dr. Susan Hadden, this research project is in the format of a two-term course. The fall 1988 term started off with an introductory lecture by Eric Drexler; the students went on to study the technology itself and the effects of other formerly new technologies such as biotechogy. This spring they will attempt to predict the kinds of social, economic, and political changes inherent in widespread adoption of nanotechnology, and will review various policy responses and their possible effects. The effort is funded by Futuretrends, a nonprofit educational group. Roger Duncan, president of Futuretrends and longtime Foresight supporter, initiated the project, which is expected to release its report in mid-summer 1989. Thanks Once again there are too many people deserving thanks for all to be listed here, but the following is a representative group: Michael Schrage for pointing the Rockefeller Foundation in our direction, Peter C. Goldmark, Jr., for investigating nanotechnology for the Rockefeller Foundation, Ray and John Alden for continuing useful advice, Peter Schwartz and Stewart Brand of the Global Business Network for help with the planned technical conference, David Gagliano for looking into research funding sources, the Seattle NSG for putting on Nanocon, Time-Life Books for covering nanotechnology, and Ed Niehaus for public relations help. Others are mentioned throughout FI publications. FI Policy The Foresight Institute aims to help society prepare for new and future technologies, such as nanotechnology, artificial intelligence, and large-scale space development, by: promoting understanding of these technologies and their consequences, formulating sound policies for gaining their benefits while avoiding their dangers, informing the public and decision makers regarding these technologies and policies, developing an organizational base for implementing these policies, and ensuring their implementation. FI has a special interest in nanotechnology: at this early stage, it receives relatively little attention (considering its importance), giving even a small effort great leverage. We believe certain basic considerations must guide policy: Nanotechnology will let us control the structure of matter, but who will control nanotechnology? The chief danger isn't a great accident, but a great abuse of power. In a competitive world, nanotechnology will surely be developed; if we are to guide its use, it must be developed by groups within our political reach. To keep it from being developed in military secrecy, either here or abroad, we must emphasize its value in medicine, in the economy, and in restoring the environment. Nanotechnology must be developed openly to serve the general welfare. The Foresight Institute welcomes both advocates and critics of new technologies. **************************************************************** Nanotechnology Online There is a nanotechnology Netnews group, sci.nanotech, on the USENET system. The USENET newsgroups form a large, distributed, hierarchical electronic bulletin board; formerly available only to those with UNIX machines, it is now accessible to anyone through services such as the WELL at 415-332-6106 (data), 415-332-4335 (voice) and the Portal at 408-725-0561 (data), 408-973-9111 (voice). In cooperation with FI, sci.nanotech carries most FI publications. The moderator is Josh Hall (josh@klaatu.rutgers.edu or rutgers!klaatu.rutgers.edu!josh), who can answer specific questions about the group by electronic mail. **************************************************************** Meet the Advisors: Meeting News January and February saw a number of nanotechnology-related events: MIT's annual symposium (see report elsewhere in this issue), a lecture at Bell Communications Research (a spinoff of Bell Labs) on Jan. 13, major coverage of protein folding and design at the AAAS meeting in San Francisco, a lecture at Silicon Valley's Software Entrepreneurs Forum on Feb. 17, a nanotechnology Physics Colloquium at the University of Seattle, and Nanocon, a regional meeting sponsored by the Seattle Nanotechnology Study Group. The last three have yet to occur as we finish this issue, so will be reported on next time. In addition to the items listed in the "Upcoming Events" column, both Hewlett-Packard and Union Carbide are planning meetings to discuss nanotechnology. "Nanotechnology: Prospects for Molecular Engineering" was the title of a symposium held at MIT on January 11-12. Sponsored this year by both the MIT Nanotechnology Study Group (MIT NSG) and the Foresight Institute, a nanotechnology event has been held at MIT annually since 1986. The introductory lecture was given by Eric Drexler, in which the technical foundations of the case for nanotechnology were laid out and basic designs described. Next, David Pritchard of MIT's Physics Department described his work on laser trapping and the use of optical standing waves to diffract beams of sodium atoms. In conversation after his talk, he noted that it is possible to use optical trapping to confine atoms to a space small compared to a wavelength of light, but that positioning is quite inaccurate on an atomic scale. This inaccuracy (with today's technology, at least) precludes "optical assemblers" for molecular structures. Adam Bell of the Technical University of Nova Scotia described computer-aided design and its role in design for nanotechnology. He emphasized the usefulness of designing a uniform language for describing systems, and the need to develop new engineering methodologies in this new domain. Ray Solomonoff of the MIT NSG spoke on "Managing Innovation" with particular emphasis on the prospects for managing nanotechnology as it arrives. His talk highlighted the practical parallels between self-replicating molecular machinery and self-improving artificial intelligence. He expects that the latter, in particular, is apt to bring an abrupt transition in knowledge, technology, and world affairs. Jeff MacGillivray, also of MIT NSG, looked at the economics to be expected in a world with nanotechnology. He asked "what will be of value?" His answers included land, resources, and human services. On the second day, several of the previous speakers were joined by Marvin Minsky (of the MIT Artificial Intelligence Lab and Media Lab) and Paul Saia (of Digital Equipment Corporation) in panel discussions of the technical basis of nanotechnology and the timeframe for its arrival and of the social impact and implications expected from nanotechnology. The second day's events were capped off by a more informal meeting for those interested in further pursuing the issues raised. MIT NSG and FI would like to thank the groups whose funding made the symposium possible: MIT's Departments of Electrical Engineering and Computer Science, Materials Science and Engneering, Mechanical Engineering, and Physics; MIT's Alumni Association, IAP Funding Committee, and Media Laboratory; and the Digital Equipment Corporation. **************************************************************** Nanotechnology & Microengineering Progress by Chris Peterson This quarterly newsletter covers a wide range of topics, arranged under five headings: Microchips: semiconductor technology Micromachines: miniature and micromachines Molecular Engineering: genetic engineering and biotechnology Sensors: a subset of micromachines Microstructures and Micromechanics: includes new materials, computer advances, brain theory, math, and semiconductor fabrication, as well as microstructures and micromechanics The items consist of abstracts describing research news, technical papers, company announcements, and patents. Company profiles and brief market analysis comments also appear. The publication's commercial slant will be useful for investors. The enabling technologies leading toward nanotech--protein and other polymer design, supramolecular (and biomimetic) chemistry, and STM/AFM based micronipution--were not covered in the premiere issue we saw, so the publication's name seems a bit of a misnomer. However, this has the advantage for potential subscribers that N&MP should have little if any overlap with Update. The newsletter does a good job at summarizing progress in various micron-scale technologies. For the technically literate reader who wants to keep up with these, for business or other reasons, this publication could easily be worth the subscription price. N&MP is available from STICS, Inc., 9714 South Rice Ave., Houston, TX 77096, (713) 723-3949. It is edited by Donald Saxman and costs $200 per year, with a 25% discount for libraries, universities, and medical schools, and an extra $20 charge for overseas airmail. A sample copy of the first issue costs $20. **************************************************************** The Foresight Institute, in cooperation with the Global Business Network, is planning a small technical colloquium on nanotechnology, to be held in Palo Alto in fall 1989. This invitational meeting will help researchers in enabling technologies make contact and communicate their goals and concerns. Potential attendees will be asked to submit position papers describing their interests. Additional information will be announced as it becomes available. Since last issue we've received news of a Hypermedia Design Workshop held