You might not know it, but among us there exist artificial beings that are lifelike enough to give you goose bumps. If you had visited robot developer Rodney Brooks at MIT in the late 1990s, you would have met his Cog (short for “cognitive”) robot. Shaped vaguely like a human head and torso, and built more or less to human scale, Cog still looked alien and machinelike because it was made of girders and electronic components. Instead of eyes, video cameras located in its head fed visual information to its computer brain.

But when I saw Cog’s intricate body language, I forgot its machine appearance. Although it did not look like a person, it acted like one. Those sensors and computers, motors and metal supports kept its “eyes” in continual motion, scanning the scene for interesting events—just as our own brains and eyes do at an unconscious level. And when the door opened and a student walked in, Cog did what you or I would; it stopped scanning and turned toward the visitor. As Cog brought its gaze and (apparently) its full attention to bear on her, the action was so eerily human that it gave me a moment of hair-raising, gut-level understanding, for in that instant, Cog seemed fully alive and conscious.

A year or two later, in that same laboratory, you would have met another robot, Kismet, created by Cynthia Breazeal, Brooks’s graduate student at the time, now an MIT professor and well-known robotics researcher in her own right. Where Cog is intimidating, Kismet has a face out of a children’s storybook, clownlike and cartoonish with exaggerated features—huge blue eyes, bright red lips, and prominent, highly mobile ears. Approach Kismet and engage its attention by waving a toy or talking, and it responds in a tiny voice, moving its head, and adjusting its face to smile, or to look sad, angry, or fearful. When Kismet was young, Breazeal brought in adults and children to interact with it. Today she says, “Kismet became a personality to them, to the point where people still ask me ‘How’s Kismet?’ They refer to Kismet as a creature rather than this thing in the lab.”

Human reactions to Cog and Kismet offer an important lesson: regardless of what is going on inside an artificial being—and the debate over what might constitute “machine intelligence” and “machine consciousness” is a deep and continuing one—the merest hint of humanlike action or appearance deeply engages us. Cog generates a surprising sense of life simply through its reactions to its environment. Kismet goes further; it reacts to people with, for example, facial expressions that humans sense in a direct and natural way.

Other artificial creatures add vigorous body movements or other levels of interaction. At the Honda research laboratories in Japan, a child-size robot, humanoid in outline, walks, balances on one foot, and nimbly climbs stairs without a hitch. At MIT, Carnegie Mellon University, and the Palo Alto Research Center, artificial creatures roll, slither, crawl, stride, and hop across the floor, or configure and reconfigure their bodies so as to locomote in the most efficient way. At the ROBODEX 2003 exposition in Yokohama, Japan, robots answered questions, reacted to human body language, sang, danced, and played soccer. At Walt Disney theme parks in the United States and Europe, and in countless Hollywood movies, entertainment androids convincingly simulate people, animals, and imaginary beings.

Most remarkably, artificial creatures are beginning to generate a kind of emotional lifelikeness because they create warm feelings in

people, as you can easily see without visiting any robotics laboratories. Just spend a few moments with any recent robotic toy such as the Sony Corporation’s AIBO dog (Artificial intelligence + robot) that went on sale in 1999, or the I-Cybie robot dog, made by Tiger Electronics and Silverlit Toys. These two could never be taken as natural creatures because both are only plastic caricatures of a dog, but like Cog, they need not work very hard to elicit human reactions. If the creature interacts with the world, has some capacity to change its behavior as it gains experience—that is, if it can learn—and displays natural-seeming behavior, it can project a well-nigh irresistible impression of life.

As you watch I-Cybie cock its head toward you when you call its name, or AIBO perform a trick at your voice command, it’s easy to feel something toward the mechanism: amazement that it listens to you or a small rush of affection. And if the synthetic being looks like a human rather than an animal, like Kismet’s face or the toy robot infant called My Real Baby released in 2000, its emotional power is far more intense.

On the face of it, it might seem unreasonable to have feelings toward “a creature that really doesn’t know you’re there,” as sociologist Sherry Turkle of MIT puts it, yet it happens all the time. Little girls have always loved their dolls, no matter how crude, and children and adults bond to objects and machines not in the least cute or petlike. We become attached to bicycles, boats, and computers, giving them names, endowing them with personalities, and projecting human or animal dimensions onto their actions. We swear at a “stub-born” or “cranky” lawnmower that won’t start, or affectionately caress a sleek car as we would a superb racehorse.

