What kind of baby shower would you throw for a robot that had just given birth to triplets? What would you say to a cyborg you met on a city street? And what would you do if a prominent and entirely sane computer genius suddenly started worrying that robotic advances were the harbinger of the end of the world?
You'd better decide soon-it's already happened.
Technology defines our past: Stone Age, Bronze Age, Iron Age, Space Age. It also defines the present: Where you want to go today-and how you want to get there-depends on your access to information and transportation. Still, those who believe technology will define the future have often been brushed aside as raving lunatics, like men who trudge the streets with signs that say, "The End is Near."
Until now. Machine dreams are everywhere, from cyberpunk culture to cybernetic research done in Duke's own neurobiology labs. More than ever, the boundaries between reality and fantasy-and human and machine-are blurring. "The interface between what is human and what is machine is constantly cropping up," says Anne Allison, a Duke associate professor of cultural anthropology who currently co-teaches a class called Cyborgs. "We're seeing the humanization of the machine and the mechanization of the human. The border is becoming increasingly murky in a thousand different ways."
Welcome to the future.
Merging man and machine has always been one of the most prominent-and frightening-sci-fi scenarios out there. Allison says the scenario already exists. The title of her class doesn't refer to robots, proper. Cyborgs, unlike robots, are partially human. "You take something that has been alive, usually the brain, and you combine it with something that's been artificially created," Allison says.
We don't have resurrected robocops patrolling the streets. But depending on who you ask, cyborgs may already be part of our daily life. What, for instance, do you call a person who's dead for all practical purposes, but lives on with the help of life support machines? A person with a pacemaker? A person undergoing gene therapy? Would we really know a cyborg if we saw one? Would it matter?
In his 1999 book, "The Age of Spiritual Machines: When Computers Exceed Human Intelligence," artificial intelligence expert Ray Kurzweil argues that everyone could end up as a cyborg anyway-that humans could eventually attain near-immortality by becoming one with technology and robotics. If that sounds like a far-fetched scenario, consider that researchers at Duke are now taking the first steps toward it. Neurobiologist Miguel Nicolelis has implanted electrodes in the brains of owl monkeys and, with the electric signals generated when the monkeys reached for a morsel of food, managed to duplicate the gesture in real time with an attached robotic arm in the lab-and another one 600 miles away at the Massachusetts Institute of Technology. The New York Times called the development "a major step toward melding minds and machines," and MIT's Technology Review named the field of brain-machine interfaces as one of the 10 most promising technologies worldwide.
Nicolelis is now working on "closing the loop" by getting the robotic arm to send feedback signals to the monkey's brain, thereby fooling it into thinking the robot is really part of the monkey's body. "If
Nicolelis says it could be "at least a decade" before the technology is practically useful for humans. But by the end of those 10 years, the technology could be used to develop "neural prostheses"-robotic body parts that obey instructions directly from human patients' brains. Or, put another way, cyborgs.
As humans take on more machine-like qualities, machines are moving in the opposite direction-they're becoming more "human." True, some are little more than glorified vacuum cleaners, but others can perform heart surgery with a precision unattainable by the world's best surgeons. Scientists at Carnegie Mellon University are working on an unpiloted helicopter that can start itself, fly safely to a designated area, track a moving object and, like a migratory bird, find its way home. NASA has a self-aware spacecraft. Some new versions of ELIZA, a computer "psychiatrist" that previously could do nothing more than spit your words back at you, can now pass the Turing test, which is supposed to distinguish artificial intelligence from real smarts. (I asked one version of ELIZA whether robots would take over the world. She responded, "Do you wish that robots will take over the world?" I said no, to which she responded, "You are being a bit negative." I'll say.)
Now, for the pièce de résistance: Honda has a humanoid robot that, among other things, plays soccer. Although it may rival Ali Curtis in terms of scoring goals, the robot doesn't play for love of the game-it can't. All of these robots are ultimately dependent on humans, and they're as far from being able to feel love-or fear or hate or happiness-as an iMac.
