The miniaturization of technology is slowly turning from promise into reality. The success of this trend has spawned perceptual difficulties for consumers: a few years ago it was still possible to impress a datacentre manager with the size and robustness of a Sun workstation. Today, an iPOD has more RAM, processor power and hard disk space than a server used to have in the year 2000.

Miniaturization also led to specialization. Gone are the days when the promise of a universal machine imitating all possible machines reminiscent of Turing's ideas led to computer architectures that literally could do everything. Dynabooks at Xerox park, multimedia-enabled Macs and overpowered PCs somehow, well, seem so 2002 cum 1976.

From Ephemeral Bits to Hard Atoms
There are clouds on the horizon and the MIT Media Lab, the other Negroponte mountain that the technoprophets moved to, has a center for bits and atoms. The soft, squishy media stuff used to grow on electronic designs that, let's face it, were hardly more than souped-up 70s technology. New devices like sensor networks, micro-machinery, nanotubes and exotic materials programmable on molecular level change the way in which engineering and manufacturing is going to be conducted.

Today, applications of nanotechnology still look slightly more pedestrian: nanoceramic materials are extending the life of batteries; IBM research scientists have successfully experimented with nanoscale oxidation to produce prototypes of new circuits and devices.

The Greeks Knew Something, too
The Presocratics had similar dreams: the 5th BC natural philosopher Anaxagoras said that "Mind is the finest of all things and the purest, it has all knowledge about everything and the greatest power; and Mind controls all things, both the greater and the smaller,... .And all things that were to be -...Mind arranged them all". To him, mind and matter were corporeal, embodied in tiny specks of material that in turn assembled bodies large and small. The concept of indivisible matter, i.e. atoms, was invented only a few years later by Leucippos and elaborated on by Democritus. Only Mind embodied in matter was able to assemble compound objects out of tiny specks of indivisible micro-spheres.

The manipulation of matter, i.e. molecules and atoms, to form new materials, machines, or indeed new forms of life appeared to be the business of the gods or indeed of a single supreme being. Unfortunately, God as a designer and engineer turned out to be a scientific hypothesis of limited usefulness since provided His existence could be ascertained, he forgot to leave blueprints for us to imitate his handiwork. Even if theoretical physics and the beginnings of chemistry, the latter very much a 16th century Chinese discovery, had been more up to the task, micromachinery stayed within the realm of Byzantine 14th century marvels and European Renaissance goldsmiths.

Periodic Tables, Monsieur
The very existence of atoms and their composites, lest we forget, was very much subject to severe doubts until Lavoisier's periodic table. Precision manufacturing only began its trend towards every smaller things once it had broken out of the confines of the medieval guilds and found itself quite literally empowered by reliable steam engines and other movable feasts. The trend towards analyzing the smallest units of matter did not even quite start once microscopes began to give medical doctors a basis for their extremely speculative etiologies. Bacteria and viruses were viewed as menaces, not as marvels of natural, if accidental, design.

Feynman's Charm Again
The 20th century American physicist Richard Feynman asked in 1959 how it might be possible to exploit the room at the bottom of the engineering scale, to whit, how it was possible to create wires, spaces, transistors and pins less than 100 atoms wide. Electronic microscopes at the time performed far above this scale - in the region of several thousand atoms or tens of angstroms of wavelength. Much of his lecture published a few months later reads like a laundry list of technological achievements that seem to be within our grasp now. Writing the Encyclopaedia Britannica on the head of a pin was the challenge that entered public awareness quickly. Others included the mechanism driving mutation, a task he proposed to resolve by looking the molecules directly using some as yet undreamt-of super-electron microscope.

The miniaturization of computer processors makes a proud appearance, a challenge that has obviously been met. Yet Richard Feynman's musings went hardly beyond the intellectual level of the Presocratics. Granted, he was a theoretical physicist much at home with the mathematics and experimentation that had turned quantum mechanics and quantum electrodynamics into fields studied by undergraduates all over the world. Claiming that "we can arrange atoms the way we want: the very atoms, all the way down" did not mean that scientists should go forth and push atoms on sub-microscopic pool tables. It was inspired scientific propaganda, but it had little foundation in the reality of physics research at the time.

Feynman's hostility to philosophy was well known; and I assume he would have appreciated the irony of having his lectures compared to the rough and ready speculations of thinkers who lived just after the threshold beyond which cosmogonic thinking and poetic license became slightly less acceptable than the philosophical constraints imposed by logic, ethics and natural philosophy. But he was aware of the impact quantum-level forces were likely to have on nano-size devices: he knew it would be difficult to deal with heat and resistance in compact solid state devices that were likely to replace transistors visible to the naked eye.