Artificial beings, however, are not limited to fully manufactured creatures of plastic and metal. We ourselves are partly artificial or “bionic”—that is, people with synthetic parts—to a surprising extent: 8 to 10 percent of the U.S. population, approximately 25 million people, and becoming more so as our population ages. Our bionic additions include functional prosthetic devices and implants, such as artificial limbs, replacement knees and hips, and vascular stents (tiny

gridlike metallic inserts that aid the flow of blood in blocked arteries, in themselves a multibillion dollar industry). There are also cosmetic or vanity bionic additions that replace what accident or nature took away, or genetic inheritance never gave, from false hair and teeth to artificial eyes and breast augmentations (more than 200,000 of which were implanted in 2001 alone).

Cosmetic additions like these might be used in a project to make an artificial creature look human, but bionic alterations beyond the merely cosmetic are significant steps toward building a whole creature. Where once we had only crude prosthetic devices such as a glass sphere in place of an eye, or an iron hook for a hand, now we are developing functional body parts that are increasingly indistinguishable from the real thing, some with neural connections—not only limbs, but replacements for lost or diseased vision, hearing, and other capacities and organs. Most startling of all, we now look beyond the physical to “bionic brains” that is; “artificial” means to alter or augment mental capacity and emotional states, from implanted drug delivery systems to computer chips connected directly to the neural network.

The combination of human with artificial components lies at one end of a spectrum of artificiality, depending on how much of the being is made of natural or living parts, or is meant to look natural, and on how self-directed the creature is—from automaton, to robot, to android, to cybernetic organism or cyborg, to bionic human. An automaton is a machine that appears to move spontaneously, although actually it moves “under conditions fixed for it, not by it” according to one definition. A robot is an autonomous or semiautonomous machine made to function like a living entity (here, “machine” includes mechanical, inorganic, or organic but nonliving moving or static parts, and electronic, computational, and sensory components). It can be humanoid, although not necessarily so; most contemporary robots take nonhuman shapes that are useful for their particular applications. An android is similarly entirely artificial but has been made to look human (the word comes from Greek roots meaning “manlike.”). Star Trek: The Next Generation’s Commander Data is a popular example of an android.

A cyborg (cybernetic organism) and a bionic human (from “biological” and “electronic”) are different from the previous three categories, in that both involve a combination of machine and living parts. In my usage, a cyborg has a machine portion that might dominate the natural part in mass and bulk but is under the mental direction of the natural part—essentially, a brain in a box. A bionic human, on the other hand, is mostly natural with a relatively small portion given over to implants or replacement parts such as a heart pacemaker or an artificial limb.

The categories from automaton to bionic human include mobile and responsive robots, amusingly lifelike toys, entertainment androids, humans with mechanical and electronic implants, and others. All are part of a technology that is beginning to realize an extraordinary achievement: the creation of partly or fully artificial beings. Although these possibilities draw on the ultimate in twenty-first-century science, they are not new in the collective human imagination; artificial beings have intrigued, terrified, and exalted us for millennia.

The reasons for this long-standing interest are basic to human nature, although it is not easy to say which of the reasons dominate. Least noble, perhaps, but understandable, is the desire to ease our lives by creating workers to till our fields, operate our factories, and prepare our meals, tirelessly and without complaint. As long ago as the fourth century BCE no less a thinker than Aristotle saw the potential for automated machinery to reduce labor, and even its potential for disrupting the job market:

Aristotle’s idea was perhaps first realized in eighteenth-century France in an innovative and efficient automated loom for silk weaving. The silk workers immediately understood that the device meant the loss of their livelihoods and objected to its adoption. This is one example of the contradictory quality typical of many aspects of artificial beings (and indeed of all technology): the good that they might bring is counteracted by undesirable side effects that might ensue.