last October. Organized by Jan Walker of DEC and John Leggett of Texas A&M, and funded by DEC. It was the first of two invited hypermedia meetings. The goal of the first was to bring together representatives of as many of the major hypertext media systems as possible, have them compare designs, and design a hypermedia storage substrate that would support the various systems. The second meeting is planned for Texas in early 1989 and will look at user interface issues and standards. The following are the participants, their organitions, and the systems they've worked on: Rob Akscyn (Knowledge Systems: KMS), Doug Engelbart (McDonnell-Douglas: NLS, Augment), Steve Feiner (Columbia: FRESS, Interactive Graphical Documents), Frank Halasz (leader of a new hypertext team at MCC; also Xerox's NoteCards), John Leggett (Texas A&M: teaches graduate course on hypertext), Don McCracken (Knowledge Systems: ZOG, KMS), Norm Meyrowitz (Brown: Intermedia), Tim Oren (Apple: HyperCard), Amy Pearl (Sun: Sun Link Service), Mayer Schwartz (Tektronix: Neptune/HAM), Randy Trigg (Xerox PARC: TEXTNET, NoteCards), Jan Walker (DEC: Concordia, Symbolics Document Examiner), Bill Weiland (U of Maryland: Hyperties). The only (known) major hypermedia systems not represented were Xanadu and Guide. Marc Stiegler, Xanadu's Director of Product Development, reports that all team members were locked in their offices, creating software and therefore unable to attend. Nanotechnology Education A Foresight Institute Briefing paper is available for students and others who want to learn the basics underlying nanotechnology. Send a stamped, self-addressed envelope to FI and ask for Briefing #1, "Studying nanotechnology." **************************************************************** by Chris Peterson The goal of nanotechnology and the engineering approach needed to reach it are receiving increasing attention within the biotechnology community, particularly among protein designers. Drawn from a pure science background, these researchers are being pulled increasingly in the direction of designing and building new structures, a task for which creative engineering skills are needed. This interest has shown up at two meetings: At the First Carolina Conference on Protein Engineering (held last October) the subject was raised by researcher Bruce Erickson of the University of North Carolina's chemistry department. As the chair of the session on Nongenetic Engineering, he led off with a reading from the book Engines of Creation and recommended it to the audience. This January's American Association for the Advancement of Science conference in San Francisco, which included substantial coverage of protein engineering, featured a plenary lecture given by Frederic Richards of Yale's Department of Molecular Biophysics and Biochemistry. In it, he highlighted one paper in particular: the 1981 PNAS paper describing a path from protein engineering to control of the structure of matter(1). While it is too soon to tell whether the protein path to nanotechnology will be the fastest, the goal is becoming clearer to researchers in that field. 1. Drexler, K.E., Proceedings of the National Academy of Sciences, 78:5275-5278, 1981. **************************************************************** Upcoming Events Nanotechnology Keynote address, April 21 evening, American Humanist Assoc., LeBaron Hotel, San Jose, CA. Part of a weekend-long conference. Contact 408-251-3030. Human Genome Project Conference, April 23-25, Alliance for Aging Research and AMA, J.W. Marriott Hotel, Washington, DC. Dinner lecture on nanotechnology on 24th. Contact 800-621-8335. HyperExpo, June 27-29, Moscone Center, San Francisco. Trade show covering hypermedia and related topics. Contact American Expositions, 212-226-4141. Second Conference on Molecular Electronics and Biocomputers, Sept. 11-18, Moscow, USSR, $150. Contact P.I. Lazarev, Institute of Biophysics of the Academy of Sciences of the USSR, Pushchino, Moscow Region, 142292, USSR. First Foresight Conference on Nanotechnology, during October or November, 1989, Foresight Institute and Global Business Network, Palo Alto, CA. Small technical meeting; see writeup in this issue. **************************************************************** The Foresight Institute receives hundreds of letters requesting information and sending ideas. Herewith some excerpts: I am enclosing an article, "Microscopic Motor is a First Step," by Robert Pool, published in the Oct. 21 Science. The article discusses recent developments in micron-scale mechanics. I have seen similar discussions elsewhere. A problem in this and other stories on micromachines is that they often confuse information about nanomachines with information about micromachines. Expecting micromachinery (which is developing sooner) to accomplish the tasks of nanomaery could disillusion proponents of micromachinery, and mistakenly discredit claims for nanomachinery. The practical distinction between these two technologies needs to be clarified. Tom McKendree Garden Grove, CA I agree that progress toward nanotechogy can't be prevented by any sort of organized suppression. If any one group, country, or group of countries tries, someone else will make breakthroughs eventually. And I believe that once it's developed, any attempt to make the technology proprietary will be short-lived. The secrets of nanotechnology will be simply too important to not have attempts made at espionage, midnight computer hacking, and theft... "The Problem of Nonsense in Nanotech" points out the problems of spreading misconceptions about nanotechnology and their interference with foresight. If nanotechnology has the capacity to utterly change society for better or worse, then the general public needs to get prepared. But I don't think people will be as interested in the details of the technology as much as the kinds of change that will be brought about. Engines of Creation points out a future, perhaps as close as the first half of the 21st century, with fantastic posbilities. To spark curiosity and interest it's necessary to discuss things like remaking the physical shape of humanity. But dwelling on the fantastic would be an open invitation to bogosity. To make best use of the next few decades, the public needs to focus on issues such as population control, active shields, environmental renewal, and space exploration. John Papiewski Elgin, IL Currently I am a student at USC, interested in pursuing a career in the field of nanotechnology...How should I structure my curriculum in order to pursue this goal? It seems that this field is interdisciplinary in nature, consisting of physics, electrical engineering, molecular biology, and chemtry. Is there a common denominator? Shawn Whitlow Manhattan Beach, CA Students and others interested should request a free copy of FI's Briefing #1, "Studying nanotechnology." Send a stamped, self-addressed envelope. **************************************************************** Will the BioArchive Work? by David Brin David Brin and Eric Drexler discuss here the challenge of using nanotechnology to restore species from preserved tissue samples, for species where habitat protection and captive breeding have failed. Dr. Brin holds a doctorate in astrophysics, works as a consultant to NASA and the California Space Institute, and teaches graduate-level physics and writing. He also writes award-winning science fiction. Dr. Brin leads off: I have one minor cavil with the notion of gene banks being sufficient to preserve the information inherent in the gene pools of species. Certainly I use this concept extensively in my own latest novel (titled Earth, it includes brief mentions of nanotech). But we should not be so blithe in assuming the Gene Library contains all. There is also Process Implemtation: getting the initial genetic mix to initiate and maintain the processes leading to a complete organism. Take the mitochondria and other purported "guest" genomes within eukaryotic cells. These symbionts are different in each species. They must be included. Then take the nurturing environment of the womb/egg. These are programmed in, all right, but at the other end of the library, where the shelves read "this is how the be a mother," not "this is how to be an embryo." This becomes incredibly complicated in placental mammals, in which certain genes are apparently turned on or off depending on whether they were delivered by the sperm or by the egg. So the concept of recreating lost species from their recorded information, while worthy and desirable, is not going to be any trivial undertaking. Even when the day comes that we can read an entire genome of, say, a blue whale, that'll be a far cry from making one, even with nanomachines. The last fly in the ointment is the apparent language of development, in which one gene doesn't express directly into one macroscopic trait. Rather, it's the pull and tug of a thousand enzyme selection sites, all playing against each other, that results in a cell here deciding to become a neuron and another over there deciding to become a bit of alveoli. Each enzyme site may take part in hundreds of cuing operations simultaneously. Indeed, it may turn out simpler just to disassemble a blue whale, cell type by cell type, and store that information, using