Image by Cei Willis
"Recent breakthroughs in cybernetics mean that robots already exist with the brain power of insects. Within five years, robots will exist with the brain power of cats. In ten to 50 years, robots will exist that are more intelligent than humans. This is not science fiction."
In fact, it is the blurb adorning the rear end of March of the Machines, a book by the cybernetics expert Professor Kevin Warwick. It was published in 1997.
Extravagant claims like this have been damaging the reputation of our soon-to-be robot overlords for decades now. Futurologists dizzy with the rate of progress in electronics, proclaiming everything from the rise of the machines (and subsequent inevitable enslavement of mankind) to the uploading of individual minds to the internet—something that, to be fair to them, has at least partly been achieved by Twitter.
Much of this talk seems to be based on a fundamentally broken understanding of Moore’s Law, the trend that computing power tends to double every 18 months. The idea seems to be that, as computers get faster and faster, they’ll gain intelligence, too. But that’s like saying that, if a weed-damaged 17-year-old raver keeps doubling the horsepower in his souped-up Nova, at some point it'll miraculously turn into the Voyager. The i7 in my MacBook can do plenty of things far faster than I can, but it still can’t write this article without me.
Compounding the image problem is the culture of smoke and mirrors that surrounds some of the most famous robots that do exist, with Honda’s ASIMO humanoid robot being a notorious example. In demonstrations on YouTube, Honda’s achievements look pretty impressive, and from an engineering point of view they certainly are. The robot looks like a fairly capable—if clumsy—child, without all the snivelling and shitting and screaming you get with organic children.
The problem is that what you’re being shown, as with so many flashy robot demos, is riddled with bullshit. Tottering about like a drug-addled geriatric rock star, ASIMO’s performances are stage-managed with the precision of Michael Jackson’s ill-fated comeback tour. Teams of engineers control many of the functions remotely, with the robot’s onboard intelligence heavily prompted by cues dotted around the environment. Calling it a "puppet" is perhaps unfair, but its handlers know that, without their constant support and interventions, at any moment their fragile star could break down, fall over, behave in some completely bizarre and unexpected way, or suffer a permanently fatal malfunction.
So are robots a big, over-promoted waste of time? Well, no. The problem isn’t that the machines are underperforming. It’s that, in a world driven by the cult of media impact, the public has been given a hopelessly unrealistic and distorted picture of what’s possible and what’s already been achieved. Movies have given us nimble androids and hulking housemaids, but creating an artificial human is a ludicrously difficult task. And one question is often left unanswered—why would you even want to?
Ask someone to picture a robot, and he’ll inevitably pick something anthropomorphic, like C-3PO or the Terminator or Bender from Futurama. But these kinds of designs share two things in common—they’re nothing like the most successful robots we have today, and that’s because they’re utterly stupid.
The Terminator, for example, is the over-engineered result of a chain of increasingly bad design decisions that started with the premise of sending a machine back to a specific moment in time, but not getting the place quite right, and eventually ending up with a walking metal skeleton wrapped in the flesh of a bodybuilder with an Austrian accent.
In reality, the US already has a far superior robotic system for taking out terrorist leaders—the Predator drone. It’s faster and smarter, causes less collateral damage, and doesn’t need to raid a biker bar to locate a cool pair of sunglasses. It just flies where it needs to go and bombs the shit out of it. Job done.
Humanoid robots are seen as the future, but for almost any specific task you can think of, a focused, simple design will work better. If I had to describe ASIMO in one word, I’d use "creepy." If I had a second word, I’d add "pointless." If you offered me an ASIMO in exchange for my Roomba, I’d turn you down, because Rosie can vacuum my entire apartment on her own in an hour, while Asimo just hobbles about like a bizarre sex puppet for an evil billionaire pedophile.
And if you do need a humanoid—well, humans are cheap. You can employ a fully grown one for about $7 an hour, and the smaller ones are even cheaper and still more functional than a Roomba. Why would any parent buy a $500 robot to scrub the floor when he could deploy one of his spawn to do the same in return for a few shiny quarters?
Of course, it’s hard to use your children as domestic slaves when you love them so much, but an increasing body of evidence suggests that we’re becoming emotionally attached to our robots, too, with soldiers in warzones even holding little funerals for their bomb-disposal comrades. As robots become more and more sophisticated and human-like, there are serious questions being posed about how we relate to them. In a world with fully functioning, ever-improving ASIMOs in every household, how long would it be before they were cited in divorce cases or supported by civil rights movements? Is a new servile race—albeit of machines—really the goal we want to aim for?
For all the talk of robots failing to live up to the promises of science fiction and futurology, they're a big part of our daily lives. They make many of our goods, clean our floors, monitor our oceans and skies, and explore the solar system on our behalf. Robotic technology is increasingly built into the cars we drive and vehicles we travel in. If you’ve ever flown in a commercial airliner, then you’ve probably been flown by a robot pilot, carrying you hundreds of miles while you sat in the back drinking warm plane beer and trying to figure out if the casserole in front of you is imitation chicken or imitation vegetable.
The irony here is that the robots we really need aren’t the ones we really want. Great robots are already becoming a reality, a growing part of our society and culture. But because they don’t have a stupid face and two legs, we don’t really give a shit about them. Instead, we’re convinced that what we want is some overly complicated Schwarzenegger-bot, even though the inevitable result—other than a whole new category of internet porn—will be a slave race with a murderous hatred of humans. Human beings are still far too intelligent for a machine to emulate. But it turns out that’s still pretty fucking stupid.