Join the conversation at CRNtalk!
CONDUCTED IN TWO PARTS
PART ONE, JANUARY 2004 [PART TWO, MARCH 2005]
Question 1: Tell us about yourself. What is your background, and what projects are you currently engaged in?
My background includes an M.S. in Computer Science from Stanford, six years at Electronics for Imaging as an embedded software engineer, and five years of dyslexia correction and research. I have been studying nanotech continuously since taking Drexler's class in 1988, and have co-authored a major nanomedical paper that was published last year.
My current project, which is taking all my time, is the Center for Responsible Nanotechnology, which I co-founded last year with Mike Treder. We realized that someone needed to talk about advanced molecular nanotechnology: its consequences, both good and bad, and how it could be administered to avoid the really bad stuff. That's not going to be easy to do!
Question 2: What are the differences between structural nanotechnology and molecular nanotechnology?
Structural nanotech is what's being done today in the labs. Working with all sorts of big chemicals and small pieces of stuff, learning how to make it and what it's good for. You can get clear sunscreen particles, glowing dots that don't fade, faster and more exotic electronics, and all sorts of other useful stuff that can be built into products.
Molecular nanotechnology, or MNT, is not about creating technologies to be built into products--it's about creating the whole product by molecular manufacturing, a few atoms at a time. Some of the products will be quite small, of course. A MNT-built computer could fit inside a bacterium. But some of the most useful products will be a lot larger. One likely way to do MNT is by mechanochemistry: using machines to position and guide chemical reactions. Each machine can only do one reaction at a time. So an important part of MNT will be building manufacturing systems that can duplicate themselves, growing exponentially to provide enough manufacturing capability to build more than a few nanograms of product.
MNT sounds like science fiction, and in fact we are only starting to develop the necessary capabilities. Some structural nanotech people seem to feel the need to criticize or debunk MNT. But they're really talking outside of their field. The science behind MNT is solid, and the technology is rapidly being developed. The only question is whether it will ever be desirable in practice, and it's pretty clear that it will.
Question 3: You have stated that we could have a molecular assembler built within a decade. Many nanotechnology enthusiasts are predicting that we will have an assembler within 30-50 years. Do you consider yourself more optimistic than other nanotechnology enthusiasts?
I don't know whether I'd call that "optimistic"--if it happens that quickly we may not be ready for it. I admit I'm a little bit ahead of the curve: back in the late 90's I was hearing a lot of people say 30-50 years, and now I'm hearing a lot of people say 20-30. I expect in a few more years a lot of people will be saying 10-20.
I don't have much doubt that technically, it would be possible to build an assembler in a decade. Maybe even less. Whether it will happen politically is another matter. It depends on a lot of factors. A group with a lot of money has to be convinced that it's worth spending the money on. Then they have to set up the right kind of R&D effort to make it happen. Of course, it will get easier as time goes on. Today, the effort might be comparable to building a completely new kind of spacecraft. A decade from now it could be more like designing a new line of automobiles.
One thing I'm pretty sure of: either the economic or the military benefits of advanced nanotech will convince someone to fund a special program that will create a working nanofactory, quite a few years before the general progress of technology reaches that level. That will be quite disruptive unless we're prepared to deal with it. I'm also pretty sure that we can't avoid it; if we try to forbid it or "relinquish" it, the most we'll do is drive it underground where it's less controllable.
Question 4: Lyle Burkhead on his site is critical of the concept of nanotechnology. He concedes that atoms and molecules can, and probably will, be manipulated in the future. But he argues that “Making things with atomic positioners will be at least as expensive as making them with biotechnology or bulk technology...atomic positioners will only be used to make things that could not be made in any other way."
What is your response to his critique of nanotechnology?