nanomachines to build an adult from scratch! Eric Drexler responds: Some thoughts on restoring species, given frozen tissue samples and advanced nanotechnology: It is indeed important to save more than just nuclear genes, especially in the minimal sequence-of-bases sense. There are also mitochondrial genes, patterns of DNA methylation, obscurely-encoded states of genetic activation, and who knows what. Freezing entire tissue samples (and, for insects, entire organisms) answers several of these concerns, because it saves numerous cell types with essentially full information. For plants, which can typically regenerate from any meristematic cell, samples will clearly be adequate. Restoring animal species will be more challenging. Setting aside several separate problems (such as genetic diversity, habitat restoration, and lost culural information), it would be adequate to reconstruct fertilized eggs, and to raise the organisms to adulthood. Starting with only somatic cells and a thorough knowledge of a hypothetical martian biology, this might be impossible. But we will enjoy the advantage (one hopes!) of having closely related species available. To restore a beetle species, for example, one would integrate genetic (and other) information gathered from frozen cells into an egg having a general structure derived from one or more related species of beetle. This would be done after studying the relationship between somatic cell information and egg structure for a number of existing, normally-reproducing species; note that early embryology is terribly conservative, in an evolutionary sense. The resulting first generation might nonetheless have a somewhat atypical phenotype, but one would expect the offspring of that generation to be typical members of the original species. Mammals require more than just fertilized eggs, but embryos from endangered species have already been brought to term by host mothers of related species, even where the relationship between the species is not terribly close. By the way, I agree with your evaluation of the relative difficulty of (1) projecting an adult organism from its genes (etc.) and (2) constructing tissues or organisms from scratch after a molecule-by-molecule study of the original. The first involves a recipe, the second a blueprint; only blueprints describe products and leave a choice of implemtation strategies. **************************************************************** by Chris Peterson Government decisions being made now and over the next few years will influence these critical points: Will nanotechnology be developed openly, or will it be classified and developed secretly in government labs? Will individuals be permitted to publish freely on hypertext publishing systems, or will the system owners be forced to censor their writings? A recent report addresses these questions without ever using the terms "nanotechnology" and "hypertext publishing." Surprisingly, the report is published not by a private high-tech think tank, but by the US Congresss Office of Technology Assessment. FI participants who care about these issues will want to order this remarkable work. It's the best short introduction we've seen to the new challenges to freedom of speech and the press, and to how these challenges affect developing technologies. Well-written, the report alternately horrifies and encourages the reader with its review of past and possible future govment actions likely to affect nanotechnology and hypertext publishing. Did you know, for instance, that the US government can classify technical work done by independent, private researchers who take no public money? Or that it can block a patent or any disclosure by an inventor, even if the government has no right to the invention in question? Although these powers might not withstand a Supreme Court challenge, they appear to be current government policy. While the report does not take an advocacy role per se, it does clearly present arguments against such restrictive policies, some of which date back to the 1940s. Readers who pursue this topic further will find that FI Advisor Arthur Kantrowitz is a major proponent of a policy of openness. (Surprisingly to some, Edward Teller, often termed the father of the H-bomb by the media, also advocates openness.) The basic argument is that an open society, in addition to being more free, will progress faster technically, and be better able to defend itself, than one which binds its minds with over-classification. According to this theory, only critical short-term military information, such as codes and troop movements during a war, should be classifiable. Although nanotechnology is still far too theoretical for classification to be likely now, the issue will eventually arise. We need to understand current policies and, if necessary, make improvements before problems become acute. The OTA report's section on electronic publishing gives a remarkably clear view of such systems, which are described as the future crucible of cultural change. While it misses the basic concept of linked hypertext, other key features of hypertext publishing are described: the system is seen as decentralized, more as a clearinghouse for exchange of news and information than as a gatherer itself, in which the users themselves can be reporters and publishers. Information is preprocessed or screened to each individual's taste, either by a host computer sending the data or by the user's personal computer, perhaps by an artificially intelligent front-end. The report tackles the critical issue of how such a system will be regulated. Legally, publishers are responsible for their output and can be sued for libel or for publishing false, damaging information. However, in an open hypertext publishing system inviduals will be free to publish their own material; it will not be prescreened by the system's owner. But if regulators or the courts regard the system's owner as the publisher, then the owner would be forced to verify and police all information on the system--a crippling, impossible task. In contrast, under today's law, common carriers such as the phone system can't be held liable for what goes over their lines, since they obviously have no control over it. Newspapers also are exempt from liability for material they are legally required to publish. Obviously hypertext publishing systems, lacking control over what is published, should likewise be exempt from liability. The report goes so far as to say that holding electronic publishers liable may conflict with First Amendment rights. As the report makes clear, whether a new system is treated as a common carrier in this way is a political decision rather than a technical one. The report makes the obvious suggestion that in such a system the actual publisher of each piece of information should be held liable. It gets confused on this point, however, by implying that for some information it may be impossible to identify a responsible party. We would disagree: a system can be set up such that every item it contains is linked to a responsible party, typically whoever paid for its publication on the system. Authorship could be kept anonymous in some cases, yet made available by a court order when necessary. Again the national security issue is raised: it has been suggested that security concerns would force the screening of database entries for militarily-valuable information, or require database subscriber lists to be turned over to the government. The former would render the system uneconomic; the latter violates the right to privacy. Those wishing to pursue the issues raised by electronic publishing will want to see Ithiel de Sola Pool's excellent book, Technologies of Freedom, which is quoted in the report. But for those desiring a compact introduction to these critical issues, we suggest ordering a copy of the report: Science, Technology, and the First Amendment, GPO stock number 052-003-01090-9. It can be obtained by mail from the Superintendent of Documents, Government Printing Office, Washington, DC 20402-9325, or by calling 202-783-3238. Checks in US currency, Visa, Mastercard, and Choice cards are accepted. The cost is $3.50 within the US, $4.40 outside. **************************************************************** Books Signal: Communication Tools for the Information Age, ed. Kevin Kelly, Harmony Books, 1988, paperback, $16.95. A Whole Earth Catalog focusing on high tech subjects, mixes serious items (e.g., FI) with lighter ones. Foreword by FI advisor Stewart Brand. Filters Against Folly, by Garrett Hardin, Penguin Books, 1985, paperback, $7.95. A respected en?ist looks at the relationship between ecology and economics over time, pointing out the problems of commonization and the error of thinking every worldwide problem is global. A systems approach to a difficult problem; highly recommended. One jarring note: Hardin's seeming belief that economics is a zero-sum game. Molecules, by P.W. Atkins, Scientific American Library Series #21 (distributed by W.H. Freeman), 1987, hardcover, $32.95. Lavishly illustrated and elegantly written in nontechcal language, it makes the molecular world understandable. Requires no prior knowledge of chemistry. Computer-Supported Cooperative Work, ed. Irene Greif, Morgan Kaufman, 