Artificial intelligence has a long way to go before computers are as intelligent as humans. But progress is happening rapidly, in everything from logical reasoning to facial and speech recognition. With steady improvements in memory, processing power, and programming, the question isn't if a computer will ever be as smart as a human, but only how long it will take. And once computers are as smart as people, they'll keep getting smarter, in short order become much, much smarter than people. When artificial intelligence (AI) becomes artificial superintelligence (ASI), the real problems begin.
If you want a sense of where the nation’s job market is headed, a good place to stand is inside the half-mile-long Skechers warehouse in Moreno Valley, California, where box after box of shoes is stacked upon row after row of shelving, which soars some 40 feet in the air. Physically, the place is a wonder—quiet, sleek, and environmentally friendly (at 1.8 million square feet, it’s the largest officially certified “LEED Gold” building in the country). But what’s most remarkable about the $250 million structure, which opened in 2011, is how few people work there.
Let's face it, we all need a break from talking about this god-awful shutdown (acknowledging, of course, that the best break of all would be to end the damn thing). In that spirit, via Technology Review, here's an interesting study out of Georgia Tech about what kind of robots young and old people are more comfortable with, and how those preferences change depending on what it is we're asking the robots to help us with. Generally, the older people preferred more human-looking robots, while the younger people preferred more, well, robot-looking robots. That would make sense if you assume that the young are more comfortable with technology. But things get interesting when you get into details about what the robots are doing:
Baxter, a friendly robot colleague who'd love to hang out with you after work. (Flickr/Steve Jurvetson)
As robots move into more and more workplaces in the coming decades—not just high-tech manufacturing but eventually everything from hospitals to supermarkets—one of the big challenges employers will face is making their carbon-based workforce comfortable with the new arrivals. That's the topic of an interesting story in The Economist (h/t Kevin Drum) that focuses not just on the technology but on how the robots make us feel, and what must be done to keep people from freaking out when they find out their new partner is made of metal and plastic. It seems that the psychology of human-robot interaction is going to be a burgeoning field in the next few years:
It's Friday, which of course means we have to talk robots. Yesterday, the Defense Advanced Research Projects Administration (DARPA) unveiled Atlas, a humanoid robot it's using as part of its robotics challenge, in which teams of engineers will compete to write software that best employs Atlas' human-eradicating capabilities. Kidding, of course—they'll actually be trying to perform a series of tasks that might be needed in a disaster scenario.
Frankly, I've always been skeptical about the potential of humanoid robots. Sure, it helps us to relate to them if they look like us, but the human body has a lot of limitations. For instance, hands are great, but should a robot have only two? Why not four or six, or eleven? The more hands, the more things you can do with them. And legs are extremely useful, especially for navigating uneven environments where wheels won't work well, like the rubble of a building that has fallen over, or the stairs in your house. But are two legs better than three or four?
Robots: Not just for nuns anymore. (Photo from Paro Robots U.S., Inc.)
In the movie "Castaway," Tom Hanks' character, stranded on an island with no human companionship, dresses up a volleyball to look vaguely like a person's head, gives it a name ("Wilson"), then spends years having conversations with it. Near the end of the film, as Hanks is making his desperate attempt to return to civilization on a raft, Wilson gets washed overboard. There's a poignant moment when Hanks tries to reach Wilson, who is drifting away from the raft, then realizes sadly that he'll have to let it go if he's going to save himself. Because no matter how much emotion he's invested it with, in the end it's just a volleyball.
Here in the actual world, there are lots of people who go through their days lacking companionship, many of whom live in nursing homes. As the Baby Boom generation ages, there are going to be a lot more of them. Which naturally leads to the question: Can we use robots to make their lives a little less miserable? Slate's Future Tense brings us the not-really-surprising (at least to me) results of a small pilot study where a group of nursing home residents were each given a Paro robot, which is a baby harp seal stuffed animal that has some sensors and actuators and responds to your touch. Here's what happened:
One of Google's self-driving cars. (Flickr/Guillermo Esteves)
Futurists have been predicting self-driving cars for decades, but for a long time it wasn't because the idea was a natural extrapolation of existing technology. Instead, from the standpoint of the 1950s or so, it just seemed like something we'd have in The Future, along with robot maids, vacations on the moon, and a spectacular network of vacuum tubes in every home. Today, almost all the technology necessary to allow cars to drive themselves is either already in existence or in the development process, and Google has already allowed its driverless cars to go hundreds of thousands of miles on their own. So the Department of Transportation has issued a policy statement laying out some of the issues that are likely to be confronted as these technologies develop, and establishing its research agenda to address the questions they'll need to answer in order to properly regulate driverless cars.