I've read Lyle Burkhead's site, and I think his arguments about advanced nanotech are well-reasoned and worth paying attention to. However, I disagree on practical grounds with the statement you quoted. Biotechnology is just not very efficient. For example, a field of plants generally turns only 1-2% of the sunlight into useful crops. And bulk technology gets a lot harder as you get smaller. We're already pushing those limits. A new semiconductor plant costs billions of dollars! By contrast, when we develop nanotech, it will be easy and efficient at a wide range of scales, from molecular to desktop and beyond.
I'm working on a paper right now that demonstrates how easily a tabletop nanofactory can be bootstrapped from a basic Merkle-style assembler. A nanofactory's products could be designed by traditional engineering techniques. They would be easy to use, and even with a fairly primitive nanofactory design they would be much stronger, more powerful, and contain vastly more computer cycles than products made by competing technologies. Assuming I'm right, we can have all the advantages of meter-scale nanotech manufacturing almost as soon as we develop a basic assembler, at a very modest additional cost.
I'm not sure that Lyle has considered the hidden costs in biotech and bulk manufacturing: things like transportation for the parts, and labor and floor space for the machines, and even the designers spending time to make designs that can be mass produced. With biotech or bulk, you need a lot of processes--usually in different buildings, sometimes on different continents--to make useful products. A nanofactory would simply eliminate many of those costs. And it would provide better materials, and far more intricate and efficient and controllable structures, than you could get with any other technology. The use of materials would be vastly more efficient. Assume nanotech-made diamond costs ten times as much as plastic. Which will cost more: a chair made of five pounds of wood, four pounds of plastic--or one-tenth of one ounce of diamond filled with water for rigidity?
Question 5: Molecular nanotechnology enthusiasts argue that molecular assemblers are both technically possible and economically and socially desirable, and are therefore inevitable. Yet in the 1950s, people made the same arguments about flying cars and jetpacks. Is it likely that future historians will lump molecular assemblers in the same category as flying cars and jetpacks?
It's possible, but only if something better than nanotech comes along. It turned out that jetpacks weren't all that practical--most people simply don't want to be pilots these days. Cars got better, and roads got better, and people didn't need jetpacks to get from place to place. If manufacturing gets a lot better, eventually we won't need nanotech. Or maybe we will have nanotech like we have helicopters today--expensive, special-purpose, and very valuable in its own niche. That fits with Lyle Burkhead's argument. But jetpacks and helicopters are useful for only a few things. Molecular manufacturing systems are so flexible that they'll be more like computers. Computers went in the opposite direction: from the famous quote by a leader of IBM about a world market for maybe five computers, to computers built into your cell phone, in only a few decades. I think the problem with nanofactories will be not their lack of viability, but how to keep them from proliferating too much and being used for too many purposes--some of them destructive.
Question 6: What is your assessment of artificial intelligence? Do you believe that is an inevitable spin-off of nanotechnology, or vice-versa?
The fastest supercomputer in the world today is the NEC Earth Simulator. It draws many megawatts of power and fills a large building. If that were done in the clunky, reliable mechanical nanocomputer design that Drexler invented as a proof of concept, it would take two watts and fit into a cubic millimeter. We will certainly have enough raw computer power to simulate a human brain. And we'll be able to probe millions of neurons simultaneously--invasively but non-destructively--in a living brain (a primate brain, if not a human brain). So in the sense of simulating a brain, I don't see any question: nanotech will enable AI. Whether we can make something better than a brain is another question. Personally I think we'll be able to reverse engineer the brain's function at both high and low level and come up with a design that still does language processing, visual processing, and creative thought, but is more useful (and has fewer ethical issues) than a simple brain simulation. And there are whole fields of AI that I haven't even mentioned, and won't get into here.
Will AI enable nanotech? That depends on the kind of AI. Systems are already being worked on that discover connections between separate fields of scientific research. That sounds helpful. But I certainly don't think we need AI to develop nanotech. And there are a lot of AI systems, like handwriting recognition, that are no help at all in engineering or invention. If we got something that was just generally smarter than a human in every way, of course that would help. But this is futuristic speculation, and I don't want to get into that. Eventually, I think both nanotech and AI are probably inevitable, and each one will simply be an enabling technology for the other. Each could help the other happen faster, but it won't be the thing that lets it happen, or forces it to happen.
