Bunch of Amateurs (19 page)

One of those nights we had a brief discussion of flow, the notion developed by Mihály Csíkszentmihályi (pronounced “Me-high Chick-sent-me-high”—perhaps the most fun name to say, ever). This Hungarian-American psychologist holds that there is a very satisfying state of mind that occurs when one is totally absorbed by some action. It may sound as if this is some rare state of being, like a kind of secular nirvana, but it’s not. We all experience it pretty frequently. It doesn’t require special meditative skills, just the love of doing something so that one gets lost in the labor. One might experience flow while painting a complex landscape or painting the front porch. Or chaperoning an incubator or testing electrical currents on a floor.

Its commonness is why we have so many phrases for this pleasant state of existence: being in the zone, losing ourselves in our work, being on the ball, or in the groove. According to Csíkszentmihályi (say it loud: “Chick-sent-me-high”), this state is marked by a total absorption such that one loses a sense of time and experiences a deep feeling of satisfaction.

“Oh,” Patterson said, suddenly recognizing what I was talking about. You mean “codespace”—as programmers call it—“where the world just sort of disappears.” She knew it well. “That’s a good head space to try new things, especially if there’s something I think should work. So I try it and see if it does: ‘Oh, that didn’t quite work the way it was supposed to. Let’s check a few settings and see if this works.’ Sweet.”

This was also the night that we were scheduled to meet up with some random hackers at a moveable geekfest known as SuperHappyDevHouse. Like Cowell’s crowd back in Boston, this is a gathering of mostly programmers who come together at someone’s house to exchange ideas and just generally relax. But we were physically incapable of leaving the apartment: The incubator required constant babysitting to keep it from overheating. As we sat in our own groove, occupied by long stretches of work, there appeared a solution, like a bubble slowly popping at the surface of our flow.

“What about one of those light timers you can get to fool burglars?” I asked. Patterson instantly got it. Yes. A light timer. We could program it to turn the incubator off and on. Then we could leave the apartment building. We raced downstairs and drove across the bridge to a late-night Home Depot in San Jose. The new models were perfect for what we were doing and, as Patterson would say, only a few bucks. Digital controls allowed you to customize the on/off experience however bizarrely you wanted—on for five minutes, off for thirty, for instance.

Problem solved. It was just one more jerry-rig for her duct tape and baling-wire lab. And we never did get to SuperHappyDevHouse; since dawn was on the way, we got back into our transistor flow, and there was always one more thing we had to improvise. But for now,
there was this intense pleasure. We both stepped back, looking at a light timer plugged into a surge protector with, absurdly, an exhilarating amount of self-satisfaction. Time to light up.

When I first decided to start hanging around amateur groups to see how they might be different from your average R&D outfit, I originally looked no farther than my backyard. One of the oldest amateur robot clubs in America still meets in Hartford, Connecticut, every month. So I dropped in a few times, hoping to catch a glimpse of amateurs at play and to see how their motivations might be different from the more ordered investigations of academic research aiming for prestige, or private development looking for profits. Meeting in a local high school classroom, most of the members were either enthusiastic teens or adult engineers with a serious robot jones. But it took no time at all to surmise that the enfant terrible of the club was a fourteen-year-old named Nathaniel Barshay, a small boy with a barbershop haircut and a hesitant but inviting smile.

His robots were magnitudes of complexity beyond everyone else’s, and everyone else knew it. Most roboticists fiddle with various programs to teach the machine how to walk a line or along a wall. When I first caught up with Barshay, he had purchased some new sensors and was teaching his robot how, in robot terms, to “see.” In the course of an afternoon, the crowd looked to him for advice, including the people leading the club. This particular club, the Connecticut Robotics Society, is sort of famous for producing these wunderkinds. Most of the members will tell you how DARPA, the Pentagon’s major research
arm, occasionally sends a scout to the club to raid them of their best members. Barshay is clearly DARPA-bound, should he choose. When I first chatted with him, we talked about his favorite movies, how he got interested in robots, computer games. Then I asked him what he had done last summer, when he was thirteen years old. And here was his answer: “I taught robotics at Tufts University.”

Despite being shy in the way that all young teens are around adults, it didn’t take long for Barshay to agree to let me watch him as he prepared to build a robot to enter into the Trinity College Fire Fighting Home Robot Contest (in Hartford). This is one of the oldest competitions, and Barshay had won it before.

The mission is to build a robot—most of these tend to be about the size of a toy truck—that can navigate an unfamiliar maze until it locates a candle flame and then puts it out, usually with an onboard squirt of water. That’s a good number of complicated instructions; many robots never get near the candle. But many of the successful solutions were ingenious. It was hard not to be impressed by the kid whose robot was armed with nothing more than a balloon, which, as soon as it neared the flame, burst, and, of course, blew out the candle.

Barshay didn’t win that year, but these contests were practically just signposts along the way of his constant tinkering in the basement (literally, his basement). A few months later, Barshay entered another contest. Sponsored by the iRobot Corporation, the competition asked amateurs to invent anything new, as long as you started with a stripped down version of their big commercial seller: the robot vacuum cleaner, the Roomba. In a weekend or two, Barshay installed a number of different sensors (cannibalized from his firefighting robot) and attached a long stick. The idea was to create a walking cane for the blind that would be better than both the traditional stick and guide dog combined. For instance, with Barshay’s idea, a blind person could simply ask the stick to take him to, say, the public library, and using information off the Internet and a built-in GPS and MapQuest
system, the stick would figure out the surest path and then guide the person there.

Many of the other entries in Roomba’s contest had a similarly practical cast to them—devices that would water your plants while you were away or take care of other domestic issues for the disabled.

