Home > Nanotechnology Columns > Center for Responsible Nanotechnology > Nanotechnology: Journey vs. Destination
Nanotechnology has acquired several distinct meanings over the last few decades. Its development has been marked by this confusion, which has led to concerns from one field of nanotechnology, molecular manufacturing, being applied to other fields. As all fields of nanotechnology continue to develop, molecular manufacturing will reach a point where it is able to accelerate the other fields.
September 28th, 2007
Nanotechnology: Journey vs. Destination
Nanotechnology has been a buzzword for many years, but it still doesn't have a real meaning. Or, more precisely, it has several meanings. The story of how nanotechnology developed is interesting; even more interesting is the story of where it's going.
The word "nanotechnology" was introduced to the public by Eric Drexler's book, Engines of Creation, in 1986. For some time before then, concepts and tools related to nanotechnology had been developing. As early as 1959, Richard Feynman had stated that machines could be built out of molecules. Feynman also asserted that a billion tiny machines could be controlled in parallel, pointing out that they could be built with perfect precision since atoms are discrete identical components. Scientists had been working with nanoscale tools like electron microscopes and nanoscale techniques such as colloid chemistry. But these were seen, correctly, as extensions of existing capabilities: noteworthy and useful, but not likely to lead to technological revolutions.
Drexler had published scientific papers prior to 1986, including a paper on protein engineering in the Proceedings of the National Academy of Science. But Engines was what brought public attention to his ideas. In that book, Drexler asserted that tiny machines, processing individual molecules under computer control, could build human-scale products. And not just any products, but highly advanced products made of diamond: rocket engines, supercomputers, medical devices, and weapons could all be built by a general-purpose manufacturing system made up of zillions of tiny cooperating molecular-manufacturing robots. Download the blueprints, and the robots would build a shimmering diamond product from inexpensive feedstock in just a few hours.
Naturally, Drexler's ideas fired the public imagination. His book, non-fiction but speculative, inspired numerous works of science fiction. He co-founded an organization, the Foresight Institute, to study the implications of such a powerful technology. In 1992, Drexler published a technical book, Nanosystems, that analyzed in detail the physics of molecular manufacturing. He and others continued publishing technical and scientific papers to flesh out the idea.
As the word "nanotechnology" continued to spread, its association with nanoscale wonders and powers became more nebulous but no less exciting. Scientists were starting to access the nanoscale in unprecedented detail, thanks to new technologies like the scanning tunneling microscope that let them detect and even move individual atoms. And those scientists were starting to report back that the nanoscale had all sorts of weird and wonderful capabilities. Small bits of matter could behave very differently than their larger counterparts: they could glow, conduct electricity, and even take on medicinal properties. "Nanotechnology" thus came to mean the study of any new nanoscale phenomenon.
Meanwhile, opposition to the original idea of molecular manufacturing was growing. In part, this was because of other ideas that had become associated with it. For example, technically grounded estimations of the computers that a molecular manufacturing system could build turned into speculations about radical artificial intelligence and simulated humans living inside computers. Another aspect of the opposition was that the projections had far outstripped the laboratory demonstrations. Finally, there was a large dollop of ordinary human "Not Invented Here" politics: many scientists of the nanoscale dismissed molecular manufacturing without understanding it.
Nanoscale technology got a major boost from the US government's decision, around the turn of the century, to spend a billion dollars a year on developing it. The National Nanotechnology Initiative (NNI) was formed to distribute this money. They chose a rather broad definition of nanotechnology: basically, building any new and interesting structure smaller than 100 nanometers in any dimension could be funded. Since the semiconductor industry was about to start doing just that, this definition guaranteed them at least one major success; on the other hand, with a narrower definition they might have had serious trouble spending a sudden $1 billion influx.
Unfortunately, just as the NNI was getting started, an old and obsolete idea from molecular manufacturing floated up to scare people. Back in 1986, Drexler's ideas of manufacturing systems were informed by biology: he had pictured robots not completely unlike programmable bacteria, floating in a nutrient bath, moving and cooperating to build large products. On the theory that any self-replicator is potentially dangerous, Drexler had warned that small self-contained self-replicators, if sufficiently flexible and "tough," might outcompete natural bacteria and turn the biosphere into inedible "grey goo." By 1992, Drexler had re-thought his approach, designing what what we now call a "nanofactory -- a more efficient and more controlled system with no free-floating components. But in 2000, Bill Joy resurrected the 1986 worries, writing in Wired magazine that one "oops" in a nanotechnology laboratory could destroy the planet. This provided a strong incentive for nanotechnologists, already skeptical of molecular manufacturing, to assert loudly that the entire field was impossible.
In recent years, concern about grey goo has faded, thanks in part to public statements by Drexler and CRN (among others) that molecular manufacturing development will not lead to accidental grey goo. Technical demonstrations of basic molecular manufacturing techniques have been performed in the laboratory. The scientists most strongly opposed have largely fallen silent, and many mainstream researchers are now willing to admit that the ideas may have technical merit and may even be feasible. Debate continues as to whether the "hard" mechanistic approach of molecular manufacturing will actually be better than a "soft" biological approach, but the debate is becoming less political and more scientific.
In the near future, nanoscale technology researchers will continue to develop impressive capabilities based on building small objects. But the kind of limited nanomanufacturing being developed today is just that: limited. It will be difficult to combine more than a few features or functions in any given nano-component, and it will be difficult to interface nano-components to the macro-scale world.
Molecular manufacturing, meanwhile, continues to advance. Doing chemistry with scanning probe microscopes is now well-established as a possibility. Simulation investigations of mechanical chemical reactions are improving in detail and reliability; recently, Ralph Merkle and Robert Freitas announced a complete set of reactions for building diamond from acetylene. Laboratory tools will continue to become better and cheaper; Freitas is spearheading a collaboration of scanning probe researchers to demonstrate these reactions in practice. Projected costs of a ten-year development program ending in a desktop molecular manufacturing capability has been falling, by some estimates, by about a factor of ten every decade. By the early 2020's it may cost only a few million dollars. At that point, if not well before, someone surely will have launched a targeted development program.
Once a programmable molecular manufacturing capability exists, it will accelerate other nanoscale technologies. Building programmable structures at nanometer scales, at near-zero cost, in hours or perhaps even minutes, will be a highly useful capability for laboratory research as well as product development. Meanwhile, a rather basic level of capability will allow tiny fabrication systems to build larger parallel systems of fabrication systems, doubling manufacturing capacity every few hours. With competent pre-design and well-architected software, it should be possible to scale up from nanogram to ton-scale manufacturing in a very few years or perhaps even months. Even planet-scale engineering appears possible, which provides some hope of dealing with the planet-scale effects of climate change -- and perhaps even undoing some of its causes.
Rapid prototyping and massively parallel construction should work to accelerate other fields as well. Whole-body real-time sensing should allow medical treatment that is both more aggressive and safer than today's techniques. In fields where experimentation is limited by the high cost of failure of expensive prototype articles like space ships, rapid, fully automated construction from inexpensive materials should allow a whole new approach to R&D. Inexpensive and efficient supercomputers, built at the molecular scale, should accelerate progress in many areas. The first nanofactories may be limited to a few materials (though building many structures and functions with those few materials); however, we may expect progress in mechanical chemistry to be rapid, so that within a few years, nanofactories will be able to build an even wider range of products.
In short, although a wide range of nanoscale technologies have been competing for the nanotechnology mindshare with molecular manufacturing, the latter will continue to advance, and before long may be the most influential of all the things that are called nanotechnology.