At another level, perhaps nobler, perhaps only a matter of enlightened self-interest, is the desire to transcend our limitations: We imagine creating beings that go beyond humanity’s natural physical and mental endowments. A related motivation is the desire to bionically repair ill or damaged bodies and minds or to enhance them for better performance, improved health, and longer lifespan—or, returning to the cosmetic theme, for greater beauty. These, too, are old ideas. Indian mythological writings 4,000 years old or older tell of a warrior queen who went into battle with a prosthetic iron leg, and in Norse mythology, Sif, wife of the god Thor, had dwarves make golden hair for her.

The desire to make ourselves healthier and more beautiful is rooted in our strongest motivation to consider artificial beings: fear of death. It is extreme fantasy, perhaps, to think that artificial creations might allow individuals or the entire race to foil nature and achieve immortality; but it is no fantasy to say that as we develop such beings, we begin scientifically exploring the incomprehensible gap between the living and dead, the animate and inanimate.

Inevitably, even cautious forays into this territory carry a scent of hubris, in the belief that we can outdo evolutionary forces or perhaps God Himself. As the science fiction author Stanislaw Lem has written: “The concept of an artificially created man is blasphemy in our cultural sphere. Such a creation [is] a caricature, an attempt by humans to become equal to God.” From the viewpoint of traditional religion, he adds, this blasphemy could succeed only if humanity were to collaborate with the forces of evil; that is why an air of the uncanny surrounds these creatures.

For those who find this eeriness unsettling or the blasphemy unforgivable, other motives for making artificial creatures might prove compelling. Beyond physical improvements, perhaps we can create beings or states of consciousness that avoid our moral and spiritual failings, thereby guiding us toward becoming better humans. And from the scientific viewpoint, what could be more important than to understand the origins and processes of life? In this spirit, research on artificial beings is a way to express our sense of wonder about life and

our place among the living, and to better understand both. And surely the potential medical benefits to humanity cannot be dismissed.

Nevertheless, when considering the creation of artificial beings, we must also consider the ambiguities and dark notes inherent in the quest. Whenever such creatures seem to cross the boundary between the living and the dead, the result is awesomely frightening, as shown in a tale from Roman times told in Gerard Walter’s biography of Julius Caesar. Supposedly, at the assassinated Caesar’s funeral, the crowd suddenly experienced “a vision of horror [and] brutality” when

In a similar vein, in Mary Shelley’s Frankenstein (1818), when Victor Frankenstein sees the first stirrings of the being he has created from dead body parts, he is shocked and horrified and spontaneously rejects his creation.

These visceral reactions represent the deepest fears that artificial beings might engender. But not every such creature represents a direct challenge to death or to God’s law—and if that challenge is absent, so is supernatural fear. When Sigmund Freud addressed this sense of dread in his essay “The ‘Uncanny,’” he did not relate it to religious guilt about blasphemy, but to knowledge of our own mortality. We feel uncanny, he says, when a deep emotion that has been repressed is made to recur. Our feelings about death are like that. Children, Freud notes, unambiguously want their inanimate dolls to come to life. Children, however—at least very young ones—have no knowledge of death. Adults do, and as Freud says, because of the “strength of our initial emotional reaction to death and the insufficiency of our scientific knowledge about it … almost all of us still think as savages do on this topic.” And so a special eeriness arises in the presence of a dead body or when we wonder whether something seemingly dead, such as an automaton, is actually alive.

If Freud is correct, then research on artificial beings can only reduce the sense of uncanniness as it explores the borderline between

the living and the dead. In any case, the technological creation of beings from inert metal, plastic, and silicon is a different matter from animating the dead. Perhaps that explains why technologists seem unconcerned about blasphemy as they try to create synthetic beings. One young researcher in the field recently summed it up when she said, “I thought it would be neat to design something that reproduces what God can do.” Call her attitude what you will—hubris, or a healthy pride in science—the scientists and engineers spearheading the creation of artificial beings and bionic people are responding to the magnetism of the technological imperative, the pull of a scientific problem as challenging as any imaginable.

Fascinating scientific puzzle though it is, the creation of artificial beings is also expected to meet important needs for society and individuals. Industrial robots are already widely used in factories and on assembly lines. Robots for hazardous duty, from dealing with terrorist threats to exploring hostile environments, including distant planets, are in place or on the drawing boards. Such duty could include military postings because there is a long-standing interest in self-guided battlefield mechanisms that reduce the exposure of human soldiers, and in artificially enhanced soldiers with increased combat effectiveness. (For this reason, the Department of Defense, largely through its research arm—the Defense Advanced Research Projects Agency (DARPA)—is the main U.S. funding source for research in artificial creatures.) Artificial creatures can also be used in less hostile environments: homes, classrooms, and hospitals and rest homes, serving as all-purpose household servants, helping to teach, and caring for the ill or elderly.

Among these possibilities, the connection between artificial creatures and human implants might be the most important because it promises enormous medical benefits. This connection might be the single greatest motivation to develop artificial beings. Yet regardless of their potential good uses, and apart from any issues of blasphemy, we have concerns about robots and androids. One fear is that the limitations we think to design out of our creations, from cosmetic deficiencies to the existential realities of illness and death, are essential human

attributes, and that to abandon them is somehow to abandon our humanity. Something in us, it seems, fears perfection, and artificial beings threaten us with an unwelcome perfection, expressed as rigid unfeeling precision.

There is another menace first conveyed nearly 200 years ago in Frankenstein, and now more compelling than ever: the fear that technology will grow out of control and diminish humanity for all of us. That concern is hardly limited to artificial creatures. It appears in many arenas—the loss of privacy associated with new forms of surveillance and data manipulation; the depersonalization of human relationships; the incidence of human-made ecological disaster; the growing gap between the world’s technological “haves” and “have-nots.” It is especially and deeply unsettling, however, to contemplate the literal displacement of humanity by beings made in the human image, only better.

Although Frankenstein is the most famous story touching on many of these matters, it is not the only one. The depth of our reactions is shown in a whole imaginative narrative of artificial beings—a millennia-old fantasy or “virtual” history, in which these creatures are the focus of a panoply of emotions, hopes, and concerns. In one thread of the virtual history, humans develop strong feelings for inanimate or artificial beings, as in the Greek myth of Pygmalion, who yearns for his statue of a beautiful woman to come alive. That thread also appears in E.T.A. Hoffman’s nineteenth-century story “The Sandman,” where a young man falls in love with a clockwork automaton, and in the classic 1982 science-fiction film Blade Runner, where a special agent dedicated to the destruction of androids falls in love with one of them. In another thread in the virtual history, artificial beings yearn to become human or accepted as human, for example the “monster” in Frankenstein, the puppet Pinocchio, Commander Data in Star Trek, and the little boy android in the 2001 film A. I.: Artificial Intelligence.

In yet other stories, robots display intelligence and ethical standards that make them trusted guides to a better future for humanity, as in Isaac Asimov’s book I, Robot, but in a contrary thread, other equally able robots and androids slaughter people, as in Karel Capek’s play

R.U.R. and the recent Terminator films. And even if artificial beings do not wish to wipe us from the earth, their superiority might still destroy us by stifling human creativity and independence, as in Jack Williamson’s story “With Folded Hands.”

No current artificial creatures can carry out these scenarios, nor are there yet bionic humans or cyborgs who are the physical or mental superiors of natural people. The abilities of robots and androids are still limited. If they behave intelligently, they do so only in specialized areas, or at a childlike rather than an adult level; though they might be mobile, they cannot yet independently navigate any arbitrary room or street; they are not conscious and self-aware, and hence are not moral beings as we understand morality; they are not emotional, and although they might elicit affection or an appreciation of cuteness as a living pet does, they evoke no deeper feelings.

They cannot pass for human in either appearance or behavior, at least not at the behavioral level proposed by the British mathematician, Alan Turing, in 1950. In what is now universally known as the Turing test, he proposed a purely verbal criterion for defining a “thinking machine” as intelligent. Imagine, he said, that a human observer can communicate with either the machine or another human without seeing either (for instance, via keyboard and printer), and can ask either any question. If after a reasonable time the observer cannot identify which of the two is the computer, the machine should be considered intelligent.

Some researchers now think the Turing test is not a definitive measure of machine intelligence. Yet it still carries weight, and now, for the first time in history, the means might be at hand to make beings that pass that test and others. Advances in a host of areas—digital electronics and computational technology, artificial intelligence (AI), nanotechnology, molecular biology, and materials science, among others—enable the creation of beings that act and look human. At corporations and academic institutions around the world, in government installations and on industrial assembly lines, artificial versions of every quality that would make a synthetic being seem alive or be alive—intelligent self-direction, mobility, sensory capability, natural

appearance and behavior, emotional capacity, perhaps even consciousness—are operational or under serious consideration.

