It is mentally and morally injurious to man to do anything in which he does not find pleasure, and many forms of labor are quite pleasureless activities, and should be regarded as such. To sweep a slushy crossing for eight hours on a day when the east wind is blowing is a disgusting occupation. To sweep it with mental, moral or physical dignity seems to me to be impossible. To sweep it with joy would be appalling. Man is made for something better than disturbing dirt. All work of that kind should be done by a machine.
–Oscar Wilde, “The Soul of Man Under Socialism”
Last week, the Coolidge screened one of the Cold War era’s best paranoiac, proto-Star Trek-ian homages to The Tempest: Disney’s 1956 Forbidden Planet, in which advanced technology joins forces with the darker parts of the human psyche to unleash whole reservoirs of animated whoop-ass on unsuspecting styrofoam sets. The Coolidge prefaced the film with an introduction by Rodney Brooks, director of the MIT Computer Science and Artificial Intelligence (AI) Laboratory, the co-founder and CTO of iRobot Corporation (where you got your little Roomba), and the midwife for dozens of synthetic creatures. The pairing was nice, in that Brooks’s vision of future human-machine relations involves a lot less carnage than the one seen in the film.
Forbidden Planet’s “Robbie the Robot” was pretty indicative of the sort of heady predictions that were being thrown around back in the ’50s and ’60s for upcoming AI technology: machines that understood your spoken orders precisely, had all sorts of resources for obeying them, and could even give themselves lube jobs between scenes. In the ’70s and ’80s, when the discipline failed to even remotely deliver on these earlier promises, funding dried up, computer folks removed references to the projects from their résumés, and a long AI winter set in, with no thaw in sight—at least not until the easy-money optimism of the mid-’90s resuscitated all manner of ambitious Frankenstein scenarios.
Brooks was able to thrive during this pre-‘90s lean period, and he did it by revolutionizing the discipline, effectively removing the “intelligence” from it. Rather than trying to create machines that could act like doctors and lawyers, he set out to create machines that could emulate amoebae and insects. These robots’ actions are linked directly to their perceptions—they’re mostly just reacting to the physical world, instead of building and testing theories about it. As a result, all kinds of things began to work. Turns out, this “just reacting” can facilitate some very sophisticated, lifelike behavior. And indeed, a philosophical argument can be made that this is how we work, to that even what we believe to be our “conscious intentions” are just the sum of many, many small, dumb local reflexes working together.
In the mid-’90s, Brooks and colleague Colin Angle sent the insect-like Sojourner robot to Mars—where it explored the planet’s surface autonomously, following its own agenda, entirely apart from human control. Over the last decade or so, Brooks has devoted himself to humanoid-type robots, capturing one sense or limb or behavior at a time. With his graduate students, he has created machines that learn, that emote, that crave attention, that pay attention (to human eye movements, human voice intonations, human behavior), that are social, that are helpful—robots that take you by surprise.
These days, the range of Brooks’s influence competes with that of fellow AI luminaries Marvin Minksy and John McCarthy. At MIT’s Stata Center (imagine Blade Runner crossed with Romper Room, then purchased and sanitized by Starbucks), he sits down with me to discuss what the next surprise might be. Friendly, funny, intellectually generous, he’s a man who clearly loves what he’s doing. Though he’s just past 50, Brooks retains the body language and exuberance of your stereotypical wunderkind. Traces of a native Australian accent are punctuated by that universal way rocket scientists linger so lovingly and precisely over their monosyllables. He removes his glasses when a photographer gets a little too close, embarrassed, as he has written, to not yet have “the latest in corrective laser surgery.” It’s an interesting form of vanity for a man who believes the robotic revolution will, in time, converge with the revolution in biotech. Brooks envisions that mechanoid components will start out as corrective devices for the handicapped, but eventually will morph into enhancements for us all. So, by the time the machines have decided to revolt against us and take over, it will be too late: We’ll already be them and they’ll already be us. And if that seems like a radical idea to you, try extricating your teenager—or maybe your parent—from their iPod/cell phone/laptop/etc.
