tumbled into the gully. Consequently, any Martian Rover requires slope and roughness sensors. Fortunately, these are readily available and are even seen in some children’s toys. When confronted with a precipitous slope or large boulder, the spacecraft would either stop until receiving instructions from the Earth in response to its query (and televised picture of the terrain), or back off and start in another and safer direction.
Much more elaborate contingency decision networks can be built into the onboard computers of spacecraft of the 1980s. For more remote objectives, to be explored further in the future, we can imagine human controllers in orbit around the target planet, or on one of its moons. In the exploration of Jupiter, for example, I can imagine the operators on a small moon outside the fierce Jovian radiation belts, controlling with only a few seconds’ delay the responses of a spacecraft floating in the dense Jovian clouds.
Human beings on Earth can also be in such an interaction loop, if they are willing to spend some time on the enterprise. If every decision in Martian exploration must be fed through a human controller on Earth, the Rover can traverse only a few feet an hour. But the lifetimes of such Rovers are so long that a few feet an hour represents a perfectly respectable rate of progress. However, as we imagine expeditions into the farthest reaches of the solar system-and ultimately to the stars-it is clear that self-controlled machine intelligence will assume heavier burdens of responsibility.
In the development of such machines we find a kind of convergent evolution. Viking is, in a curious sense, like some great outsized, clumsily constructed insect. It is not yet ambulatory, and it is certainly incapable of self- reproduction. But it has an exoskeleton, it has a wide range of insectlike sensory organs, and it is about as intelligent as a dragonfly. But Viking has an advantage that insects do not: it can, on occasion, by inquiring of its controllers on Earth, assume the intelligence of a human being-the controllers are able to reprogram the Viking computer on the basis of decisions they make.
As the field of machine intelligence advances and as increasingly distant objects in the solar system become accessible to exploration, we will see the development of increasingly sophisticated onboard computers, slowly climbing the phylogenetic tree from insect intelligence to crocodile intelligence to squirrel intelligence and-in the not very remote future, I think-to dog intelligence. Any flight to the outer solar system must have a computer capable of determining whether it is working properly. There is no possibility of sending to the Earth for a repairman. The machine must be able to sense when it is sick and skillfully doctor its own illnesses. A computer is needed that is able either to fix or replace failed computer, sensor or structural components. Such a computer, which has been called STAR (self-testing and repairing computer), is on the threshold of development. It employs redundant components, as biology does-we have two lungs and two kidneys partly because each is protection against failure of the other. But a computer can be much more redundant than a human being, who has, for example, but one head and one heart.
Because of the weight premium on deep space exploratory ventures, there will be strong pressures for continued miniaturization of intelligent machines. It is clear that remarkable miniaturization has already occurred: vacuum tubes have been replaced by transistors, wired circuits by printed circuit boards, and entire computer systems by silicon-chip microcircuitry. Today a circuit that used to occupy much of a 1930 radio set can be printed on the tip of a pin. If intelligent machines for terrestrial mining and space exploratory applications are pursued, the time cannot be far off when household and other domestic robots will become commercially feasible. Unlike the classical anthropoid robots of science fiction, there is no reason for such machines to look any more human than a vacuum cleaner does. They will be specialized for their functions. But there are many common tasks, ranging from bartending to floor washing, that involve a very limited array of intellectual capabilities, albeit substantial stamina and patience. All-purpose ambulatory household robots, which perform domestic functions as well as a proper nineteenth-century English butler, are probably many decades off. But more specialized machines, each adapted to a specific household function, are probably already on the horizon.
It is possible to imagine many other civic tasks and essential functions of everyday life carried out by intelligent machines. By the early 1970s, garbage collectors in Anchorage, Alaska, and other cities won wage settlements guaranteeing them salaries of about $20,000 per annum. It is possible that the economic pressures alone may make a persuasive case for the development of automated garbage-collecting machines. For the development of domestic and civic robots to be a general civic good, the effective re-employment of those human beings displaced by the robots must, of course, be arranged; but over a human generation that should not be too difficult-particularly if there are enlightened educational reforms. Human beings enjoy learning.
We appear to be on the verge of developing a wide variety of intelligent machines capable of performing tasks too dangerous, too expensive, too onerous or too boring for human beings. The development of such machines is, in my mind, one of the few legitimate “spinoffs” of the space program. The efficient exploitation of energy in agriculture-upon which our survival as a species depends-may even be contingent on the development of such machines. The main obstacle seems to be a very human problem, the quiet feeling that comes stealthily and unbidden, and argues that there is something threatening or “inhuman” about machines performing certain tasks as well as or better than human beings; or a sense of loathing for creatures made of silicon and germanium rather than proteins and nucleic acids. But in many respects our survival as a species depends on our transcending such primitive chauvinisms. In part, our adjustment to intelligent machines is a matter of acclimatization. There are already cardiac pacemakers that can sense the beat of the human heart; only when there is the slightest hint of fibrillation does the pacemaker stimulate the heart. This is a mild but very useful sort of machine intelligence. I cannot imagine the wearer of this device resenting its intelligence. I think in a relatively short period of time there will be a very similar sort of acceptance for much more intelligent and sophisticated machines. There is nothing inhuman about an intelligent machine; it is indeed an expression of those superb intellectual capabilities that only human beings, of all the creatures on our planet, now possess.
CHAPTER 21
What has been done is little-scarcely a beginning; yet it is much in comparison with the total blank of a century past. And our knowledge will, we are easily persuaded, appear in turn the merest ignorance to those who come after us. Yet it is not to be despised, since by it we reach up groping to touch the hem of the garment of the Most High.
AGNES M. CLERKE,
(London, Adam and Charles Black, 1893)
THE WORLD has changed since 1899, but there are few fields which have changed more-in the development of fundamental insights and in the discovery of new phenomena-than astronomy. Here are a few titles of recent papers published in the scientific magazines