1988, hardcover, $36.95. A collection of papers on groupware and hypertext. Includes classic visionary papers by Vannevar Bush and Douglas Engelbart, interesting work by Thomas Malone, Robert Johansen, Xerox PARC, others. Text, Context, and Hypertext, ed. Edward Barrett, MIT Press, 1988, hardcover, $35. Diverse set of papers on how computers and hypertext have changed the way people write using computers. Strong emphasis on computer documentation. Quality is uneven, with some overlap, but includes some noteworthy papers. The Ecology of Computation, ed. Bernardo Huberman, Elsevier Science Publishers, 1988, paperback, $39.50. Now available in a somewhat more affordable edition. Open-systems perspective on advanced computing. Includes a set of three papers on agoric market-based computation. For the computer literate. Proteins: Structures and Molecular Properties, by Thomas E. Creighton, W.H. Freeman, 1984, hardcover, $37.95. Invaluable reference for protein designers and nanotechnologists thinking about molecular self-assembly. Quanta, by P.W. Atkins, Clarendon, 1974 (reprint 1985), paperback, $29.95. Qualitative explanations of quantum theory concepts with a bare minimum of mathematics, in dictionary format. A reference rather than a beginner's text. **************************************************************** +---------------------------------------------------------------------+ | This material is based on and builds on the case made in the book | | "Engines of Creation" by K. Eric Drexler. | | It is reprinted with the additional permission of the author. | +---------------------------------------------------------------------+ by K. Eric Drexler Artificial intelligence, like nanotechnology, will reshape our future. nanotechnology means thorough, inexpensive control of the structure of matter, and early assemblers will enable us to build better assemblers: this will make it a powerful and self-applicable technology. Artificial intelligence (that is, genuine, general-purpose artificial intelligence) will eventually bring millionfold-faster problem solving ability, and, like nanotechnology, it will be self-applicable: early AI systems will help solve the problem of building better, faster AI systems. AI differs from nanotechnology in that its basic principles are not yet well understood. Although we have the example of human brains to show that physical systems can be (at least somewhat) intelligent, we don't understand how brains work or how their principles might be generalized. In contrast, we do understand how machines and molecules work and how to design many kinds of molecular machines. In nanotechnology, the chief challenge is developing tools so that we can build things; in AI, the chief challenge is knowing what to build with the tools we have. To get some sense of the possible future of AI--where research may go, and how fast--one needs a broad view of where AI research is today. This article gives a cursory survey of some major areas of activity, giving a rough picture of the nature of the ideas being explored and of what has been accomplished. It will inevitably be superficial and fragmentary. For descriptive purposes, most current work can be clumped into three broad areas: classical AI, evolutionary AI, and neural networks. Classical AI Since its inception, mainstream artificial intelligence work has tried to model thought as symbol manipulation based on programmed rules. This field has a huge literature; good sources of information include a textbook (Artificial Intelligence by Patrick Winston, Addison-Wesley, 1984) and two compilations of papers (Readings in Artificial Intelligence, Bonnie Lynn Webber, Nils J. Nilsson, eds., Morgan Kaufmann, 1981, and Readings in Knowledge Representation, Ronald J. Brachman, Hector J. Levesque, eds., Morgan Kaufmann, 1985). The standard criticism of AI systems of this sort is that they are brittle, rather than flexible. One would like a system that can generalize from its knowledge, know its limits, and learn from experience. Existing systems lack this flexibility: they break down when confronted with problems outside a narrow domain, and they must be programmed in painful detail. Work continues on alternative ways to represent knowledge and action, seeking systems with greater flexibility and a measure of common sense. (A learning program called Soar, developed by Allen Newell of Carnegie Mellon University in collaboration with John Laird and Paul Rosenbloom, is prominent in this regard.) In the meantime, systems have been built that can provide expert-level advice (diagnosis, etc.) within certain narrow domains. Though not general and flexible, they represent achievements of real value. Many of these so-called expert systems are in commercial use, and many more are under construction. Evolutionary AI When one reads "artificial intelligence" in the media, the term typically refers to expert systems. If this were the whole of AI, it would still be important, but not potentially revolutionary. The great potential of AI lies in systems that can learn, going beyond the knowledge spoon-fed to them by human experts. The most flexible and promising learning schemes are based on evolutionary processes, on the variation and selection of patterns. Doug Lenat's EURISKO program used this principle, applying heuristics (rules of thumb) to solve problems and to vary and select heuristics. It achieved significant successes, but Lenat concluded that it lacked sufficient initial knowledge. He has since turned to a different project, CYC, which aims to encode the contents of a single-volume encyclopedia, along with the commonsense knowledge needed to make sense of it, in representations of the sort used in classical AI work. Another approach to evolutionary AI, pioneered by John Holland, involves classifier systems modified by genetic algorithms. A classifier system uses a large collection of rules, each defined by a sequence of ones, zeroes, and don't-care symbols. A rule "fires" (produces an output sequence) when its sensor-sequence matches the output of a previous rule; a collection of rules can support complex behavior. Rules can be made to evolve through genetic algorithms, which make use of mutation and re?tion (like chromosome crossover in biology) to generate new rules from old. This work, together with a broad theoretical framework, is described in the book Induction: Processes of Inference, Learning, and Discovery (by John H. Holland, Keith J. Holyoak, Richard E. Nisbett, and Paul R. Thagard, MIT Press, 1986). So far as I know, these systems are still limited to research use. Mark S. Miller and I have proposed an agoric approach to evolving software, including AI software. If one views complex, active systems as being composed of a network of active parts, the problem of obtaining intelligent behavior from the system can be recast as the problem of coordinating and guiding the evolution of those parts. The agoric approach views this as analogous to the problem of coordinating economic activity and rewarding valuable information; accordingly, it proposes the thorough application of market mechanisms to computation. The broader agoric open systems approach would invite and reward human involvement in these computational markets, which disguishes it from the "look Ma--no hands!" approach to machine intelligence. These ideas are described in three papers (Comparative Ecology: A Computional Perspective, Markets and Computation: Agoric Open Systems, and Incentive Engineering for Computational Resource Management) included in a book on the broader issues of open computational systems (The Ecology of Computation, B. A. Huberman, ed., in Studies in Computer Science and Artificial Intelligence, North-Holland, 1988). Ted Kaehler of Apple Computer has used agoric concepts in an experimental learning system initially intended to predict future characters in a stream of text (including written dates, arithmetic problems, and the like). Called "Derby," in part because it incorporates a parimutuel betting system, this system also makes use of neural network principles. Neural nets Classical AI systems work with symbols and cannot solve problems unless they have been reduced to symbols. This can be a serious limitation. For a machine to perceive things in the real world, it must interpret messy information streams--taking information representing a sequence of sounds and finding words, taking information representing a pattern of light and color and finding objects, and so forth. To do this, it must work at a pre-symbolic or sub-symbolic level; vision systems, for example, start their work by seeking edges and textures in patterns of dots of light that individually symbolize nothing. The computations required for such tasks typically require a huge mass of simple, repetitive operations before patterns can be seen in the input data. Conventional computers simply do one operation at a time, but these operations can be done by many simpler devices operating simultaneously. Indeed, these operations can be done as they are in the brain--by neurons (or neuron-like devices), each responding in a simple way to inputs from many neighbors, and providing outputs in turn. Recent years have seen a boom in neural network research. Different