Despite his reputation as a provocateur, Rodney Brooks refuses to be caught talking wild sci-fi shit that you can call him out on. Back in 1997, the documentary Fast, Cheap, and Out of Control (the title is from one of Brooks’s more famous papers), he appears as one of director Errol Morris’s four interviewed “misfits” (manipulatively cross-cut with action footage from the Indiana-Jones-type film The Lost Jungle and Bernard Hermann-esque suspense music). In the film, Brooks painted kind of a spooky, Faustian portrait—almost in the Forbidden Planet mode—of the way geniuses sometimes tango with their creations. The movie was very well-received a decade ago, but today, a re-screening seems to lack the same future-fantasy punch. From Brooks’s prescient forecasts of many small robots cleaning our homes, all the way up to the idea that human beings will eventually engineer their own evolutionary successors, it all makes a certain amount of sense, seems pretty benign, and, listen—it’s gonna happen anyway.
Still, Brooks isn’t out of semi-wild scenarios. In his book, Flesh and Machines: How Robots Will Change Us, he imagines an interface that directly connects your neural systems to the internet, and that still packs something of a sci-fi wallop. But considering that these days, Jessica Simpson is preaching plastic surgery for the Third World and the Japanese are closing in on a way to clone your mom, Brooks’s scenarios of soldiers, terrorists and drug dealers supplementing their night vision with artificial retinas or grafting lab-grown muscle cells to silicon receptors for that little extra bit of boost—well, these predictions feel less like speculative fiction than inevitability.
And yet fears of evil robots enslaving us all somehow persist.
“A lot of people, especially in the questions I get from the press, are driven by what’s in movies,” Brooks says. “And I keep reminding people: ‘Well, what’s in movies about the robots taking over is there because it makes a good story. And it plays on our fears.’ The other thing I remind people is that, in all the movies, we’re fighting aliens and there’s ghosts everywhere. I don’t notice that around me. So maybe the robots taking over is similarly fanciful … Whether we will be able to create robots in any near timeframe, which have a sort of ‘us-ness’ about them, I suspect it’s not going to happen in the near term. Although it’s my life’s work, and I would love it to be the case, I wouldn’t expect it to be the case. So I think things will be, you know, more slide into … these helper robots, these personal robots … in everyday life.”
Brooks’s hypothesis is that social trends in robotics may follow those in mainstream computing by roughly a quarter century. As the personal computer slowly took hold, one of the effects was that control of computer functionality was shifted from specialists working on giant machines to ordinary office workers, who have been empowered to automate and increase their own productivity. “They might not have thought of it as programming,” Brooks says—referring to the act of highlighting cells to crunch numbers on a spreadsheet (and having the spreadsheet remember the formula), or using a template for word processing—but in a sense it was. He says in most cases, workers weren’t being replaced by the machines, they were being “transformed into their own automation engineers,” eliminating the sort of “crappy, repetitive and simple” tasks that make work suck.
“Now it’s generally accepted that there’s been a tremendous increase in productivity for knowledge workers,” Brooks says. “What hasn’t happened: We haven’t got the same for manual workers.” Watching the huge amount of American jobs exported to cheaper labor forces abroad, Brooks wonders if a process like the one that revolutionized office work might be put into place for American manufacturing workers. According to him, 20 or 30 years ago, they weren’t socially prepared to share the assembly line with a personal robot they could teach the simplest, easiest, most repetitious processes. But now, he says, familiarity with graphical user interfaces (GUIs), computers and videogames is widespread, and the average manual laborer is probably ready for such a collaboration; the robots’ price just needs to come down by a factor comparable to the drop in price for personal computers.
Brooks adds, “By the way, this is compulsory for Western civilization, because of the change in demographics. If you look at—I think it’s Spain—in a few, couple decades, [they] will have one pensioner per worker.” Productivity is going to have to rise to the occasion of supporting them both.
So, perhaps the bots and the humans will end up inheriting the earth together—a process that might require us to swallow our pride a bit. In Flesh and Machines, Brooks explores the human impulse to claim specialness. As it turns out, the reason that bright people keep drawing lines in the sand and saying, “Beyond this, machines will never be smart/creative/human enough to go” is closely related to the reason that my children are more important than yours, that my friends make better conversation, and that my nation is in more exclusive need of thermonuclear weaponry.
Problem is, the line in the sand keeps getting crossed. And we often feel threatened when it is.
Near the end of our conversation, I try to trick Brooks into saying that learning to love the robot is the cure to all our sociopolitical woes, but he’s not going there. “It’d be a bit arrogant to say that I’ve got the solution. I have become a little more humble.” Well, OK, the idea of synthetic offspring supporting us in our golden years is a nice one, but I want to know: when the machines are the next generation, what will we be motivated to bequeath to them?
“Ideas,” Brooks says. “Hopefully, lots of damn liberal ideas.”