Still, some scientists do think robots could eventually become sentient, self-aware and to some degree self-sufficient. Researchers at Brandeis University acknowledge that "to realize artificial life, full autonomy must be attained." These are the same researchers who last year enabled robots to literally evolve and replicate almost entirely without human intervention. Hod Lipson and Jordan Pollack, leaders of the GOLEM Project-named after a Jewish legend of robots run amok-designed and built a robot that can design and build other robots. The mother robot draws up hundreds of body plans dissimilar to its own and discards the ones that don't move well, à la natural selection. Then it manufactures them via a 3-D printer and waits for scientists to strap motors to its creations as per its instructions. The results are crawling baby bots made of plastic and wires.
So far, the offspring do little but squirm on the floor, and in the August Nature article announcing their success, Lipson and Pollack were careful to say that "there are many, many further steps before this technology could become dangerous." When they mentioned danger, the researchers prolonged a debate that raged for much of the 20th century. Footnoted in Lipson and Pollack's article was a piece ominously titled, "Why the Future Doesn't Need Us."
Bill Joy is the founder of Sun Microsystems and one of the Internet's original developers. He is the man behind Java, Jini and Unix. His credentials are spotless. But last April, when he penned "Why the Future Doesn't Need Us" for Wired magazine, he started sounding like a ranting futurist.
Imagining an Armageddon driven by machines, Joy wrote that the triple threats of genetics, nanotechnology and robotics, if put in the wrong hands, could be "so powerful that they can spawn whole new classes of accidents and abuses.... I think it is no exaggeration to say we are on the cusp of the further perfection of extreme evil."
The article fractured the scientific community. Within hours, readers flooded the Wired e-mailbox, and the magazine eventually went so far as to establish an e-mail address solely for correspondence about Joy's manifesto. Editor-in-chief Katrina Heron wrote that "many readers express immense gratitude to Joy for giving their fears and anxieties a voice; others object to certain passages; still others disagree with almost everything he wrote."
"Why the Future Doesn't Need Us," still at www.wired.com/wired/archive/8.04, became a touchstone for dissidents and doomsayers alike. They started gathering for conferences, forming committees, stoking debates. Scientific fact was looking more and more like science fiction.
What Joy imagines as a reality has already been imagined time and time again as a fantasy by science fiction writers. Our love/hate relationship with robots stems partially from the fear that machines could be both sentient and soulless-a fear that has been aggravated and exaggerated by science fiction. If machines become sentient, the argument runs, they will demand freedom. But that's not in line with what we demand from them.
The word "robot" derives from the Czech word for "forced labor" and was first used in Czech author Karel Capek's 1921 play, "Rossum's Universal Robots." In it, the inventor Rossum strips humans of everything unrelated to work-i.e., everything that makes them human-resulting in thousands of robots produced as substitutes for workers. When Helena, one of the robots, spontaneously develops a soul, other robots follow her lead. They eventually rebel, killing almost everyone in the factory and proclaiming that the world should be populated by "nothing but life! nothing but robots!"
Helena served as Rosemary to the hundreds of sci-fi demon babies we know and love, or hate, today. And like Rossum's robots, most of these modern, fictional sentient machines-HAL in "2001," the Borgs of "Star Trek," the nameless monarchs of "The Matrix"-are harmless only when they are tamed and under human control. The sentient, self-aware and self-sufficient robot who also bears no ill will toward humanity is a rare creature indeed.
Capek was no scientist-his robots sprung from chemical vats, not mechanical labs. He was also no visionary. In 1935, he wrote with horror of the possibility that "metal contraptions could replace human beings, and that by means of wires they could awaken something like life, love or rebellion." At best, he feared, machines would enslave us. At worst, they would exterminate us.
In 1935, Capek sounded like a worrywart. Now he sounds a lot like Bill Joy.
Regardless of how much we wring our hands, or how silly we think futurist fears really are, the future itself will march on. But where will it go? Joy's three threats may end up not
Get The Chronicle straight to your inbox
Signup for our weekly newsletter. Cancel at any time.