Armageddon Science (19 page)

All these simple nanostructures do have the potential to harm human beings. Because of the physics of the very small, the tiny particles or structures can react in unexpected ways—for instance, penetrating the skin and slipping through biological membranes that normally keep foreign matter out, or by being carried into the lungs to cause damage. We have seen in the past how much damage asbestos fibers—once treated as if they were harmless—can do. It seems only reasonable to take care when dealing with nanoparticles and other nanostructures.

Some of the early reactions to nanoparticles have, however, demonstrated massive ignorance. In January 2008, the Soil Association, the biggest organic certification body in the United Kingdom, banned nanoparticles from organic products. But in doing so, the Soil Association specifically banned only
man-made
nanoparticles, claiming that natural ones (like soot) are fine because “life has evolved with these.”

This totally misunderstands the threat we face from a nanoparticle. A nanoparticle is most likely to be dangerous because of its scale, because the physics (rather than chemistry) of particles of this size is quite different from that of the objects we are familiar with. Where this is the case, that danger is just as present whether the particle is natural or it isn’t. Even where scale isn’t the only risk factor, natural nanoparticles can be dangerous because of the way they act or their ability to interact with the body. Viruses are natural nanoparticles, and like soot, they aren’t ideal for the health.

The Soil Association defends its position by saying that its approach parallels a sensible attitude to carbon dioxide in the air, where there is no problem with the natural carbon dioxide, only the man-made contribution. This is a specious argument, both because carbon dioxide is carbon dioxide, and if levels are too high it doesn’t matter where the molecules are coming from, and because there is no comparison between CO
₂
and a nanoparticle that could be directly physically dangerous to humans.

To make matters worse, the Soil Association also says that it can’t control natural nanoparticles present in the environment. They’re just there. However, this is not relevant—the Soil Association isn’t an environmental control group. Its role is to control what goes into organic products, and there is nothing to stop a manufacturer from putting natural nanoparticles into a product either by accident or intentionally. You might as well say we don’t mind a manufacturer putting salmonella into organic food, because it’s natural. If the Soil Association believes nanoparticles are a bad thing, it should ban all nanoparticles from a product that gets its seal of approval, not just artificial ones.

It is when they try to summarize their argument that the Soil Association lets slip the reason it takes this strange attitude. “The organic movement nearly always takes a principles-based regulatory approach, rather than a case-by-case approach based on scientific information.” In other words, theirs is a knee-jerk reaction to concepts, rather than one based on genuine concerns about the dangers of nanoparticles. It is all about words like “natural” and “artificial,” not about the nanoparticles themselves. In practice, we need to be concerned about all types of nanoparticles.

There are genuine and sensible worries about potential health risks from nanoparticles. As we have seen, because of their unique physics, they can penetrate barriers that stop other contaminants, whether these are natural barriers like skin or simple breathing masks. We do know it makes sense to minimize our exposure to breathing nanoparticles, and we should make sure there is long-term testing of the effects of any nanoparticles in substances we apply to our skins or ingest. But it seems unlikely that nanoparticles could cause worldwide devastation.

Moving up a step of complexity, we get to nanoscale machines, the sort of mechanical “hands” on the scale of molecules that Feynman described. Much of the work that has happened so far in this field has come not out of engineering shops as Feynman envisaged, but biology labs. There’s a good reason for this: because the biological world, from the complex chemicals that operate within our bodies to stand-alone nanoscale entities like viruses, is replete with molecular machines operating at just the level we are considering.

Take proteins, the workhorse molecules that carry out the instructions of DNA in living cells. Proteins don’t just carry signals by plugging into other molecules, or act as reinforcements in cell-based structures like cartilage. What are often long, string-like molecules fold into specific shapes—and the way a protein folds will determine how it then acts. It is arguable that because of the mechanical action of folding, such proteins are the simplest form of biological machines. We are used to thinking of machines as complex devices like a car or an iPod—but we should remember that basic devices like the lever, the pulley, and the screw are all machines in the technical sense.

Whether or not you class a simple folding protein as a machine, there are certainly machine functions in living creatures that are powered at the level of nanotechnology, whether it’s the kind of ratchetlike “grab and pull” that proteins undertake in a muscle, multiplied millions of times to make your legs or arms move, or a truly complex machine like the flagellum found on some bacteria. These microscopic propellers, beloved of the supporters of intelligent design, have an ion-powered motor and a rotary socket. However, it’s hard to see how these particular machines can cause real damage. They don’t represent the kind of threat we are looking for in a nanoscale Armageddon.

For real devastation, we probably need to be looking at nanoscale robots—nanobots—which remain the ultimate aim of many nanotechnologists. Thinking back to Richard Feynman’s vision, to be able to manufacture items atom by atom, we had three issues: mapping an object to know what to build, being able to work on enough atoms at a time, and being able to manipulate individual atoms. This last problem would require as a starting point special nanoscale machines—assemblers. Exactly how these would work is not clear, but they would effectively be nanobots, invisibly small robots whose sole role in life would be to take atoms and reassemble them in a new form.

To make this possible, we would need to manufacture these nanobots and to power them, and we would have to be able to give them the instructions they need to make whatever we require. While Feynman’s “small hands, making smaller hands, making smaller hands” could be seen as a way to kick-start the process, inevitably the only practical way to make nanobots would be to have them capable of making themselves. That’s because we would need not just a few million nanobot assemblers, but countless billions.

Remember the numbers for assembling an object around the size and weight of a human being. We would need 300 trillion such assemblers to achieve the task in a year at a rate of a million actions per assembler per second. We could achieve this only if, in effect, assemblers could breed.