Question 7: Speaking of artificial intelligence, how long do you think that Moore's Law will continue? Many are saying that Moore's law will end by 2010, and some are voicing skepticism that competing technologies (such as quantum computing or molecular electronics) will ever be viable. Is a nanotech/AI revolution still feasible if Moore's law soon dies?
I've already said that I think nanotech could happen by 2010, so if Moore's law lasts till 2010 then its eventual demise won't have any effect on our ability to do nanotech. In fact, the only thing we need faster computers for today is better mechanochemistry simulation; the rest of the system could easily be designed and implemented by today's level of computer power. You asked how long I think Moore's law will continue. It's not my field, of course. But ten years is a very long time in technology, and the idea that none of the molecular electronics schemes will be workable by then seems pretty implausible. There are at least three separate schemes, and some of them have been demonstrated already! I don't see any technical reason for Moore's law to stop until we get to at least the level of performance of a Drexler mechanical nanocomputer--and there are probably better designs after that. Of course, Moore's law could stop for economic reasons.
Question 8: Tell us about CRN.
The Center for Responsible Nanotechnology was created by myself and Mike Treder in 2002. We realized that 1) advanced MNT manufacturing is probably coming sooner than most people expected, and 2) people and institutions probably aren't going to be ready--unless someone does what we're trying to do.
I think most people don't realize just how disruptive MNT could be if it arrives before, say, 2040. It could create problems such as an unstable arms race, a black market in very powerful technology, irresponsible people building and releasing "grey goo" out of curiosity, economic disruption, social disruption, and so on. It's pretty unlikely that all the different competing special interest groups (including at least three different kinds of groups, based on force, money, and information) will accidentally reach a solution that avoids all of these problems.
So, we are working to demonstrate the power and timing of the technology, explain the risks, and develop a workable kind of administration. Then we have to go around convincing all the special interest groups that they'll be better off if they look at the big picture and sign up with a workable plan. It's a big job, but we think it's necessary. We've published a paper on the possibility of building restrictions into nanofactories so they can be widely used with comparative safety--regulation of the manufacturing and regulation of the products can be addressed separately. And the paper I mentioned about rapid nanofactory bootstrapping should help to demonstrate that nanotech can happen surprisingly rapidly.
Question 9: Most space enthusiasts are disappointed with the current state of space exploration. To what extent will nanotechnology facilitate the exploration and colonization of space? Is the colonization of space inevitable?
Nanotech would be a great help in reaching space and living in space. Aerospace hardware would be many times lighter, which saves fuel. Avionics would be literally billions of times lighter. That's just with first-stage molecular nanotech that can only build diamondoid. Advanced mechanochemistry would allow 100% recycling--life support--in a very small box. No more worries about how to grow wheat in zero-G. And MNT would allow the construction of things like Josh Hall's Space Pier, which make it a lot cheaper to reach orbit because most of the energy input is not from chemical fuel.
I have to say that the widespread colonization of space may not be a good idea; it's a lot harder to keep track of the use of advanced technologies when the users are on the other side of a big speed-of-light lag. I'm sure I'm going to be pilloried by space enthusiasts for that one! But the fact is, we just don't know yet whether it'll be safe to let unrestricted nanotech fly off in a spacecraft. Maybe it'll be possible with enough restrictions built in: you could visit space, even live there, but there would be some technology you couldn't use. This is one of the things that CRN will be studying.
Nanotech will also reduce the need for space access. When we can build anything we want with carbon, we won't have to go after metal-rich asteroids. When our technology is 10 to 100 times more efficient, we'll be easier on the environment down here. When we have the materials and construction techniques to build mile-high buildings cheaply, we'll be able to have cities a lot denser than we do today. A lot of people like to live in dense cities, and nanotech will be able to accommodate them and save some land. And, of course, we won't need as much farmland; with cheap manufacturing, greenhouses become quite viable, and they take a lot less land.
