Building Athens Without the Slaves

The heart is a pump, the eyes tiny cameras, the brain a computer: The human body is a machine. Yet when Descartes first expressed this idea over 300 years ago to his royal student, Queen Christina of Sweden, she came up with a cogent question. "How," she asked, "can machines reproduce themselves?"

Until recently Queen Christina's question had no answer. A lathe can make many useful things, but it can't duplicate its parts and put them together to make a baby lathe identical to itself. To do so would require deep knowledge of its own construction-the lathe would have to know itself. No machine ever has, but recently engineers and scientists at NASA have shown that a self-replicating machine could be built. With a relatively modest investment, a self-replicating factory could be built by early next century. The Japanese have reached a similar conclusion.

Self-replicating machines would be more than high-tech pets. Since they would produce factories as well as products, self-replicating machines would be the ultimate tools for increasing productivity. One application investigated by the NASA group was to use a self-reproducing machine on the surface of the moon to make solar cells, which could be cheaply hauled to vast solar-power satellites that would beam solar energy back to earth. The researchers considered the task of making one million solar cells out of the moon dust. A $1 billion, 100-ton conventional factory would take about 6,000 years to perform this task. A self-replicating factory of the same initial size and cost could do the job in just 20 years, because it makes not only Solar cells but more solar-cell factories. Marvin Minsky, artificial- intelligence expert at the Massachusetts Institute of Technology and former president of the American Association for Artificial Intelligence, recently wrote, "Teaching computers how to build copies of themselves could begin a flood of automatic self-replication machines making more machines at very low cost. Then we'll have to learn to cope with the resulting explosive growth of wealth and productivity."

It was not until 1948 that scientists became convinced that machines could be taught to replicate themselves. In that year John von Neumann, the Hungarian-American mathematician who helped design the first stored-program computer, gave a series of historic lectures at the University of Illinois. He showed that to duplicate itself, a device must have available its own blueprint, some manipulators-hands, essentially-and a set of rules for building things. The beauty of von Neumann's approach was that he was able to prove, with mathematical rigor, the possibility of building self-reproducing machines.

The simplest self-reproducing machine von Neumann imagined would sit in a giant stockroom filled with replacement parts-extra arms, legs, eyes, circuit boards, and the other paraphernalia from which it was built. The machine would have a memory tape containing all the instructions needed to build a copy of itself from these spare parts. Using its robot arm and its ability to move around, the machine would find and connect parts. The tape program would instruct the device to reach out and pick up a part, look to see if it's the right one, and if not, put. it back and grab another. Eventually the correct one would be found, then the next, and the two joined in accordance with the master checklist. The machine would continue to follow the instructions, never knowing what it is making, until it Finishes assembling a physical duplicate of itself. But the new robot would be "uneducated"; it would have no instructions on its tape. The parent would then copy its own memory tape onto the blank tape of its offspring. The last instruction of the parent's tape would be to activate its progeny.

Skeptics may scoff at this as mere instruction-book-style assembly. The parts in the warehouse already were "almost" robots. Somebody had to build them ahead of time. But some prefabrication of starting parts must be allowed. "After all," writes Dr. W. Ross Ashby, expert in biophysics at the Burden Neurological Institute in England, "living things that reproduce do not start out as a gaseous mixture of raw elements." Proteins, carbohydrates, and fats are the sophisticated, prefabricated materials from which humans build copies of themselves.

Some simple examples of machines that reproduce without human intervention do exist. Small mechanical contrivances capable of self-assembly from simpler parts are easy to build. Many years ago Roger Penrose, a British physicist, fabricated a set of simple blocks that could hook together using a set of ratchets and levers. When single parts are placed in a box-a track shaped like a long trough-and shaken, they do not join together. However, when an interlocked, two-block unit is placed in the box and shaken, a simple form of reproduction takes place. Collisions between the two-block unit and other parts in the box cause new two-block units to form, each identical to the original. Penrose successfully tested replicating units of up to four blocks, using several different block types.

Compact, self-replicating robots, capable of doing more than dumbly duplicating, are far more difficult to build. Minsky says, "Today it might be possible to build a self-replicating machine, but it might have to be the size of a factory." Minsky believes that in the future these machines could be as small as animals-also self-reproducing units. Charles Rosen, founder of the Robotics and Automation Division of Stanford Research Institute and chief scientist of the Machine Intelligence Corporation, agrees. "It is important to recognize that a complex robot is not essential to make a useful machine. It is possible to design a modular growing robot system, like the early lathes that could produce parts that could make better lathes, which could then produce parts that made better lathes, and so on. But the exercise has not been done. It is an exciting concept that has yet to be fully explored, although I have little doubt that it will be."

