Before molecular biology, such speculations seemed wild and unfounded, but Feynman is once again proving to have been correct. The detailed study of the structures of the cell revealed that nature had engineered machinery from the insubstantial substance of a few atoms strung together; the search for a "vital force" only revealed a bewildering array of mechanisms ­- enzymes, ribosomes and other tiny structures ­- which demystified the cell even as it revealed the incredible versatility of atomic-scale chemistry.

Feynman postulated that once the tidy language of atoms had been decoded, it would be possible to engineer molecules precisely, placing one atom against another to create the smallest possible artifacts. What kinds of tools might we create with these ultra-miniaturized forms? Feynman imagined a molecular "doctor" that would be hundreds of times smaller than an individual cell. It could be injected into a human body and go to work, reading the health of cells, making repairs, and generally keeping the body in perfect health.

Science fiction, his peers said. Absolute fantasy tossed off by the master storyteller of physics. During the heights of the Industrial Age, "big" carried an importance of its own ­- big science, big engineering projects, big dreams. Even computers, in the 1950s, consumed whole floors of buildings. But even as Feynman made his address, engineers at Texas Instruments put the finishing touches on the first integrated circuits, and the world began to grow small.

MARVIN MINSKY, THE FOUNDING FATHER of Artificial Intelligence, possessed a mind fertile enough to nourish Feynman's dream. Throughout the 1960s and 70s, Minsky lead the world in future thinking. He consulted with Stanley Kubrick and Arthur C. Clarke on the cinematic reality of HAL 9000, directing a small squadron of graduate students into the emerging fields of machine intelligence, and always speaking broadly about the nascent possibilities of tiny technologies. By the mid-70s, when a graduate student named K. Eric Drexler came to Minsky seeking a sponsor for his master's thesis, Minsky's word was as good as gospel in the halls of MIT's AI Lab. Drexler, fascinated with Feynman's tiny devices, wanted to explore their possibility and Minsky ­- who had never forgotten the physicist's vision ­- immediately agreed. Thank you very much, Drexler replied, and went on to produce a vision that would come to shape the world.

When I attended MIT, in the early 80s, Drexler had just received his Masters Degree in computer science and had, like some Pied Piper of Cambridgeport, lured a small coterie of students into his orbit. Not yet "hackers" (though they certainly practiced the art of hacking, in the positive -- and now nearly lost -- sense of the word) they found in Drexler's ideas a blueprint for a future as programmable as they could imagine. In salons at his flat, Drexler entertained younger minds with a set of ideas he christened nanotechnology. Bring a bottle of wine, pull up a chair, and help design the future.

How could any red-blooded hacker resist such an opportunity?

I went to one of these salons d'idées, and by the end of the evening considered Drexler a prophet of the next age of Man, a time when nearly anything seemed possible. Nanomachines -­ or, more commonly, nanites - which could repair cellular-level damage and guarantee a nearly eternal, healthy existence; kitchen appliances which, fed on garbage, produced an endless supply of high-quality "meat"; an inexhaustible supply of incredibly strong building materials made of diamond, grown in forms of any conceivable volume. Drexler promised a material world nearly entirely subservient to the whim of the human imagination, programmed according to need.

Like many others in Drexler's orbit, it took me many years to absorb the full implications of nanotechnology. During this time Drexler worked hard to explain this revolution to laymen ­- his Engines of Creation: The Coming Era of Nanotechnology was published in 1986, with a forward by mentor Marvin Minsky ­- and he moved from Cambridge to Palo Alto to pursue doctoral work at Stanford. His thesis work, Nanosystems, published in 1992, grounded the wild speculations of nanotechnology in the hard-and-fast sciences of mechanics and atomic chemistry. A cookbook of atomic-scale machinery, with gears, rotors and motors, Nanosystems provided a codex atomicus for the nanomolecular universe.

Next, Drexler set out to change the world. But Drexler sensed that the implications of his research would be as profound as the work itself, and unlike Robert Oppenheimer, the Faust of our nuclear age, he laid the consequences out alongside the potential. It's important, in any discussion of nanotechnology, to bring these into full view, especially now that we know enough about nanotechnology's realities to imagine some rather harrowing nanotech futures.