projects follow diverse approaches, but all share a connectionist style in which significant patterns and actions stem not from symbols and rules, but from the collective behavior of large numbers of simple, interconnected units. These units roughly resemble neurons, though they are typically simulated on conventional computers, and the resemblance in behavior is often very rough indeed. Neural networks have shown many brain-like properties, performing pattern recognition, recovering complete memories from fragmentary hints, tolerating noisy signals or internal damage, and learning--all within limits, and subject to qualification. A variety of neural network models are described in the two volumes of Parallel Distributed Processing: Explorations in the Microstructure of Cognition (edited by David E. Rummelhart and James L. McClelland, MIT Press, 1986). Neural network systems are beginning to enter commercial use. Some characteristics of neural networks have been captured in more conventional computer programs (Efficient Algorithms with Neural Network Behavior, by Stephen M. Omohundro, Report UIUCDCS-R-87-1331, Department of Computer Science, University of Illinois at Urbana-Champaign, 1987). A major strength of the neural-network approach is that it is patterned on something known to work--the brain. From this perspective a major weakness of most current systems is that they don't very closely resemble real neuronal networks. Computational models inspired by brain research are described in a broad, readable book on AI, philosophy, and the neurosciences (Neurophilosophy, by Patrica Smith Churchland, MIT Press, 1986) and in a more difficult work presenting a specific theory (Neural Darwinism, by Gerald Edelman, Basic Books, 1987). A bundle of insights based on AI and the neurosciences appears in The Society of Mind (by Marvin Minsky, Simon and Schuster, 1986). Some observations For all its promise and successes, AI has hardly revolutionized the world. Machines have done surprising things, but they still don't think in a flexible, open-ended way. Why has success been so limited? One reason is elementary: as robotics researcher Hans Moravec of Carnegie-Mellon University has noted, for most of its history, AI research has attempted to embody human-like intelligence in com-puters with no more raw computional power than the brain of an insect. Knowing as little as we do about the requirements for intelligence, it makes sense to try to embody it in novel and efficient ways. But if one fails to make an insect's worth of computer behave with human intelligence--well, it's certainly no surprise. Machine capacity has increased exponentially for several decades, and if trends continue, it will match the human brain (in terms of raw capacity, not necessarily of intelligence!) in a few more decades. Meanwhile, researchers work with machines that are typically in the sub-microbrain range. What are the prospects for getting intelligent behavior from near-term machines? If machine intelligence should require slavish imitation of brain activity at the neural level, then machine intelligence will be a long time coming. Since brains are the only known systems with general intelligence, this is the proper conservative assumption, which I made for the sake of argument at one point in Engines of Creation. Nonetheless, just as assemblers will enable construction of many materials and devices that biological evolution never stumbled across, so human programmers may be able to build novel kinds of intelligent systems. Here we cannot be so sure as in nanotechogy, since here we do not know what to build, yet novel systems seem plausible. It is, I believe, reasonable to speculate that there exist forms of spontaneous order in neural-style systems that were never tested by evolution--indeed, that may make little biological sense--and that some of these are orders of magnitude better (in speed of learning, efficiency of computation, or similar measures) than today's biological systems. Stepping outside the neural realm for a moment, Steve Omohundro (see above) has found algorithms that outperform conventional neural networks in certain learning and mapping tasks by factors of millions or trillions. Thus, although there is good reason to explore brain-like neural networks, there is also good reason to explore novel systems. Indeed, some of the greater successes in current neural network research involve multi-level versions of back-propagation learning schemes that seem rather nonbiological (and Omohun's algorithms seem entirely nonbiological). In summary, AI research is rich in diverse, promising approaches. Our ignorance of our degree of ignorance precludes any accurate estimate of how long it will take to develop genuine, flexible artificial intelligence (of the sort that could build better AI systems and design novel computers and nanomechanisms). If genuine AI requires understanding the brain and developing computers a million times more powerful than today's, then it is likely to take a long time. If genuine AI can emerge through the discovery of more efficient spontaneous-order processes (or through the synergistic coupling of those already being studied separately) then it might emerge next month, and shake the world to its foundations the month after. In this, as in so many areas of the future, it will not do to form a single expectation and pretend that it is likely ("We will certainly have genuine AI in about 20 years." "Poppycock!"). Rather, we must recognize our uncertainty and keep in mind a range of expectations, a range of scenarios for how the future may unfold. Genuine AI may come very soon, or very late; it is more likely to come sometime in between. Since we don't know what we're doing, it's hard to guess the rate of advance. Sound foresight in this area means planning for multiple contingencies. **************************************************************** Ultimate Computing by Ralph Merkle Stuart R. Hameroff's Ultimate Computing: Biomolecular Consciousness and nanotechnology (Elsevier, 1987, $78), is an uncritical mix of fact, fancy, and fallacy. Hameroff says "...this book flings metaphors at the truth. Perhaps one or more will land on target..." Perhaps,but the reader must sort the hits from the misses. One miss is his central premise, that "...the cytoskeleton is the cell's nervous system, the biological controller/computer. In the brain this implies that the basic levels of cognition are within nerve cells, that cytoskeletal filaments are the roots of consciousness. (Italics in original.) Unfortunately, there is every reason to believe this is completely wrong. This casts something of a pall over the book. Hameroff's chapter on nanotechnology is better than his average, although it adopts the curious perspective that nanotechnology really began with Schneiker in 1986, with Drexler mentioned only in passing. (Readers can check Drexler's 1981 PNAS paper and decide for themselves.) This is explained by the acknowledgements which say that "Conrad Schneiker [Hameroff's research assistant] supplied most of the material on nanotechnology and replicators for Chapter 10..."Hameroff covers a lot of ground. He has chapters on the philosophy of the mind, the origin of life, the cytoskeleton, protein dynamics, anesthesia (a good chapter--Hameroff is an anesthesiologist), viruses, and nanotechnology. He gives his own qualifications in a dozen fields as "...an expert in none, but a dabbler in all..." He's mostly right. There are better books written by more qualified people--the reader is advised to select from among them. Dr. Merkle's interests range from neurophysiology to computer security; he also lectures on nanotechnology. **************************************************************** Dr. Mills has a degree in Biophysics and assists in the production of Update. by Russell Mills Electronics Researchers at Caltech, JPL, and Univ. Sao Palo, Brazil have designed (but not built) a molecular-sized shift register -- a memory storage device with 1000 times the density and a ten-thousandth the energy consumption of its VLSI equivalent. Bits are stored by bumping individual electrons into the energy levels of a polymer, where they are moved along the polymer as more bits are written. The design has been worked out in some detail, and specifies the orbital energy levels of the molecules, the rates of competing (error-producing) electron transitions, spacing of the polymers, and the timing of the read/write cycle. The authors state that the register and associated read/write devices could be implemented with current technology; they provide chemical formulas of candidate molecules. [Science 241:817-820 (12Aug88)] Molecular-sized conducting wires of lengths down to 3 nanometers have been made at the Univ. of Minnesota from polyacenenone and imide subunits. The researchers hope to generate 3-D networks under 10 nm in size using similar chemical techniques. [New Scientist (19May88)] Chemistry Herschel Rabitz at Princeton Univ. proposes using femtosecond laser pulses to excite molecules in solution, measuring their response, and using the data to craft another pulse--thus homing in on the pulse structure needed to produce a desired chemical reaction. Once the correct pulse structure is