Notice anything? As my hang time with roboticists grew, I kept encountering this. Most of the robots getting hacked together, if they weren’t just toys, were extremely
functional
—putting out fires, helping the handicapped …

How quickly we seemed to have shot past those early days of BattleBots and other super-futuristic robotic visions. To begin with, how frumpy is it that America’s premier robot is a vacuum cleaner? You hear the word “robot,” and you’re primed to expect science fiction, not a soap opera commercial. Besides the vacuum cleaner, the company’s other big seller is exclusively for the Pentagon: “the iRobot 710 Warrior,” not to mention the 510 Packbot and the 210 Negotiator (which performs surveillance in lethal areas). If you’ve got a comic book image of a fighting machine, or the pilotless drones that drop bombs in Afghanistan, think again. The Warrior can locate mines, but it resembles a dentist’s extendable light on tiny tank treads. Crucial stuff, don’t get me wrong. It’s just that when one says “war-fighting robot,” the expectation is more Sergeant York and less mechanical ET.

At first I thought this was just human nature, that all robots, like college graduates, begin with grand intentions but slowly settle into the mundane world of vacuum cleaners and office clerks (respectively). That is, until I set up a series of alerts to keep me constantly updated on robot innovations. Robot creativity seems to center on three essential sources—Japan, the United States, and Europe. The cultural differences of the three locations have heavily influenced the kinds of robots produced by both the corporations and the amateurs. America is still the land of Ben Franklin. There has always been a plain-Jane kind of pragmatism to our inventive calling. Franklin is
the guy who invented the lightning rod and the bifocals and the public library—all because, on some level, he didn’t want his house to burn down while he read his borrowed books well into middle age.

It is largely American imaginations that have turned their robots into not-all-that-glorified appliances. For the aged, we’ve invented the “intelligent walker” that responds to voice commands (“take me to the kitchen”); for construction workers, a serpentine robot that can climb to dangerous places for inspections; in hospitals, robots run medicines around from room to room while other robots down the hall are performing surgery. Even America’s most famous fictional robot, Wall-E, is a trash compactor.

Japanese robot innovations, on the other hand, are so much more about mimicking the human form. They invented the Qrio (Quest for cuRIOsity) that walks about on two legs. It’s two feet tall and looks humanoid with a head and a face. In fact, it looks like it walked right out of a sci-fi movie. Its mission, in the usual poorly translated instruction lingo: “Makes life fun, makes you happy.” You’ll never read even a superbly translated slogan like that on an American robot.

Japan also produced Murata Boy, who can ride bicycles, and Honda’s Asimo, the robot that looks like somebody decked out for a moon trip with a backpack and space helmet (not to mention five fingers that articulate). It can run across a room in perfect biped fashion.

Then there was the Aibo, the yapping, flipping robot dog. The word means “pal” in Japanese. By contrast, Boston Dynamics in the United States invented Big Dog, definitely not your pal. The promotional videos on YouTube are haunting. A headless pack animal, the robot looks like a coffin with four legs (each bent at the knee). It ambulates very fast, with the frantic, desperate gait of two guys in a horse suit trying to run without falling over. This robot can walk over most terrain and can get up from falling. When I first saw Big Dog walking, it spooked me in some primal way that I can’t describe.

But theorists of robotics can describe it. In fact, there has been a lot of research about how we humans react to artificial-life machines
as their appearance or motions get closer on some existential level to real life—to us. We are intrigued by robots for a while. Then, there is a plunge in acceptability as we become repulsed. As the robot more closely approximates lifelikeness, it becomes attractive again.

In robotics circles, this dip of disgust in the spectrum of like-ability is called
bukimi no tani
—the “uncanny valley.” The various stages down into it and back out have been analyzed and named from the most harmless, “industrial robot,” on to “android” and “moving corpse,” to “prosthetic hand,” then “handicapped person,” “bunraku puppet” (life-sized, very lifelike theater puppets manipulated in a creepy way on stage, proving that the Japanese were exploring this valley long before Karel Capek wrote his 1921 play coining the word “robot”), and then on to “unhealthy persons” and, finally, “healthy persons.” Maybe it’s no surprise, since all humans have a zombie fascination, that the moment robots begin to resemble “moving corpses,” we humans get the willies.

Americans seem to avoid it assiduously in our creations. The Japanese imagination, on the other hand, luxuriates in the uncanny valley and has practically colonized the place. One Japanese roboticist, Hiroshi Ishiguro, took it as his challenge to build an exact replica of himself so that it would be difficult to tell him apart from his robot. Staring at pictures of Ishiguro with his arm slung around the shoulder of Geminoid HI-1 definitely triggers a Rod Serling–esque creep-out. You’ve just crossed over into … the Uncanny Valley.

As a result, a lot of robot research in Japan involves skin texture, eye blinking, and the slight fidgets and twitchy movements that define the human physical existence.

An enormous number of robots invented in Japan are young girls around the barely legal age. This is creepy in a different way. At the 2005 World Expo in Tokyo one encountered the Repliee Q1, a pretty girl robot who’d made breakthroughs in blinking, gesturing, and even breathing. Also that year: Cute J-Girl, who looked like an airline stewardess from 1975. The following year, Japan gave us Actroid robot
girl, dressed sometimes in black form-fitting vinyl or the
Hello Kitty
look—although she donned a formal kimono for her press conference when she traveled to Washington, DC. In 2009, cybernetic human HRP-4C appeared at a Tokyo fashion show. With a cute Manga-style face above a robot body with unnecessarily perfect form-fitted breasts, she pouted and flirted with reporters. Given her $2 million price tag, she’s not yet a major threat to the blow-up plastic doll market.