Nanotechnology Arises
The very concept of nanotechnology made a rather sudden appearance at a time when the world was in the throes of the First Oil Crisis and the end of the Vietnam War. In 1974, Norio Taniguchi, a Tokyo Science University professor had this to say in a peer-reviewed scientific paper: "'Nano-technology' is the production technology to get the extra high accuracy and ultra fine dimensions, i.e., the preciseness and fineness of the order of 1 nm (nanometer), 10 -9 m in length. The name of 'Nano-technology' originates from this nanometer. In the processing of materials, the smallest bit size of stock removal, accretion, or flow of materials is probably of one atom or one molecule, namely 0.1~0.2 nm in length."

Drexler the Prophet?
This definition somewhat dryly reduces the lure and promise of nanotechnology to just the size and not the function, which it is supposed to fulfill. We might find the definition disappointing, but as Karl E. Drexler was to demonstrate to an enthusiastic audience in 1986, the ability to move atoms from place to place would have dramatic implications for the way in which industrial manufacturing and our daily lives are likely to be conducted.

It isn't only the nano-scale sciences between quantum mechanics up to molecular biology that came into view over the last few decades. It was rather more specific ideas that impressed Karl E. Drexler so much that he felt he had to dedicate his whole life to them. Drawing on the rich heritage left behind by John von Neumann's theory of self-reproducing automata, Drexler asked whether the model of cellular reproduction found when the relationship between DNA and RNA had been disentangled would yield algorithms and even material lessons to be adapted to a still-to-be-invented craft of molecular engineering. He published the rabble-rousing call to arms, "Engines of Creation", baffling many in the not-yet techno-utopian community of physicists and computer engineers how to build a new discipline with such earth-shattering consequences.

In essence, Drexler was asking whether it would be possible to build molecular assemblers able to produce any molecular design engineers care to develop. The extremely small parts required for this type of assembly would draw on so-called replicators, little virus-sized machines whose copying mechanisms show more than a passing resemblance to Turing machines.

Of Stonecrunchers and Miniature Masons
To crunch raw materials into molecule-sized chunks, disassemblers would work their way through rock, waste materials and other insalubrious matter feeding assemblers of varying sizes; disassemblers and their counterparts would reside in linked hierarchies of nano-machines. Drexlers waxed almost lyrically about the possibilities, founding the Foresight Institute in the process and marrying the inventor of the concept of Open Source, Christine Peterson, herself a technology guru of international stature. Nanotechnology and the open source movement share common ground, since both need to keep their "code" open, although the programming of assemblers would reset the structure of molecules and assemble ever greater, and owing to its error-correcting abilities, increasingly correct and more stable macro structures. Programming the very structure of matter itself turned out to be possible. Of which more anon.

Gravy Turns into Goo
But even before Drexler's paean to the green fields of nano-engineering, somewhat dystopian ideas had eaten their way into the Darwinian world of the sciences of the very small. Greg Bear, a Science Fiction author with some biological expertise and almost as much imagination as Drexler, wrote his first novel on the replacement of all biological life with an alien life form that showed properties strongly resembling universal assemblers. Greg Bear added the rather uncomfortable idea that they would mimic human beings, geological structures and suchlike, and they would show mindlike properties while ruthlessly replacing the very surface structure of Earth itself.

From Drexler to Doomster
The idea of Grey Goo was born. Drexler, a Californian born and bred, had been working at MIT on nanotech ideas since 1977, but science fiction writers were more willing to pick up on the doomsday aspect of a nanobot or nanite plague than on the then ill-understood potential benefits. Intriguingly, Drexler himself had warned of the same scenario rather early. Grey Goo, the idea of exponentially multiplying nanoreplicators running amok like a non-biological cancer replacing "healthy" rock and organic materials resonated with some of mankind's oldest fears: that of a molecular-sized, unseen monster, a plague created by man. Bill Joy of BSD and Sun Microsystems fame turned Cassandra and warned that the technology should remain a dream, since realization would almost certainly lead to military applications that would be difficult to contain.

Neumann's Paradise
John von Neumann's idea of self-reproducing cellular automata provided mathematical formalisms and a few handy metaphors to give ideas around nanotechnology wings. They were cellular not in the biological sense, of course, but referred to the fact that a horizonless grid was covered with an organism consisting of about 200,000 cells. The description and actions of the organism was given by adding a state to each of the 200,000 cells, a fairly complex arrangement, since 29 states were possible for each cell.

