CHRIS: I'll be talking today about near-future technology and the difference it could make in our lives. I'd like to start with a quote from Chris Peterson, who is the President of the Foresight Institute: "If you're looking ahead long-term, and what you see looks like science fiction, it might be wrong. But if it doesn't look like science fiction, it's definitely wrong." Less than 20 years ago, IBM came out with a very successful computer. It had a speed of 4 megahertz, about 1/30 of one megabyte of RAM, and no hard disk at all. So if I say anything today that sounds incredible, please remember how today's computers would look in 1982.
HOST: I'll try to keep that in mind. You mentioned the Foresight Institute. What is that?
CHRIS: A few years after that PC I mentioned, Eric Drexler saw the need for widespread discussion of future technology, how it would affect us, and how we could best prepare for it. He started the Foresight Institute to generate awareness of one particular technology, "nanotechnology", which will create major improvements in our ability to make material do what we want.
HOST: How does nanotechnology work?
CHRIS: You know all matter is made up of atoms. The way the atoms are stuck together can produce anything from a snowflake, to a bacterium, to a skyscraper. When you cook something, you're changing the way the atoms fit: sugar turns into caramel, vegetables become soft but eggs become hard, and so on. The cells of your body are manipulating atoms all the time to turn food into muscle and bone. The idea of nanotechnology is to grab hold of individual atoms and put them exactly where we want them. Rather than "cooking" trillions of them at once, and letting the processes happen randomly, each process would be controlled individually. We can't unscramble an egg--but a chicken can eat grain and create a new egg. It can do that because in each of its trillion cells, chemicals called enzymes are individually creating each molecule of the egg and putting it where it needs to go.
HOST: So what can nanotechnology make besides eggs?
CHRIS: Eggs are actually pretty hard to make. We'd want to start with something simpler. Crystals are pretty simple, because the atoms line up in regular formation. And some crystals can be very strong--diamond for example. There are more than a few people interested in using diamond as a structural material, if we can figure out how to build it cheaply enough. But at the atomic level, you just have to grab carbon atoms and put them in the right place.
HOST: Let me get this straight. You want to build things out of diamonds?
CHRIS: Yes. Big chunks of diamond are brittle. But think of glass vs. fiberglass. Make glass thin enough, and it'll bend just fine. The same is true of diamond. We could build very strong things out of fiberdiamond.
HOST: How big a thing could you build?
CHRIS: There are an immense number of atoms in anything that's even big enough to see. To build, say, a chair out of diamond, you would have to build lots of tiny little diamond pieces and then stick them together. This is not trivial. But the flip side is that, if we can figure out how to do this, we can stick those diamond pieces together a different way and build a table. Or a TV set, or a plate, or just about anything you can design.
HOST: Wouldn't that cost a lot?
CHRIS: Now we're getting to the really interesting part. Remember that egg. We couldn't build an egg in the lab today, and if we could it would cost millions of dollars. But you can buy them at the supermarket for a few dollars a pound. The reason is that chickens make eggs, and chickens also make more chickens. Chickens are pretty cheap, and in a short time you can have as many of them as you need to make eggs cheap too. So the way to make diamond cheap is to design a diamond factory that is built out of diamond. Then that factory can copy itself.
HOST: Is that possible?
CHRIS: Chickens make new chickens all the time. Of course they're living, and a factory is not living. But all it really takes is engineering. I've heard that the main Ford factory has just about enough machinery installed to make a new factory from scratch. Of course that's millions, maybe billions, of dollars worth of machinery. But the point is that it can be done. The other thing that Ford factory has is people running the machines. It would cost many millions to pay those people to work long enough to copy the factory. But the cells in the chicken don't have people running them, and we're getting closer to having fabrication machines that can work without people having to run them. Also, there are some tricks you can play at the atomic scale that make it easier to build a replicating factory.
HOST: What do you mean by "The atomic scale"?
CHRIS: A human hair is about 50,000 atoms wide. Three billion atoms would fit into a meter or a yard. So "atomic scale" means something capable of moving or sensing in steps of a billionth of a meter, more or less. That's called a nanometer, which is where the word "nanotechnology" comes from. We have microscopes today, called scanning probe microscopes, that can actually detect individual atoms and make pictures of them.
HOST: So what are some of the tricks you mentioned?