The Japanese are already investigating the possibility of self-reproducing machines. Ten years ago the Japanese Ministry of International Trade and Industry began a feasibility study called "Methodology for Unmanned Manufacturing." In 1977 the ministry plunked down $60 million in seed money for a seven-year research and development program aimed at "complete automatic manufacturing" by the start of the next century. The first prototype "unmanned" production line was started two years ago by Fujitsu FANUC, a leading international manufacturer of machinery. At the $40 million factory, computer-controlled industrial robots build other industrial robots with minimum human intervention. Only 100 human supervisory and manufacturing personnel are needed, just one-tenth the usual number. An automated plant where robots make robots is an important first step toward practical self-replicating systems.

A recent NASA study, conducted at the University of Santa Clara in 1980 with the support of Ames Research Center and the American Society for Engineering Education, examined the automation technologies needed to achieve machine self-replication. These requirements were compared to the existing state of the art in robotics and computer science. The surprising conclusion: According to industry experts and the NASA analysis, a primitive self-reproducing robot able to assemble copies of itself from a couple of dozen components drawn from a parts bin could be running within five years, for a total investment of from $5 million to $50 million using a specialist staff of only about a dozen engineers.

The NASA study proposes to tackle the practical difficulties of machine replication in a four-phase program. The first phase is just von Neumann's simple robot in a warehouse. But by the fourth phase the NASA scientists hope to have an entire, self-sufficient factory capable of building copies of itself from raw material-natural ores and soils. This final "parent" replicating system would be about the size of a football field, weigh 100 tons, and consume a few megawatts of power, and it could reproduce itself about once a year.

Former NASA administrator Robert A. Frosch clearly understood the power of self-reproducing factories. "It appears we can use machines in a pseudo-biological way to establish a productive machine economy," he observed during a 1979 speech on the future of robotics in NASA. "The key is the construction of a machine which automatically, or with minimal human [remote-control] guidance, can use solar energy and local materials on the earth or elsewhere in the solar system to build a replica of itself. Then these machines will construct further generations of machines."

Replicating systems are best employed when massive output is desired at little cost or in hostile or inaccessible locations. A "seed" factory is planted, which builds additional facilities so that it can produce useful products. Possible products include large solar-cell arrays for electrical power plants in the southwestern United States, desert irrigation and soil-conditioning equipment covering vast areas, mass-production of automated agricultural machines or military material, fleets of ocean-bottom mineral- and metals-retrieving robots patrolling the continental shelves, and expansive fields; of solar-power satellite groundr-receiver elements.

Replicating factories on the moon could make space exploration affordable. "On the lunar surface we could build thousands of lunar rovers to cover the moon like ants, measuring, exploring and mapping, surveying the entire surface in just a few years," says Georg von Tiesenhausen, assistant director of the Advanced Systems Office at the Marshall Space Flight Center in Huntsville, Alabama. "By conventional methods this exploration might take a century or more." Ultimately, replicating machines could be turned loose in space to build giant manufacturing complexes on the moon, automatically assemble orbital space colonies, rework the surfaces of entire planets such as Mars to make them fit for human habitation, or make millions of automatic probes to be sent to other stars in search of extraterrestrial intelligence.

Pessimists, such as, von Tiesenhausen, believe that within 20 years after the project is begun, the United States could produce the first robot factory able to replicate itself from raw materials. Frosch is more optimistic. "We are very close to understanding how to build such machines," he says. "I believe that the technology is presently available and that the necessary development could be accomplished in a decade or so."

An early workshop sponsored by Gerard K. O'Neill, president of the Space Studies Institute in Princeton, New Jersey, and an advocate of space colonies as the future home of humankind, found that self-reproducing machines could dramatically reduce the cost of the industrialization of space. Instead of the initial cost estimates in the hundreds of billions of dollars, O'Neill found that self-replicating systems "appear capable of achieving high levels of productivity for investments considerably less than $10 billion. They would be in the range of the Alaska pipeline ($7 billion), and much lower than the Churchill Falls, Quebec, electric power system, both of which were private ventures." According to the NASA study, the proposed self-replicating factory could repay its own initial start-up costs after only a few years on the job. Comments British Agriculture Minister Peter Walker: "Uniquely in history, we have the circumstances in which we can create Athens without the slaves."

References To This Page

• The Artificial Self-Replication Page by Moshe Sipper