known, it could be used routinely to carry out the reaction while dispensing with the elaborate techniques now required to protect one part of the molecule while another part is being modified. If Rabitz's method works, it may shorten many of the paths to nanotechnology by drastically simplifying the assembly of complicated molecules. [Science News 134:6 (2Jul88)] Micromanipulation A technique has been developed at Bell Labs for trapping and manipulating micro-organisms without damaging them. A lens is used to focus a laser on the organism; light refraction results in a force that pushes it toward the focal point of the beam. Viruses and bacteria can be trapped and immobilized by the technique; larger cells, such as yeast or protozoa, can be dragged around by moving the beam. The investigators even found that they could reach inside a cell with the laser beam, grasp internal organelles and move them around. One wonders whether a similar technique could be used to assemble components of micromachines like those discussed elsewhere in this article. [Science 241:1042 (26Aug88)] Physicists at the National Bureau of Standards are now able to confine groups of sodium atoms between a set of laser beams and then slow down their motions to under 20 cm/sec. Under these conditions the properties of atoms can be studied with very high precision; such information will someday be needed for the design of nanomachines and zero-tolerance materials. [Science 241:1041-1042 (26Aug88)] A step forward in our ability to handle individual molecules has been made by Japanese researchers at Osaka Univ. who have directly measured the tensile strength of an intermolecular bond--by pulling on it until it broke. The bond is that between protein subunits in a skeletal muscle filament. The filaments are chains of "actin" molecules held together by non-covalent bonds; two such chains wind around each another to form an actin filament. Another protein, "myosin", contains the motor apparatus of the muscle. The researchers obtained a value of 108 piconewtons for the tensile strength of actin filaments. They proceeded to measure the force exerted by each myosin "motor as it pulls on an actin filament--about 1 pN. Since each actin filament is pulled on by roughly 50 myosin molecules, there would seem to be a safety factor of 2 built into our muscles. [Nature 334:74-76 (Jul88)] Viewing Biochemists at Cornell Univ. are now able to take 120 picosecond x-ray diffraction exposures of organic molecules and enzymes. This breakthrough is made possible by a magnetic undulator that produces an intense x-ray beam. Until now, x-ray diffraction analysis has required long exposures, especially for large molecules. Molecular motion would cause the images to blur, thus limiting the resolution obtained. With exposure times now reduced by a million-fold, it should be possible to watch enzymes change shape as they catalyze reactions and to troubleshoot nanomachines by observing them in action. [Science 241:295 (15Jul88)] Micromechanics An electric motor less than half a millimeter across, miniature air-driven turbines, and gear trains--these are among the various micromachines recently fabricated at the Univ. of Calif. at Berkeley, Cornell Univ., and Bell Labs using the techniques of integrated circuit manufacture. Intended to provide measurements of friction, wear, viscosity, lubrication, stress, deformation, fatigue and other factors at the scale of microtechogy, they may be forerunners of practical devices: tiny fans for cooling integrated circuits, drug-dispensing mechanisms for smart pills, cutting tools for unblocking blood vessels, cell sorters for diagnostic tests. Similar methods might be used to make even smaller machines, but true nanomachines are probably beyond the range of these techniques. [Science 242:379-380 (21Oct88)] Public attitudes Victim of numerous court-ordered delays inspired by unfounded fears, the U.S. biotechnology industry has finally realized that it can no longer take public awareness for granted. Some companies have dealt with the problem by hiring public relations firms to promote positive attitudes toward them; often this approach has led to company-sponsored public meetings in communities where the testing of genetically modified organisms is being planned. The effectiveness of the effort is already evident--more than a dozen field tests have been conducted recently without controversy. [Science 242:503-505 (28Oct88)] nanotechnology proponents: take note! Technophobia is an easy nut to crack when moderate resources are devoted to the effort. Protein engineering Wm. DeGrado's group at the duPont Co. has continued to make remarkable progress in protein design and production. Having designed a four-helix protein that self-assembles into a stable bundle, they proceeded to synthesize the gene for this protein, insert the gene into a bacterium, and show that the bacterium produces the desired protein. Although this effort aimed at studying the relationship between amino-acid sequence and 3-dimensional structure of proteins, the designed protein will probably be used as a platform for adding functional features. [Science 241:976-978 (19Aug88)] The molecules responsible for photon-capture in photosynthesis were mapped in detail several years ago. To find out how they work, scientists at MIT and Washington Univ. (St. Louis) are making amino-acid substitutions in the reaction center of photosynthetic bacteria. When they altered an important amino acid linking a chlophyll molecule with its protein support, one of the chlorophyll subunits lost its magsium atom--yet the system still functioned at about 50% efficiency. This suggests that photosynthesis does not depend critically on the molecular structures arrived at through traditional evolution, and that better and simpler molecules may be developed for powering some kinds of nanomachinery. [Science News 134:292] Biological membranes are equipped with a variety of channels connecting the inside and outside of cells or organelles. These channels, made of protein, can be opened and closed; when open they allow certain ions to pass through the cell membrane. Wm. DeGrado's group at duPont has designed and synthesized a number of simple ion channel proteins and tested their ability to form functional ion channels in a phospholipid membrane. The proteins were chains of 14 to 21 serine and leucine residues, arranged into helical structures with the polar serines running down one side and the apolar leucines along the opposite side. A number of these helices would then aggregate in parallel to form a cylindrical bundle around a central channel. The researchers determined that 21-residue proteins spanned the membrane and created a conductive path for ions. The amino-acid sequence of the proteins determined the number of helices in a bundle, and this in turn determined the size of ions that could pass through the channel. [Science 240:1177-1181 (27May88)] Protein engineering advances swiftly. In each of the following three summaries, researchers have programmed Esterichia coli bacteria to produce and secrete redesigned antibody molecules. Bacteria are far easier to program and grow than eukaryotic (nucleated) cells, but in earlier experiments bacteria would not output functional proteins. In the latest work the bacteria have been persuaded to produce "antigen-binding fragments" (Fabs) with the same specificity and affinity for their substrates as the original antibodies. Researchers at Max Planck Institute developed a bacterial expression system mimicking the one eukaryotes use. In eukaryotic cells, an antibody's protein chains are synthesized in the cell's cytoplasm, then transported into an organelle called the "endoplasmic reticulum," where they are trimmed, folded, bonded, and paired into a functioning configuration. The researchers first examined the 3-dimensional structure of the antibody MCPC603 and decided which portions of it to keep. They next constructed a custom plasmid (mini-chromosome) consisting of: the DNA sequences coding for the antigen-binding portions of the antibody's protein chains, two bacterial "signal sequences" coding for protein appendages that tell the bacterial cell membrane to secrete the proteins, and several other sequences required for replication and translation of the DNA via RNA into protein. When this plasmid was introduced into Escherichia coli, the bacteria used the new DNA to make and secrete the Fab protein chains. The chains then folded and bonded themselves correctly. [Science 240:1038-1041 (20May88)] A group at International Genetic Engineering, Inc. used essentially the same technique to produce a chimeric Fab consisting of antigen recognition domains taken from a mouse antibody, and the remainder taken from human antibody (presumably to forestall an immune attack on the Fab if it should be used therapeutically in humans). This particular Fab was chosen because it attacks human colon cancer cells. [Science 240:1041-1043 (20May88)] Genex Corp. researchers have gone a step further in simplifying antibody molecules. Traditional antibodies are composed of four polypeptide chains. In the Genex design, two of these chains are eliminated and the other two are joined by a short chain of amino acids. The result is called