TWO OF THE MOST CRUCIAL -- and, as yet, unbuilt -- devices in nanotechnology are the nanocomputer and nanoassembler. The nanocomputer, as its name implies, is a molecular machine capable of executing a string of instructions and producing a result. In function, it differs little from today's microprocessors, although it bears a curious resemblance to the antique, mechanical computers designed by Charles Babbage in the middle of the Victorian era, with rods and registers creating something like a grown-up adding machine (an adding machine a million times smaller and a billion times faster than any microprocessor yet designed).

Once the nanocomputer exists, it becomes possible to create a nanoassembler, a device constructed at the atomic level, which can arrange atoms precisely into most any desired form. Today, working at the atomic level requires bulky and expensive Atomic Force Microscopy (AFM), which uses electric fields to "push" atoms into position. But a nanoassembler could simply "pluck" atoms from a "bin" and, like a post-industrial loom, knit them into position. In our cells, ribosomes do something similar, copying DNA into RNA, and then gathering the correct amino acids to create the proteins which make up our physical nature. The nanoassembler, which contains a nanocomputer at its core, does much the same thing, translating instructions into molecules.

The nanoassembler is the Holy Grail of nanotechnology; once a perfected nanoassembler is available, almost anything becomes possible. This, then, is both the greatest hope and biggest fear of the nanotechnology community. Sixty years ago, John Von Neumann -­ who, along with Alan Turing founded the field of computer science ­- surmised that it would someday be possible to create machines that could copy themselves; a sort of auto-duplication which could lead from a single instance to a whole society of perfect copies. Although such a Von Neumann machine is relatively simple in theory, it has never been made. It's far easier, at the macromolecular scale, to build a copy of a machine than it is to get the machine to copy itself. At the molecular level, however, this balance is reversed; it's far easier to get a nanomachine to copy itself than it is to create another one from scratch. This is an enormous boon; once you have a single nanoassembler you can make as many as you might need. But it also means that a nanoassembler is a perfect plague. If, either intentionally or through accident, a nanoassembler were released into the environment, with only the instruction to be fruitful and multiply, the entire surface of the planet ­- plants, animals and even rocks -- would be reduced to a gray goo of nanites in little more than 72 hours.

The "gray goo problem" is well known in nanotechnology, where it acts as a check against the unbounded optimism which permeates scientific developments in atomic-scale devices. Drexler believes the gray goo problem to be mostly imaginary, but does concede the possibility of a "gray dust" scenario, in which replicating nanites "smother" the Earth in a blanket of sub-microscopic forms. In either scenario, the outcome is much the same. And here we encounter a technological danger unprecedented in history: If we'd blown ourselves to kingdom come in a nuclear apocalypse, the cockroaches at least would have survived. In a gray goo scenario, nothing ­- not even the bacteria deep underneath the ground ­- would be untouched. Everything would become one thing: a monoculture of nanites.

But despite the dangers, molecular nanotechnologists study how to build machinery up from the atomic scale, while molecular biologists study how to "strip down" the organelles of the cell into atomic-scale devices. Given the immense commercial pressures of the biomedical industry, it seems unlikely that molecular biologists will stop learning how we work, so one way or another, it seems, we'll soon know enough about nanocomputers and nanoassemblers to construct both.

This inevitability inspired Drexler to create a scientific and educational foundation, The Foresight Institute, to act as a clearing-house and think-tank for research into nanotechnology. In its 14 years, Foresight has grown to become a focal point for the community of nanotechnology researchers -­ and ethical discussions about the nature of the collective project have an important place in it. Public debate on matters nanotechnological is virtually nonexistent. Today the field belongs to research scientists and a growing community of amateurs. Only with a community structure in place, Drexler argues, can we have any degree of safety in a coming age of nanotechnology.

IN MID-OCTOBER, THE FORESIGHT INSTITUTE held their annual conference at a mid-grade hotel in Santa Clara (which is Silicon Valley's ground zero for the revolution in microelectronics and software). According to attendees, there was a new buzz in the air; recent developments in molecular-scale manufacturing have resulted in the invention of some of the very basic components Drexler described in Nanosystems -­ the same components which will be essential features of nanocomputers and nanoassemblers. The pieces are coming together.