Robots, as yet unarmed, created for the military by Boston Dynamics.
Last week, Human Rights Watch released a report raising alarms about the specter of "killer robots." The report urged that we develop an international treaty to prohibit the development of fully autonomous robotic weapons systems that can make their own decisions about when to use deadly force. So is that day coming any time soon? The Pentagon wants everyone to know it has no plans to allow robots to make decisions on when to fire weapons; Spencer Ackerman at Wired points us to this memo from Deputy Secretary of Defense Ashton Carter released two days after the HRW report, making clear that the DoD's policy is that robots don't get to pull the trigger without a human being making the decision (or in bureacratic-speak, "Autonomous and semi-autonomous weapon systems shall be designed to allow commanders and operators to exercise appropriate levels of human judgment over the use of force"). It seems obvious that we don't want a bunch of Terminators walking through our streets deciding whom they're going to shoot. Or is it?
As exciting as it is to watch Olympic sprinters tear down the track, the truth is that running fast for short distances is just not really human beings' thing. Usain Bolt, the fastest human ever to walk the planet, has reached a top speed of 27.78 miles per hour, which is an amble to a cheetah or a gazelle. Heck, your dachshund can almost certainly outrun you, even with its stubby little legs. What gave our ancestors an evolutionary advantage was their stamina, the ability to chase down prey by running and running until the poor wildebeest ran out steam and dropped.
The bright side of this story is that in the not-too-distant future, robots will be able to hunt and capture your slowpoke self without too much trouble, should the authorities determine that you have a suspicious bulge in your pocket or you need to be punished for jaywalking. Boston Dynamics, a robotics company that uses your tax dollars (in the form of grants from DARPA, the Defense Advanced Research Projects Agency) to make crazy advanced robots that mimic animals in all kinds of interesting ways, just announced that their Cheetah robot has reached a speed of 28.3 miles per hour, faster than any human. What's different about it is that it has legs, not wheels, potentially making it (and robots like it) highly maneuverable. Gaze upon it and tremble:
As someone who has gone on record in support of driverless cars, I simply must raise my voice in objection to this ad targeting Florida state rep. Jeff Brandes, who is running for state senate. An inconsequential local race, you say? Not when this kind of vicious anti-technological filth is sent out to paralyze our nation's seniors with fear of walking the streets! If you think American politics is no fun, just take a gander:
Seemingly intuiting my desire for a quick diversion from politics into a more important topic, Kevin Drum links to this post by Stuart Staniford discussing the day, not long in coming, when Planet Earth's robots outnumber its humans, including a semi-serious projection that shows Them outnumbering Us some time in the early 2030s. Should we be worried? Well, yeah, but not because they're going to kill us all. The problem is capitalism.
The geek superstars at the University of Pennsylvania's GRASP (General Robotics, Automation, Sensing, & Perception) lab have taught their nano quadrotors—and if you know robots, you know that UPenn's are among the coolest of quadcopters—to play the James Bond theme. My judgment that this is awesome is unaffected by the fact that I went to grad school at UPenn. Just watch until the end for the guitar:
Google, we learn from Monday’s New York Times, has a secret lab in an undisclosed location in the Bay Area where it is developing robots. We don’t know what the Google-oids are working on there, but we do know that the company has developed and built a driverless car that has already traversed 100,000 miles on California roads without getting either a ticket or a scratch.
If robots are ever going to do all our housework and pick up all our garbage, they're going to need to improve their dexterity. The human hand is a marvel of coordination, strength, and subtlety. Figuring out how to produce robot hands that can do anything like what a human hand can do has been a famously difficult engineering challenge.
But the Defense Advanced Research Projects Agency (DARPA) has been working on it, and now some scientists working with a DARPA grant have come up with a marvelously low-tech and creative solution, as Popular Science tells us: