Report homepage > Current
issue front page
February 7, 2005
Sidney Perkowitz is Charles howard candler professor of physics. this essay is based on his book, digital people (joseph henry press/national academy, 2004).
If you saw the 2001 Steven Spielberg movie A.I., you probably remember the wrenching scene where David, a near-perfect robot version of a cute little boy, is abandoned by his adoptive human mother. Although David is a machine that only pretends to eat and never sleeps, he looks and acts sufficiently childlike to engage his “mother’s” emotions.
David is an example of how robots can elicit human responses as these machines become more like people. Other films show the opposite process, where humans become more machine-like. A prime example is the 1987 cult film RoboCop in which a police officer named Alex Murphy, killed in the line of duty, has his brain and personality resurrected within a massive steel body. The result is a hybrid man-machine whose physical abilities far exceed human limits, but who unexpectedly retains human emotions.
You might dismiss such tales as sheer fantasy, but in our technological age, science fiction has an uncanny way of becoming real. These films and others (for example, last year’s I, Robot, based on the famous stories by Isaac Asimov) show where real-life robotic and bionic (biological plus electronic) technology is headed.
Certainly, robots are becoming more “human”—from ASIMO, a robot made by the Honda Corp. that walks with cocky arm-swinging confidence; to Kismet, an MIT robot that responds to praise and scolding; to QRIO, a Sony creation that memorizes your face and greets you by name. These attributes, once characteristic only of living people, are now being replicated by machines.
Conversely, people also are becoming more machine-like. This progression began in early history; the ancient Greeks built replacements for missing or damaged body parts. Initially these were crude devices (think pirates with wooden peglegs or iron hooks for hands), but now we have a sophisticated array of devices implanted in an estimated 8–10 percent of the U. S. population, about 25 million people.
Many bionic additions are cosmetic, such as breast implants; others are functional and even lifesaving, from dental implants set into jaws to artificial legs good enough for athletic competition, to pacemakers and even artificial hearts. Even more striking are implants that connect directly with a person’s neural system or brain. Most prevalent is the cochlear implant, an electronic device inserted into the inner ear of a deaf person to restore hearing. These devices stimulate the same neurobiology as natural hearing, and about 30,000 have been implanted since 1999, some even in very young
Although not yet commercially available, more ambitious projects are in the works, such as neural implants to restore vision to blind people. The act of seeing employs millions of light-sensing rods and cones in the retina, which in turn activate the brain’s visual cortex. Researchers have inserted devices based on video cameras or solid-state light sensors into damaged retinas, and one scientist has connected a video camera directly to the visual cortex of a blind person, through an electrical socket mounted on the skull.
None of these devices has provided anything remotely as good as natural vision; however they illustrate the possibilities of brain-machine interfaces (BMI). For instance, an artificial limb could be placed under direct control of the brain, or a paralyzed person could use a BMI to control an external device such as a robot limb or a vehicle.
Researchers even have created animal “cyborgs.” In 2000, Northwestern University’s Sandro Mussa-Ivaldi removed a section of brain from a sea lamprey and connected it to light sensors and motors mounted in a small, two-wheeled mobile robot. Aided by the sensors, the piece of brain directed the robot to move either toward or away from light sources, just as it once controlled physical actions of the fish. This contrivance was a true cyborg—a living, organic brain responding to sensory data and controlling an artificial body.
Neuroscientist Miguel Nicolelis at Duke University recently connected a BMI to the brain of a living monkey, enabling the animal to move an external robotic arm by thought alone. And some researchers actually have linked a person’s brain to a machine, although not yet anything so complex as a working limb. Neurologist Phillip Kennedy of Atlanta-based Neural Signals has invented an electrode that, when inserted into a living brain, allowed a nearly completely paralyzed man to operate a computer and hence communicate with the world from which he had been utterly locked out.
Contemplating these animal-machine and human-machine combinations might make you uncomfortable. Despite the long history of replacement body parts, when the association becomes too intimate—as in linking to the brain, the seat of one’s very mind and personhood—it may seem unnatural. But compared to such wondrous outcomes as allowing a paralyzed person the ability to communicate, our discomfort becomes insignificant, does it not?
However, there are other potentially troubling issues, such as undesirable side effects. For instance, there have been reports that children with certain types of cochlear implants are at increased risk to develop meningitis. A more subtle effect is a sense of strangeness or isolation; many cochlear implantees call the quality of the sounds they hear “robotic.” Quite unexpectedly, some report that, rather than rejoining a world from which they had been excluded, they end up feeling alienated both from deaf people and those who can hear, without full membership in either community.
Then there is the question of end use. Although much current research in bionics is motivated by medical applications, a good deal of the work in the United States is sponsored by the Department of Defense, which wants to make soldiers more effective through such means as direct neural control of weaponry. A high-tech military in which a soldier’s fighting ability is multiplied through robotic or bionic means has distinct appeal—it puts machines rather than people in harm’s way—but it could also promote the dangerous belief that wars can be fought at little or no human cost.
Suppose it becomes possible to routinely integrate artificial parts into a person, and not just to replace limbs or organs. What if a person’s injured brain is repaired with a silicon chip, or his entire brain is transferred into an artificial body? Is he still a “person” in every legal, moral and spiritual sense, albeit different than before? Would we need to redefine the word “human?”
It’s never too early to start thinking about how new technology will affect people. The current debates over cloning and stem-cell research show how essential it is to understand and discuss technology that may change human expectations. I’m optimistic about where bionics will take us, because I have seen what the technology can do. I’m also influenced by RoboCop. The cyborg in that film, who began as Alex Murphy, struggles with his mixed nature. His machine attributes make him an extraordinary “good cop,” but he must also cope with the loneliness of separation from his family and from humanity.
Yet, at the end of the story, he comes to terms with his duality. A bystander asks this intimidating steel figure, “What’s your name?” RoboCop answers “Murphy,” and so proclaims his essential human spirit.
My hope is that the merging of human and machine will lead not to some kind of monstrous hybridization, but to a deeper meaning of what it means to be human.