Anything You Can Do, iCan Do Better
This article is based on a longer version published in The Tracinski Letter.
If computers and information technology are the brain of a new industrial system and computer networks like the Internet are the nervous system, then the system needs eyes and ears--and it needs legs and hands.
Today, robots are getting smaller, less expensive, more agile, more versatile--and more like humans. In the not-too-distant future, virtually any physical task that can be performed by a human will be able to be performed by a machine. And soon after that, robots will be able to do these things faster and more reliably, and to work longer hours.
In other words: Anything you can do, iCan do better.
(If you don't have Ethel Merman singing in your head right now, let me help you out with that. You can thank me later.)
One of the most important of these new robots doesn't look like a robot. It looks like your car.
By now, you may have seen the video of Google's driverless car chauffeuring a blind man around town. The fully driverless car is still a long way from making it to the road, but aspects of driverless technology have already been introduced, such as "adaptive cruise control," which automatically slows down your car when the cars in front of you slow down. The general consensus is that cars are already going driverless by degrees, introducing more of this technology a bit at a time and billing them as safety features that supplement the work of a human driver-until eventually he will have no work left.
It's not just Google that's working on this. It's GM and Audi and Mercedes and Toyota. An engineer at Ford says that "There is no technology barrier from going where we are now to the autonomous car. There are affordability issues, but the big barrier to overcome is customer acceptance."
I don't think that will be much of an issue. I spend a lot more time driving than I would like, and I would jump at the chance to get some of that time back. Imagine the other advantages: mobility for the blind and for the elderly, a decrease in accidental deaths from fatigue and intoxication, the leading causes of automobile fatalities. A drunk leaving the local bar can pour himself behind the wheel and sleep it off while the car takes him safely home.
An autonomous vehicle's driving habits can also be optimized for fuel efficiency, and the driverless car is the technology that may finally make electric cars into a practical means of transportation. The idea is that electric cars can be used as automated taxis in big cities, whisking customers around on short trips that are well within battery range, then returning to their charging stations between fares. The task of battery management, which makes long-range trips in electric vehicles so annoying, can be taken care of by machines. And surprisingly few people are discussing what might be the biggest economic application of driverless technology: long-haul trucking. Imagine long "trains" of driverless trucks that can go from East Coast to West Coast without stopping, and then turn right back around.
All told, one observer estimates that driverless car technology is going to become a multi-trillion-dollar business.
But of course, the most compelling stories are about the robots that look like us. Engineers are developing robots that can walk, climb stairs, or run really fast. Then there are humanoid robots that play soccer. It may seem like a trivial application, but what better proving ground is there for designing robots that are agile, responsive, and able to detect and react to what's going on around them? In fact, there is already a Robocup robot soccer competition.
The robots still need a fair bit of training, inspiring a hilarious video combining footage of the Robocup with some rather overheated sports announcers from a real, human soccer game.
Laugh if you like. But then again, how long did it take for a computer to beat a human at chess, or at "Jeopardy"? If those computational tasks could be overcome, the computational tasks of walking, running, and kicking a ball can be mastered, too.
These experiments are with humanoid robots, but there is a vigorous debate on the merits of legs vs. wheels. And then there are robots that do without either: the quadrotors--small helicopters with four rotors, which makes them both stable and extremely maneuverable.
Swarms of autonomous quadrotors can build walls out of blocks or build structures out of modular beams. Or they can be programmed to throw and catch poles. So far, all of this has about as much economic value as the soccer-playing robots--but they are being used to prove basic concepts for more practical applications. And then there is the Tacocopter--a fake ad for a business that supposedly allows you to order up a fast-food taco on your smartphone and have it delivered by an autonomous quadrotor drone.
Again, laugh if you like. But in ten years, this--or a self-driving car--is how your pizza delivery might arrive. It certainly has advantages over the usual method.
Some of these ideas may be, shall we say, a bit ahead of their time. But there is already a growing private market in commercial drones. One of the biggest markets is agricultural, with drones operating as crop-dusters or providing aerial views of farmers' fields. One drone entrepreneur describes the development of the business by analogy to the early days of the PC: "We're past the Apple II, and we're kind of closing in on the Mac."
And we're just getting started. Speaking of agriculture, there's the autonomous winebot built to tend to vineyards. There are dexterous robots designed "to perform such delicate tasks as assembling smartphones--something now typically done by nimble-fingered women in China." Then there are robot noodle chefs.
Much higher-skilled labor is also being replaced by machines, including robot surgeons. Oh, and did I mention IBM's Watson, which beat humans on the "Jeopardy" quiz show? Now IBM is sending Watson to medical school--appropriately enough, given its namesake--and filling its databases with the information necessary to make medical diagnoses. That sums up the progression from an expensive and elaborate toy to a powerful practical tool. If it is happening to a quiz-show computer, it will happen to soccer-playing robots and acrobatic drones.
The usual disclaimers apply. Some of these robots are more flash than technology. (The noodle chefs are really just glorified windshield wipers.) Some of them still have big technological hurdles to overcome, and many have no immediate commercial application. Then again, as with steam engines and automobiles and computers, there will also be many applications for the new robots that we cannot even imagine ahead of time.
To understand what is coming, you have to understand the implications of Moore's Law, according to which technology is "constantly doubling upon itself in everything from power to storage"
"If Moore's Law holds true over the next 25 years, the way it has held true over the last 40 years, then our chips, our computers, and, yes, our robots will be as much as a billion times more powerful than today."
At the pace of change implied by Moore's Law, the (literally) stumbling efforts of today's robot soccer athletes will quickly be solved. Robot technology that was impossible a few years ago and is inadequate now will soon become feasible, then reliable, then highly advanced--and soon robots will be doing many tasks that are no longer trusted to slow, clumsy, inconsistent humans.
About a year and a half ago, I warned that "We Are All Expendable Now," that information technology is already beginning to replace many forms of skilled white-collar labor. As information technology become more fully integrated with transportation and manufacturing, look for robots to replace many forms of skilled blue-collar labor, too.
We can debate the social and economics implications of this next stage of the Industrial Revolution--as we have ever since the days of Ned Ludd. But we can only adjust to this new wave of innovation. We cannot stop it. The technology is already here and the economic incentives for developing are too great.
So keep humming this tune in the back of your mind: Anything you can do, iCan do better.