a "single-chain antigen-binding protein." Genes to encode several such proteins were constructed and expressed in E. coli. The proteins produced by the bacterium proved to have the same specificity and affinity for the substrates as the original antibodies. Single-chain antigen-binding proteins are expected to replace monoclonal antibodies in such areas as cancer and cardiovascular therapy, assays, separations, and biosensors. [Science 242:423-426 (21Oct88)] Amidases are enzymes that catalyze the hydrolysis of amide bonds. Of particular interest to biotechnologists are amidases specific for the amide bonds connecting amino acids together in proteins; what is needed are tools for cutting a protein at any desired place along its amino acid sequence. Researchers at Scripps Clinic and Penn State Univ. have overcome a major hurdle by developing a Fab that catalyzes the hydrolysis of a somewhat different amide bond joining two aryl components. Mice were immunized with a compound resembling the transition state of amide hydrolysis; whole antibodies collected from the mice were then enzymatically trimmed. The resulting Fabs sped up the hydrolysis reaction by a factor of 250,000. [Science 241:1188-1191 (2Sep88)] **************************************************************** by James C. Bennett Looking over Stewart Brand's bio, the verb that jumps out at one is "founded": he founded the Whole Earth Catalog, Point Foundation, Coevolution Quarterly (now the Whole Earth Review), the WELL (a regional computer teleconferencing system), and co-founded the Global Business Network. He also has taught at U. Cal. Berkeley and the Western Behavioral Sciences Institute, serves on the Board of Trustees of the Santa Fe Institute, been a Visiting Scientist at MIT's Media Lab, and written books including The Media Lab: Inventing the Future at MIT. The following is a discussion between Stewart and Jim Bennett, cofounder and Vice President of the American Rocket Company. Jim serves on FI's Board of Directors and will be profiled in a later issue. Editor FI: Foresight's ambition is to begin the debate about nanotechnology on a more reasonable, less polarized basis than previous debates about technology, such as that about nuclear power. How reasonable do you think this ambition is? SB: I don't know; it will be interesting experiment. The only previous attempt at anything like this that I can think of is the Asilomar conference on genetic engineering, where they got a lot of professionals together and tried to predict what the negative conquences of recombinant DNA experiments might be, and what measures would be reasonable to take to prevent such consequences. It's not clear how beneficial that conference was. A lot of opponents of genetic enneering took the statements made there, and, in effect, said "See, even the scientists had some doubts about this, so we should really be worried. FI: There you had a situation where a number of people were already polarized, so they essentially took advantage of the situation? SB: Yes, but now that I think of it, that was a first attempt. The second attempt at anything is usually quite different from the first time around. FI: If you were going give us "Stewart Brand's Rules for Productive Debate," what would they be? SB: Don't know yet. What's important is to get very smart people, who have ears as well as mouths. Some very smart people can't listen. FI: So one thing to do would be to be selective as to who to invite? SB: Word gets around as to who's good at conferences. Most people who are high up in science and technology spend a lot of time in conferences, and it's fairly easy to tell who are listeners as well as talkers. You can also tell a lot by how people talk on the telephone: some people just preach at you. FI: To what extent is it useful to get people who don't have scientific and technical backgrounds involved in the debate, and at what point is it useful to do that?SB: I think it's worth having people who are politically active involved at all stages of the process. You want to have both people who are astute technically and who are sophiscated politically. Some who are competent in science also have a practical knowledge of politics. Especially in fields like conservation biology, you need to have a comprehensive view of things so that the Costa Rican farmer doesn't get left out of the campaign, for example. You need to get people who know what it takes to negotiate agreement. And to negotiate disagreement, by the way. FI: I can see a lot of cases where you're not going to get to quick agreement among people. You are at least going to have the disagreements be productive rather than destructive. SB: You need to have people come in and say, "Yeah, we agree on 80% of this stuff," and then identify the items they disagree on, so that as further evidence or information becomes available those items can be resolved. FI: I think that the open-minded people you're describing here are the sort of people who would be interested in seeing the new information come in to resolve such points, rather than fearing being proven wrong by it... SB: Yes. Edward O. Wilson, the sociologist, is an example. When he first came out with his theories on sociobiology, based on his work on insect behavior, a lot of the liberals attacked him, because he contradicted their current beliefs. And he was willing to change and modify his views on the basis of argument and new information. A Noam Chomsky, on the other hand, tends to be more overbearing and hurt his field of linguistics with heavy-handedness. In the sociology instance, by the way, the liberals were probably as much wrong as right, not that they're likely to admit it. FI: Speaking of trying to bring in new information, to what extent do you think that new information technology such as hypertext, or other things such as you describe in The Media Lab, can improve the quality of debate? SB: It would be interesting to do an article on what is sometimes called "grey lit"ture"papers informally passed among scientistsdiscussing how that's progressed over the years. First it was just the exchange of letters among scientists, eventually formalized by the Royal Society, then it took a jump in the level of traffic with the arrival of the typewriter and carbon paper. The arrival of Xerox copying caused another major jump in traffic. Computer networks, starting with the ARPANET, caused another major jump. Maybe we're at a virtual hypertext level now. FI: I think the difference between what we have now and what hypermedia is intended to be is the ability to screen the material. With the mass of material we are now beginning to have, there has to be some way of indicating which material is worthwhile. SB: A lot of that'll be automatable, and I also expect that there will be a lot more humans editing material. A library is far more useful with a good librarian. FI: There has been some material on nanotechnology printed in Whole Earth Review. What kind of interest have you noticed from the WER readers? SB: The two populations which have shown interest are computer enthusiasts and the major corporations. I have spent some time in the past few years among major corporations, and they have a lot of interest in what the future has in store for them. Computer enthusiasts have a strong interest in it as wish fulfillment, while the corporate person is asking "what will this mean to my company?" FI: What about the people primarily interested in environmental issues? SB: They've been blind, deaf and dumb on the issue, as far as I can see. FI: When you look at the degree to which an anti-technological viewpoint is entrenched in some people, I don't see this as going away quickly. SB: I'm not sure you want it to go away quickly. Nanotechnology is the sort of thing which could take off exponentially, and could result in a lot of change happening very rapidly, things changing more rapidly than people can adapt. The no-sayers can help flatten that curve, make it arithmetical rather than exponential; of course, they want to see it stopped altogether. No-sayers have their place. I wouldn't want to see them go away. The Alaska pipeline is an example--the first proposal was strongly criticized by environmentalists; they said that it would wipe out the caribou, and so on. They were right in that it was a lousy pipeline design. But it was a bad pipeline design that was improved by delay, and by the pressure to go back and re-think the proposal. It's useful to have no-sayers, to slow up the process. But at the end you did have a pipeline, and it didn't do the terrible things they thought it would. So no-sayers have a role, even if they aren't always reasonable. Sometimes it's useful to have unreasonable people. Also remember that a good many environmentalists are highly reasonable, and can be extremely astute on technical issues. Beware of characterizing comments; they invite reply in kind, such as that all nanotechnologists are unstable Mensoid nerds. Anyway, both reason and unreason have value in the big picture. FI. Doesn't that depend on whether you have a political and social system that can take people who are hard-over nos and have the result be a compromise, rather than giving them a veto over things? SB: There's a danger of change