Another eagerly sought-after prize at the conference was a copy of Robert Freitas' Nanomedicine. More than anything before it, Nanomedicine attempts to articulate the promise of Feynman's ultra-miniaturized "doctor," and lays out a path of step-by-step technological hurdles which must be overcome on the way towards nanomedical devices. Freitas is neither a doctor nor a molecular physicist. Despite his post at the Institute for Molecular Manufacturing (IMM) ­- which Drexler founded as the R&D and grant arm of Foresight -- he is really an amateur, uncredentialed in the field he describes. This would be an unrecoverable fault in more established fields of scientific discourse, but nanotechnology still lives in the liminal gap between imagination and reality. As in "Homebrew Computer Club" in the 1970s, there's plenty of room for amateurs; in a sense, everyone working in the field was still an amateur. The Homebrew Computer Club gave Steve Jobs and Steve Wozniak a platform to share their work and sell the Apple I, gave Lee Felsenstein the opportunity to demonstrate the first portable computers, and legitimized the amateur in a field dominated by corporate "big-iron" interests. Foresight, the IMM, and other nanotechnology interests have a similar feel ­- hackers on the edge of another revolution. And hackers are necessary to its development for many of the same reasons.

NANOTECHNOLOGY IS, HOWEVER, still beneath the cultural radar. Even the corporations sponsoring research into nanotechnology don't quite know what to do with it. Dr. Ralph Merkle, who has done more than any other individual, aside from Drexler, to advance the science of molecular engineering, had a post at Xerox PARC, but he left it last month. (Xerox wanted Merkle to split his focus between nanotechnology and public-key cryptography, a field which Merkle helped to define. But Merkle, unwilling to give his intellectual passion half-duty, left to become a Research Fellow for the Zyvex Corporation, the first of a new generation of nanotechnology startups.) Meanwhile, Congress, which seems resistant to funding any R&D that doesn't have immediate benefits for the military or medicine, doubled government funding for nanotechnology research. Some of that money will shower down on NASA's Ames Research Center in Mountain View, California, where a small team is working on the design of nanocomputers. Why is NASA interested in nanotechnology? Size, mostly. Current computers, such as those found on the Mars Pathfinder, are large, power-hungry and prone to failure. Using nanobots, NASA could send a hundred million tiny eyes and ears to the Martian surface in a package weighing a few grams. Who cares if half of them fail? There are still 50 million left!

The nanobot is still just a dream; to create one, researchers will have to crack the problem of the nanocomputer -­ the focus of NASA's research group. It's a problem they don't expect to solve until around 2011 (but then, this estimate was made back in 1997, a lifetime ago in nanotechnology). But every day researchers are posting new breakthroughs, gleaned from materials science or molecular biology, that are propelling them towards a future we'd find nearly unbelievable.

Forty years after Feynman, the promise of nanotechnology looms as the most important technological development in our history. It promises perfection and apocalypse. In the perfect worlds of fantasy science fiction, all want has been satisfied and all disease cured. Without the inequities that produce politics or the sufferings that drive melodrama, the human story rings hollow, as if our pains give birth to our drives. It makes for bad storytelling: no room for heroes or noble acts, no sacrifice to create moral legends. But the approach to such a perfect world seems fraught with pitfalls, the ascent to perfection allowing ample opportunity for the darker forces of our nature to present themselves in their full dimension.

If this were an entirely hypothetical question, we could hand it off to the ethicists and moralists, who could study the problem for a thousand years. But in less than a thousand weeks, we will be confronting these questions collectively, and no less than the fate of humanity hangs in the balance. Already "Nightline" spends a week examining the impact of bio-terrorism in the American city, and CIA analysts lie awake at night wondering who among our enemies (and our friends) has the capacity to wreak destruction on our very cells. If the threat ended there, if we could simply inoculate ourselves against the terrors that our neighbors might infect us with, we could content ourselves in believing that the future has much the same form as the past, that we know the shapes of the things that go bump in the night. But more and more it becomes clear that we are opening into a new day, and everything we know matters not at all.