CHRIS: First of all, every carbon atom is basically like every other carbon atom. If you tried build a table saw from scratch today, you'd need to make washers, screws, wires, plastic, and so on, and it would take a lot of specialized technology to make all that stuff. Today, we are not able to design a robot that could do all the forging, stamping, extruding, cutting, and everything else that would be needed. But if you could build a table saw from just bits of diamond, all you'd need is the ability to take carbon atoms (which always come pre-made) and stick them together to make diamond bits, and then stick the diamond bits together like a jigsaw puzzle to make the gizmo. It'll probably be a lot easier to design robots to do each of those tasks. So by starting from pure atoms, and doing a few simple manipulations, you can build machine tools or anything else.
(Pause for questions)
Another trick has to do with precision. If you cut a piece of wood, the cut surface is not very flat. In factory-made stuff, the precision is better--you may be able to get a thousandth of an inch, or a ten-thousandth. But even when making computer chips--some of the most precise stuff around--the lines still wiggle by tens or hundreds of atoms. Now computer chips are made by eroding a solid chunk of silicon--it's still what we call a "bulk process" meaning that stuff is happening everywhere at once and it can't really be controlled precisely. If you built stuff instead by putting each atom where you want it, it will turn out perfectly precise. Straight lines won't wiggle by more than atom, because the atoms weren't placed incorrectly in the first place. And they won't wiggle by less than an atom, because there's no such thing as a fraction of an atom.
HOST: I've heard that things do wiggle--that heat makes them wiggle, and that this is a serious problem for nanotech.
CHRIS: Yes and no. Yes, things do move around because of heat. But this actually helps us. You see, if you build it right, they don't move enough to break or get seriously out of line. Helicopters have massive vibration problems, but we build them so they don't fall apart. But this vibration actually can reduce friction in moving parts. A dish can dance off a table during an earthquake. And you know how if something is stuck, once you get it moving it moves more easily. So the heat motion is constantly unsticking the moving parts, if you design them right, and believe it or not, it seems reasonable to talk about machines with zero static friction. This is using standard physics. So the devices we build could have extremely high efficiency. Also, they would never wear out, because the atoms would never get out of line enough to pull each other out of position--they'll just slip past each other without damaging anything. By the way, there's recently been an experiment that seems to confirm both of these claims.
HOST: That's pretty amazing. So you're saying we could have factories making copies of themselves, and running efficiently. How fast could a factory produce another factory?
CHRIS: Hold on to your seat. We don't know the exact answer. It will take a lot more design work. But a reasonable guess is probably an hour, more or less. This means that if you had one of these factories in your home, you could set it to producing copies to give to your neighbors, and you could just give them away.
HOST: Whoa. A factory inside a home?
CHRIS: Sure. Remember that a billion atoms fits along the edge of a yardstick. That's a single line of atoms. Put another line next to it, and you have another billion atoms. Do that a billion times. Now remember that all those atoms are used to make a sheet only one atom thick, and you have to make a billion more sheets to make a cubic yard of stuff. And within reason, you can choose which spaces to put an atom and which spaces to leave empty. So one cubic meter of nanotech-built stuff could be more complex than all the factories in the world today.
Let me give you another image. Iimagine this room packed floor to ceiling with machinery. Put some computers here, a metal shop in that corner, a chemistry lab over there, some generators to power it all... keep going until you've completely filled the room. Now imagine a six-story building packed floor to ceiling in every room with machinery. That would be a good-sized factory. Now put more buildings next to that to fill a city. Detroit. Now imagine the entire United States covered with six-story buildings full of machinery from sea to shining sea. Not a pretty picture, but that's OK, because we're going to shrink the machinery. Today's mechanical devices tend to be built with parts on the scale of millimeters--one thousandth of a meter. Nanodevices will be buildable with parts on the scale of a nanometer--a billionth of a meter, or a million times smaller. So we can fit the same functionality, if we build it with nanotech, into a space that's a million times smaller in every dimension. If you shrank that carpet of machinery by a million times, it would be about this size. (Produce a thin plastic dropcloth.) I bought this at a hardware store for a buck twenty-seven. It weighs less than a pound. If it were built with a nanotech factory, it could still cost a buck twenty-seven, but it could also contain all the machinery I just had you imagine. That's more than enough complexity to make a copy of itself. In fact, it seems likely that a diamondoid factory capable of copying itself could be as small as one of the cells of your body, which can also copy itself.