Assuming for a moment that we had these self-replicating nanobots, they wouldn’t have to turn into the sort of voracious monsters portrayed in Michael Crichton’s book in order to become a threat to humanity. One possibility is that humans would become so dependent on nanobots that, should the technology fail, the human race would be doomed.

It’s just possible this would happen if all manufacturing were replaced by assembler production lines. Imagine products being assembled molecule by molecule so anything and everything could be constructed by a single device. If such technology were perfected and cheap, it’s hard to see why any conventional form of manufacturing would remain in use. Then, once we had become totally dependent on them, the failure of nanobots would cripple our technology-dependent society.

The impact on our lives could be much greater than just losing manufactured goods, many of which are luxuries we don’t need to survive. Imagine a food-production device like those on the TV show
Star Trek: The Next Generation,
where food is assembled to order. If all our food were assembled instead of grown, then the world would face starvation if nanobots stopped working. But there is another possible way to become totally dependent on the technology. We could get to a stage where our bodies needed nanotechnology to survive from moment to moment.

Those like Ray Kurzweil who imagine a future where human beings can effectively live forever believe that we will reach a point where we inject nanobots into our bloodstreams to fix our cells, and to replace entirely many of the functions of fallible human organs like taking over pumping blood from the heart. In principle an assembler can make anything, unstitching atom from atom and reconstructing the building blocks of nature into anything from a scarf to a TV to a piece of beef. There’s no reason why its role shouldn’t include making (or acting as) more efficient parts of a human body. If we became totally dependent on these nanobots to stay alive, then any large-scale failure of the technology could spell disaster for the human race.

One way this could happen is if nanotechnology suffered a failure that paralleled the problems affecting the natural nanomachines in our bodies. If we are to achieve Kurzweil’s dream of tiny intelligent machines inside our bodies that keep us alive, we probably have to go some way down the assembler route. The only sensible way to build such nanobots is to use other nanomachines. We would then be susceptible to failures in the nanobots that resemble the biological problems that plague real living, reproducing creatures.

It’s the replication process that produces a threat. As we already know, things can go wrong with replication in the natural world—when mutation occurs. If something is being replicated many, many times, there is a chance of an error in the copying. Usually that error will result in failure, but occasionally it can make a change that will make the replicating creature better.

Natural selection will ensure that the “better” form of the creature thrives, assuming it can pass on its difference, and eventually it will take over from the earlier version that lacks the enhancement. That’s evolution in a nutshell. It happens with the biological machines that populate the world, and it could happen to nanomachines. Once machines have the ability to replicate, and to pass on changes in design, they can evolve, or pass on a fatal flaw to the whole population.

We’ll come back to what might go wrong due to mutation when we look at gray goo, but before that we need to consider a different kind of failure. Perhaps the biggest danger facing a world dependent on nanotechnology would come not from the accidental evolution of a nanobot, but from the intentional handiwork of a hacker: a nanotechnology virus.

Every day in my e-mail inbox, I receive, besides the genuine e-mails I want, a whole host of others I don’t want. There is spam, trying to sell me Viagra or encourage me to provide my bank details so I can receive a huge bequest. But worse still, there are e-mails carrying viruses, trojans, and worms, all hoping to take over my PC or cause damage. Some virus writers produce them for fun or as an intellectual challenge, but others are, in effect, electronic terrorists who hope to cause disruption and confusion. The result of this relentless impact from would-be attackers is that I have to have three programs running all the time—antivirus, a firewall, and antispyware, constantly battling to protect my computer.

These electronic vandals don’t stick to a single technology. As long as it’s widespread enough to be worth their attention (the reason Apple computers are relatively unscathed), they will get involved. Now that many cell phones are powerful pocket computers, virus writers have spread their attentions to this technology. In June 2004, the first cell phone virus emerged into the wild. One of the particularly unnerving things about cell phone viruses is that they behave in a way that’s more like the real thing than anything that arrives on your computer.

The use of the term “virus” when referring to a malicious computer program has always had the potential to cause confusion. When the public first became aware of computer viruses, I was running the PC department of a large company, and once had a phone call from a worried executive who had recently become pregnant. She was worried about catching the computer virus and the danger of it causing damage to her unborn child. Cell phone viruses don’t put humans at direct risk any more than computer viruses, but they certainly can hit our wallets, and they spread in a worryingly natural manner.

Here’s a typical scenario featuring an attack by the phone worm Commwarrior. You are in a bar and need to make an urgent call. Your cell phone beeps—it asks if you want to accept a Bluetooth connection from someone you don’t know. Sensibly, you click No, as you don’t want to connect to a stranger. But before you can do anything else, up comes the request again. And again.

It keeps coming so fast that you can’t place your call. So you finally say Yes just to get it out of the way—and with that Yes, your phone is infected. The Commwarrior virus has jumped from a stranger’s phone to yours. Because of the way Bluetooth works, cell phone viruses and worms that use it jump from phone to phone when they are in close proximity. Your cell phone literally catches a bug like this by being near a phone that is infected. And once your phone is infected, the virus has the potential to start siphoning cash from your account.

It would be naïve to think that hackers who can jump on the bandwagon so effectively with smart phones wouldn’t try to do the same with nanomachines. If anything is going to give us cause to pause and think whether or not we want to go down this route, it is the possibility of hacking. Yet it’s reasonably easy for this to be avoided. Computer viruses, whether on a PC or on a phone, are just programs. It’s entirely possible to make intelligent electronic devices that can’t have a program run on them other than the one that is built in.

We normally allow reprogramming of electronic equipment because software often needs updating and we don’t want to throw the hardware away if the program is wrong. But nanobots exist in a different kind of world. They don’t need to be reprogrammed—we can literally rebuild them molecule by molecule instead. Any programming is hardwired in the atoms. This is an antihacking advantage, and it makes the software less complicated.