increasing exponentially. I don't think it's a matter of vetoes; I think that they end up just acting as a kind of brake. As far as how the U.S. political system works, I think it's worth reading Jerome Weisner's article in the January Scientific American. Look at what's happened with the Science Advisory Council, which was set up by Eisenhower as a response to Sputnik, and gave good advice to him and to Kennedy, but was reduced to ineffectiveness under Nixon and since then. The Challenger accident showed that correct technical information was not filtering up through the Cabinet agencies to the President. Perhaps if the Science Adviser's office been functioning properly, that information might have gotten through. FI: In fact there NASA has been waging a very strong and usually successful war against any other independent source of thinking on space in the Executive Branch. That's what's happened to the White House Office of Science and Technology Policy, for example. SB: Yes, but they couldn't control what was being said or done in the Soviet Union, or Europe, or elsewhere. FI: Or even in the American private sector. SB: Or even there. **************************************************************** by Mark Gubrud One of the goals of the Foresight Institute is to stimulate debate on the public policy consequences of advanced technologies such as nanotechnology. This essay will start off the discussion on military aptions of nanotechnology. The essays in this series are the opinions of the authors and not neccesarily those of FI. Editor When we contemplate the application of nanotechnology to weapons we find virtually unlimited room for fantasy. A number of clich's have arisen in the nanotech community: omnivorous robot locusts, omnipresent surveillance gnats, microbes targeted for genocide, mind control devices, and so on. But what makes good science fiction does not necessarily make an effective tool of combat. Will nanotechnology make nuclear weapons obsolete? Perhaps in peace, but not in war. Nuclear energy will remain preeminent in total war, for at least three reasons. First, it is infinitely lethal; chemical bonds cannot resist nuclear energy. Second, it is cheap, and nanotechnology will make it cheaper. Third, and most important, it is quick; the bomb goes bang and that's it, end of discussion. Nanotechnology might seem to make SDI's Rube Goldberg schemes workable, but space weapons will only create a final front. The principle of preemption--getting in the first blow, and aiming for a knockout--is an ancient and essentially unalterable fact of military life. Missiles are now targeted on missiles. And in a war involving space weapons, the first strike will be in space. Battles with first-generation, bulk technology space weapons will already be so swift that we will have to trust a machine to decide when to start shooting. Nanotechnology could produce huge numbers of such weapons, and also nuclear and chemical explosive-driven directed-energy weapons that will reduce the decision time praccally to zero, below even what a computer can cope with. We see it most clearly in space, but on every front the speed and numbers of today's high-tech and tomorrow's nanotech weaponry collapse decision time and undermine the basis of mutual deterrence. One does not have to callate that a first strike will succeed, one has only to fear that the other side may try it, perhaps as some conflict escalates or as some situation gets out of control. Preparing to attack is not generally distinct from preparing to defend or deter; defenses are needed against retaliation, and second strikes may aim at the same targets as first strikes. As in World War I, motion may be a slippery slope leading inexorably to war. Today that in?ity is mitigated by the gap between the time scale of crisis and combat and that of production and deployment. Nanotechnology will reduce and eventually eliminate this margin of safety. Replicating assemblers could be used at any time to initiate an arms buildup, one that could reach fantastic proportions in the time frame of historical military crises. The buildup would be exponential, and traditional order-of-battle correlations would still apply, so it would seem that whoever initiated the buildup (assuming equal technologies) would have supremacy--not falling behind would be a security imperative. Finally, the strike time compression of massively pro?ated and lightspeed weaponry would undermine mutual deterrence at the brink. These are the basic characteristics of the nanotechnic era that combine to make it militarily as different from the present as the present is from the pre-nuclear era. The difence is that no level of armament will be even metastable, not even complete disarmament. Perhaps nuclear disarmament and major conventional disarmament will be achieved, but each proud, independent nation still retain its vestigial military--including one nano-supercomputer, busily planning rearmament and war. Then one day a dispute could arise, and quickly develop into an awesome, nuclear-powered, nanotechnic struggle for the control of territory and matter. Large-scale space deployment would not change the essence of this situation. We cannot depend on the balance of terror to hold the peace, for even if there is ultimately no defense against nuclear weapons, especially not in space, there may still be temporary shelter in dispersal and/or underground. Deep tunnels and closed-cycle life support systems can provide a redoubt for entire populations, while their machines struggle for control of the open land, sea, air, and space and to penetrate the enemy's shelters. Nano/nuclear war could be a drawn-out struggle, and the victor would have means to clean up the mess and to remake the world. Or so it might seem. But in practice, hot war would probably break out before anyone was ready for it. There would be no assurance of destruction to hold back the first strike; rather, there would be great pressure to preempt, since the outcome might be decided in the first few microseconds. One could not afford to concede land, sea, air and space without a fight, despite the inevitable vulnerability of predeployments in these environments. On the other hand, a well-prepared, long war of attrition, with detralized and versatile assembler-based production, might kill everyone before one regime could neutralize all the others. The challenge of the nuclear era has been to limit arms and to resolve disputes between armed soveign states without recourse to war. The challenge of the nanotechnic era will be to abolish the armed sovereign state system altogether; otherwise military logic will always point toward fast rearmament and then to war. In the near term, the challenge will be to avoid star wars and a new Cold War. To governments, nanotechnology will suggest power, and power is dangerous in a divided and militarized world. For the world as a whole, nanotechnology will mean change, and even slow change has often been amplified by the world's complex and discontinuous system to produce violent results. To prevent such results, our development of nanotechnology must be fully open, international, and accompanied by a rising worldwide awareness of its significance and earnest planning for swift, neccesary, and unavoidable change in economic and security arrangements. Any leading force must include all potential nanotechnology powers, which does include the USSR at least! And it must lead, not force. In answer to the question of the military uses of nanotechnology: it must never have any at all. Mark Gubrud is a policy intern at the Federation of American Scientists. Your responses to his comments are welcome. +---------------------------------------------------------------------+ | Copyright (c) 1988 The Foresight Institute. All rights reserved. | | The Foresight Institute is a non-profit organization: Donations | | are tax-deductible in the United States as permitted by law. | | To receive the Update and Background publications in paper form, | | send a donation of twenty-five dollars or more to: | | The Foresight Institute, Department U | | P.O. Box 61058 | | Palo Alto, CA 94306 USA | +---------------------------------------------------------------------+
jac@paul.rutgers.edu (Jonathan A. Chandross) (07/01/89)
josh@planchet.rutgers.edu [from Update #5] > January and February saw a number of nanotechnology-related events: > MIT's annual symposium (see report elsewhere in this issue), a lecture > at Bell Communications Research (a spinoff of Bell Labs) on Jan. 13, Bell Communications Research (BCR) is not a ``spinoff of Bell Labs.'' When the Bell system was broken up by Justice, BCR was created to perform functions for the operation companies that had been previously handled by Bell Labs. It is a totally separate organization. The chart now looks like: Before: AT&T + Bell Labs / \ / \ / \ / \ After: AT&T RBOC (Region Bell Operating Companies) + + Bell Labs Bellcore Jonathan A. Chandross Internet: jac@paul.rutgers.edu UUCP: rutgers!paul.rutgers.edu!jac [Indeed, in a recent speech I heard the head of Bellcore claim it was bigger (now) than Bell Labs. However, I'm sure the reference was intended merely identify what kind of organization Bellcore is in as few words as possible. --JoSH]