HOST: What could you do with all that?
CHRIS: Well, for one thing, if this plastic were packed with nanocomputers, it would have more computer power than the entire World Wide Web. Or imagine if it contained molecular machinery to collect sunlight and turn carbon dioxide back into oxygen. You could make a nifty space suit. And it would be more than strong enough because diamond is tens of times stronger than steel. Or you could have it turn carbon dioxide into fuel, store the fuel, and turn the fuel into electricity as needed. Totally self-contained solar power, with efficient energy storage to last for weeks of cloudy days. No more rolling blackouts. Or for medical applications--a little piece of this could synthesize insulin to cure diabetes, or make a scalpel blade that reconnected the skin after it passed through.
HOST: Sounds like PG&E would be out of business.
CHRIS: That's a good question. Who would be put out of business if we all had factories in our homes? We're seeing a little of that issue today, with Napster and Gnutella and other programs for sharing music files over the Internet. I'm sure that lots of commercial interests will want to restrict this technology. I think that would be a bad idea.
HOST: Why?
CHRIS: Two reasons. First of all, we have a technology that can make physical objects "too cheap to meter" the way computer data transmission is today. I know, "too cheap to meter" is a promise that's been broken too often. But computers have finally shown that it's possible. And nanotech will let us manipulate atoms the way a computer manipulates information. This could be very good for humanity, bringing an age of abundance and creativity to surpass the Renaissance. The second reason why restricting nanotech is a bad idea is that too much restriction will simply create a black market. Prohibition didn't work. The War on Drugs isn't working. And if a nanofactory too small to see can create millions of dollars of illicit wealth, there's no way we'll be able to suppress a black market. I think a black market would be very bad indeed.
HOST: And why is that?
CHRIS: Because if you have criminals with a strong cash flow and a nanofactory, they can hire technicians to make that factory produce unpleasant things like weapons. Today's guns are dangerous enough, and there was a huge stink when they started making plastic guns that were lighter and harder to detect. A diamond gun would be an almost flimsy thing that might not even look like a gun until it was fired. There are arguments that no technology can be completely suppressed, but I'd rather undermine the criminals as much as possible by reducing the incentive to own an illicit nanofactory.
HOST: Sounds like a tough choice.
CHRIS: There are many tough choices, and there's nothing new about that. And I haven't even mentioned warfare yet. Every department of defense will want to control this technology. An arms race would be very bad, and it could get even riskier than nuclear technology and the Cuban Missile Crisis. I hope not. Personally, I think the best way to avoid that is to have a central "pool" of technology that everyone can make use of, and have as much development as possible feed back into that pool. Of course governments will still want to have a few cards up their sleeves, and commercial organizations will want to keep making a profit. Somehow we need to find a balance, because I can't think of a single extreme position that looks like a good idea.
HOST: So you're saying this technology could be dangerous.
CHRIS: Every technology can be dangerous. Every technology has two sides. The same techniques that help infertile families have children can be used for human cloning. If we had a century to get used to nanotech, we'd probably adjust pretty well. As it is, we may have only a decade or two. So we have to make some decisions in advance, and we have to avoid extremists of any description running away with the technology. It just means we have to be more responsible about how we use it. But look at it another way. There are about fifty million people around the world dying each year. Most of them would rather not die yet. Most of them die from preventable causes, such as hunger, disease, and accidents. If you add up all the useful devices that nanotech makes possible, you could probably save forty million lives per year. So the risk of not using nanotech is forty million unnecessary deaths per year. That's a lot. We shouldn't let it stampede us into ignoring the consequences of nanotech, but we should keep it in mind. Also, don't forget the ecological problems we have that keep getting worse. Even farming creates major problems, including the land itself that is covered with crops. As manufacturing becomes more self-contained, and people's physical needs become more self-contained, we will live more and more lightly on the planet. Believe it or not, technologies have been getting cleaner all along. Oil is bad, but it's much better than coal. Nanotech is the ultimate in clean. Nanotech manufacturing allows zero industrial waste and 100% recycling. That offsets a whole lot of risk. Imagine no more smog! By the way, that will also offset any temporary population increase caused by fewer unnecessary deaths.
HOST: Sounds like we have some exciting times ahead. [wrap up]