Not everyone engaged in these efforts is a robotics engineer or computer scientist. Researchers in other fields are working to help ill and injured people: Some of the most exciting efforts are in biomedical research laboratories, in hospitals and clinical settings, where physicians and engineers are developing artificial parts, such as retinal implants for the blind, that might eventually enhance human physical and mental functions. The medical applications and the engineering technologies enhance each other, and as they grow together, the potential for therapeutic uses brings significant motivation and a clear moral purpose to the science of artificial beings.

There is, however, considerable debate about the possibility of achieving the centerpiece of a complete artificial being, artificial intelligence arising from a humanly constructed brain that functions like a natural human one. Could such a creation operate intelligently in the real world? Could it be truly self-directed? And could it be consciously aware of its own internal state, as we are?

These deep questions might never be entirely settled. We hardly know ourselves if we are creatures of free will, and consciousness remains a complex phenomenon, remarkably resistant to scientific definition and analysis. One attraction of the study of artificial creatures is the light it focuses on us: To create artificial minds and bodies, we must first better understand ourselves.

While consciousness in a robot is intriguing to discuss, many researchers believe it is not a prerequisite for an effective artificial being. In his Behavior-Based Robotics, roboticist Ronald Arkin of the Georgia Institute of Technology argues that “consciousness may be overrated,” and notes that “most roboticists are more than happy to leave these debates on consciousness to those with more philosophical leanings.” For many applications, it is enough that the being seems alive or seems human, and irrelevant whether it feels so. Even our early explorations of artificial beings show us that the goal of seeming alive and human might be less challenging than we might expect because—for reasons only partly apparent—we tend to eagerly embrace artificial beings. As

in the common reaction to Kismet or the robotic dogs, it takes only a few cues for us to meet creatures halfway, filling in gaps in their apparent naturalness from the well of our own humanity. In a way, an artificial being exists most fully not in itself, but in the psychic space that lies between us and it.

And yet … there is the dream and the breathtaking possibility that humanity can actually develop the technology to create qualitatively new kinds of beings. These might take the form of fully artificial, yet fully living, intelligent, and conscious creatures—perhaps humanlike, perhaps not. Or they might take the form of a race of “new humans” that is; bionic or cyborgian people who have been enormously augmented and extended physically, mentally, and emotionally.

New humans could also arise from a different thread in modern technology. Purely biological methods such as cloning, genetic engineering, and stem-cell research offer another way to enhance human well-being and change our very nature. While astonishing progress has been made in these areas, we have yet to see definitive, broad-scale results. Moreover, a program for changing humans at the genetic level has ethical and religious implications that trouble many people, and the consequences of human-induced changes propagating in our gene pool trouble many scientists. The creation of fully or partly artificial beings has its own set of moral issues; these, however, might ultimately prove more acceptable to society than those arising from genetic manipulation.

At its furthest reach, and as a great hope for the technology of artificial beings, we might be able to create a companion race—self-aware and self-sufficient, perhaps like us in some ways but different in others, with its own view of the universe and new ways to think about it. Fascination with the notion of communicating with another race of beings has been a main incentive in the search for intelligent life elsewhere in the universe—a hope that engages many people, as witness the great interest in the 1996 announcement that traces of ancient life were found on Mars. But that announcement was mistaken, and although the search continues (for instance, with the 2004

landing of two NASA robotic planetary explorers on Mars), we have yet to find evidence of alien beings anywhere that our spacecraft and telescopes can reach. Perhaps we never will, so the creation right here on Earth of a race that complements humanity has special appeal.

No matter what emerges from controversies about robotic consciousness or the morality of making artificial beings, no matter what approach to artificial intelligence proves effective, one thing is clear: Without digital electronics and digital computation, we could not begin to consider artificial intelligence and artificial sensory apparatus, the physical control of synthetic bodies, and the construction of interfaces between living and nonliving systems. Although the history of artificial beings has presented many ways to create them, animate them, and give them intelligence, now we are truly entering an era of digital people.

This page intentionally left blank.