Question 10: You've criticized Michael Crichton's new book, Prey, as being technically inaccurate. Although the subject of molecular nanotechnology is currently being explored by many science fiction writers, few seem to be writing plausible, technically accurate novels. Why are so many authors struggling to write compelling, accurate science fiction stories pertaining to nanotechnology?
I'm not sure how much effort Crichton spent on being accurate. He wanted to get a message across, and I think he was more concerned about being compelling. I hope no one tries to learn about nanotechnology from his book! I haven't had much time to follow nanotech in science fiction, so I don't know what other authors have been producing or how accurate it is. Obviously some of it is just for entertainment, and is more like futuristic fantasy than like real science fiction.
It's just plain hard to write a plausible, accurate story about really advanced technology. You have to imagine the effects it will have on society, and also how society will use the technology. Of course if an author does not take the time to get educated, and probably to get the book reviewed by other technical people, they will make some pretty dumb mistakes. Some authors don't seem to care. Others spend a whole lot of effort getting it right, because they do care. I don't think there's any other way to write good science fiction.
When an author takes the time to try to get it right, it shows. I'll mention a couple of stories that show it's possible: "We Were Out Of Our Minds With Joy" by David Marusek and The Cassini Division by Ken MacLeod. It's obvious that they spent a lot of time on the technical details. There are some authors--I won't mention names--who simply try to weave nanotech into a pre-established or even recycled plot, and they tend to make all sorts of careless errors. I'll forgive one or two small technical errors as long as they're not the result of simple carelessness, or worse, not even caring about accuracy. Like it or not, people learn science from science fiction, and that's why I hold science fiction authors to such a high standard.
Question 11: How will mature nanotechnology affect the economy? Is it necessary or even wise for young people to make long-term investments, given the changes that will occur?
These days, it seems like "long term investment" means three years as opposed to three months. I doubt that any businesses are getting ready to change their business models yet. I don't think molecular nanotech is going to happen soon enough--at least unclassified; I don't know what the military may be planning--to have a serious effect on the economy yet. Structural nanotech, of course, is already making money for people, and it may be worth looking for money-makers there. But structural nanotech produces a whole bunch of little technologies that have to be built into products before they can make money. Anyone wanting to invest there should consider the soundness of the business model, not just the coolness of the technology.
If you're asking about IRA's and 401(k)'s, I don't know what to say. There's enough uncertainty about the future--from several sources, not just nanotech--that people may want to keep their money more accessible. But I can't possibly advise on something like that. I can't even tell you whether nanotech manufacturing will eliminate money or not, much less what would happen to your IRA if it did. In a lot of ways, it's good if we can keep using money post-nanotech; money is a good way to keep track of some things, and capitalism is a great system for solving certain kinds of problems. Of course nanotech will create some new possibilities for non-money systems. I've been thinking about how we can keep using money while still benefiting from the inherent abundance of nanotech manufacturing, but my ideas need more research before I'll be confident in them.
Question 12: What are your plans for the future?
I expect CRN to take all my attention for the foreseeable future. I think the questions that CRN is addressing will be very important to the survival of society--possibly the entire species--a decade or two from now. In the near term, I'm going to be researching and writing like crazy; we plan to publish more than a dozen papers in six months, and I'll probably be doing a lot of that writing myself. Of course, we plan to attract more co-authors and co-researchers. After CRN gets established and people start paying attention to our work, I expect I'll be traveling and speaking as well. I don't have any post-CRN plans.
PART TWO, MARCH 2005 [PART ONE, JANUARY 2004]
Tell us about the recent activities of the Center for Responsible
Nanotechnology. Are you pleased with the progress of CRN?
This interview was conducted by Sander Olson. The opinions expressed do not necessarily represent those of CRN.
Copyright © 2002-2008 Center for Responsible Nanotechnology TM CRN was an affiliate of World Care®, an international, non-profit, 501(c)(3) organization.