The organism, 400 cells long and 80 cells wide, contained three units: the factory, the duplicator and a computer. The factory gathers raw materials from its environment to arrange them in a new pattern; the duplicator reads instruction and copies them and the computer controls all actions within the organism. The rest of the cells consist of the blueprint for the entire organism so it can be read as preparation for reproduction.

This organism worked as a giant automaton encompassing 200,000 finite state machines and the rule applied to the whole machine changes not only the internal state of the organism, but also changes the environment. The transition rules, which naturally are far more complex than that of a standard Turing machine, enabled the machine to build a new organism and add all the information contained in the new blueprint. The infinite grid now contains two new self-reproducing cellular automatons. And so on and so forth ad infinitum.

The constructor contained in the factory had to be told explicitly what to do with the new raw materials, i.e. the cells. Drexler's nano-replicators and assemblers were based on similar principles. Neumann finished his work in the late 1940s, but publication only took place in 1966 and the book on self-reproducing automata was only made possible by the additional work accomplished by Arthur Burks. Contrary to a well-established rumour, Neumann's universal constructors were not able to construct everything. Neither are Drexler's nanobots. But for all intents and purposes, they would be amazingly versatile.

So one would assume that our little non-furry molecular friends might be the ultimate goal of any chemical engineer and computational physicist on this planet. But not only are there considerable practical problems, there are also competing approaches and some ideas that sound even more outlandish, but turn out to eminently practical.

Programmable Matter
Wil McCarthy, who combines the enviable talents of a novelist, entrepeneur and physicist, dreamt the Democritean dream of matter to be arranged in ways its unchangeable and indivisible nature would not have allowed, had 5th century Atomists had their way. Molecular machines seemed almost quaint when one considers that it is possible to change the properties of electron clouds around atomic nuclei. Programmable matter or quantum dots - two concepts used interchangeably these days - have some very useful properties. Essentially, programmable matter can be created by taking P-N-P junctions inside semiconductors and make the N layer so thin that is consists of little more than electrons behaving like a standing electron wave in keeping with wave-particle dualism: if the N layer is less than 10 atoms thick, it is not likely that there are any atoms left for electrons present in both P layers to crowd around. What we have is electron orbitals without nuclear centres.

It is Matter, Jim, But not as We Know it
Electrons, as we know, determine the chemical behaviour of atoms. Rather weirdly, we have created artificial atoms without a core. Given that they behave like a standing electron or Broglie wave within the P-N-P semiconductors, we have created those quantum dots that mystified us earlier.

Since the number of electrons determines the way in which the atom is situated in the periodic table, we can increase and decrease the number of electrons without a nucleus in the quantum well by adding and decreasing the voltage at the P-N-P boundaries. It doesn't actually matter how the electrons are confined. What matters is that quantum dots a few nanometers across can be grown together in so-called nanocrystals whose blinding speed and efficiency is finding its way into new types of hard drives and extremely small transistors.

Alchemical dreams and weird materials have become reality: turning iron into gold does not seem impossible anymore. This of course is just the way we turn the reality of one species of matter into the appearance of another, producing strange materials with, say, fluorescent properties. Atoms with thousands of electrons can be simulated, and the properties, which do not occur in nature since the periodic table only allows for 92, are decidedly strange.

Anyone who has managed to follow this article so far, is probably wondering what Drexler's molecular assemblers have to do with artificial atoms. On the face of it, we can't postulate much more than the size of atoms and molecules that are being manipulated and programmed to fulfill purposes defined by man. Molecular assemblers, if they ever become fully-fledged reality would be wonderful tools to have, but programmable matter is being experimented with now and has found applications in harddrives and other places. Molecular assemblers as a concept are also being questioned; other approaches have been tried to the same problems and some results have been obtained.

From Nanodreams to Real Tech
Nano-scale manipulations have been only been done fairly recently. When the Swiss IBM researchers Heinrich Rohrer and Gerd Binnig started tracking changes in the electrical current on sharp nano-scale diamond tips in 1982, the scanning tunneling microscope (STM) and therefore the first nano-scale technology had been born. The step from seeing atoms to manipulating them is not quite as far as it might seem, although Gerd Binnig in Zrich and Calvin Quate at Stanford had to generalize their approach to non-conducting surfaces first. The chemical signature of individual elements represents enough information for a computer to build an image of the imaged surface on an atomic scale.

5 years later, Erhard Schweizer at IBM's center in Almaden succeeded in spelling out "IBM" using 35 xenon atoms on a nickel surface. It was the first time that Feynman's and Drexler's vision of nano-scale manipulation had been realized. But there is a caveat that is continuing to cause problems: to this day we have fairly little idea how we would go from what essentially amounts to "hand"-crafting individual nano-scale sheets of atoms to mass-producing functioning nano-scale devices.

There are early applications, however, that have found their way into the practice of researchers and industry. Nano-crystallites can be grown from cadmium-selenite semiconducting materials. They shine in every so many pretty colors whenever they change size. They have held triumphant entry as dies in biology experiments and their commercial potential has not gone unnoticed.

London Tube Miniatures
A rather more eye-catching and thought-provoking discovery were carbon nanotubes, in essence extremely stable, conducting flask-shaped molecules whose geometrical characteristic determine their behavior: they conduct either like metals or semiconductors. It seems that they will be changing diode and transistor production. But the nanotechnology community seems profoundly skeptical of Drexler's claims that the next step beyond using nano-particles and interesting molecules in the traditional sense to true nanotechological devices lies just around the corner.

On the one hand it is not easy to demonstrate that nanotechnological devices actually work. It has been done: transistors using carbon nanotubes have been researched and made to work since the late 1990s. On the other hand, the holy grail of nano-tech devices mass production has not yet been revealed to any White Knight in the industry. There is no doubt that the billion dollars or euros of taxpayers' money invested in basic research have yielded fairly impressive results. But given the high expectations and the pitfalls of interdisciplinary cooperation, we cannot expect a huge number of results any time soon.

We are not yet facing an all-encompassing manufacturing revolution, but to make nanotechnology an everyday occurrence, one is needed within the next few years. Nanotechnological devices in medicine and biological research are emerging slowly, and the application of nano-materials in the semiconductor industry and in medical sensors are not a new phenomenon. But we are facing, if not doomsday caused by gazillions of nanobots, the slow evolution of artificial devices. Democrit's random atomic collisions have been mastered and they have begun to yield deliberate molecular design.

Postscript
The 5th century Presocratic philosopher Democritus is credited with the elaboration of philosophical atomism. His somewhat older contemporary Leucippus is traditionally assumed to be the inventor of philosophical atomism, although it is next to impossible to ascertain which part of Democritus' philosophical concerns were due to discussions with his predecessor and what parts were developed by Democritus in a more independent fashion.

Democritus believed that the universe is infinitely large and contains an infinite number of extremely small, invisible particles. He did not think it possible to divide the ultimate parts of the universe any further, thereby leading to the property of being indivisible (atomos). His attempt was the first to suggest that the universe was non-continuous, i.e. that the universe consisted of discrete parts. Digitalization is an extremely old idea, as we can see.

Although one of the most prolific authors of antiquity, he suffered the indignity of having none of his books transmitted to modernity. Aristotle's work and the 6th century philosopher Simplicius' commentaries largely contained what we know about his natural philosophy.

Modern Atomism was motivated by modern chemistry and an experiment conducted in a controllable environment; intellectual history provides indirect connections at best. But more importantly, Democritus stipulated that atoms are by no means identical to each other, being distinguished by shape, arrangement and position. Bigger objects are composed from smaller ones, thereby scaling from single atoms to suns and planets.

There was no explanation for the composition of bodies except through random attachments, apart from Democritus' contemporary, the Athenian philosopher Anaxagoras, who speculated that matter was capable of organization. This made it intelligible and therefore provided one of the conditions for emergence of intelligence. It is interesting to see that all the themes of nanotechnology were already present, but there was no attempt to restate the problems of atomism and attempt to explain it as the result of deliberate design. The very notion of machines doing mankind's work was in its infancy at the time and the lure of the extremely small had not begun to invade the imagination of scientists and philosophers.

Frank Pohlmann

See also: Little secrets - How not to launch a new technology

References
General Nanotechnology
www.foresight.org
www.nano.org.uk
www.zyvex.com/nano

Democritus
www.seop.leeds.ac.uk/entries/democritus
www.seop.leeds.ac.uk/entries/lucretius

Richard Feynman's talk on nanotechnology
www.zyvex.com/nanotech/feynman.html

Karl E. Drexler's Engines of Creation
www.foresight.org/EOC

Programmable Matter
www.nanotech-now.com/Wil-McCarthy-interview-06132003.htm
www.sciencebar.com/pmfaq.htm

Christine Peterson's role in Open Source
www.oreilly.com/openbook/freedom/ch11.html

John von Neumann's Theory of Self-Reproducing Automata
www.zyvex.com/nanotech/vonNeumann.html