Jay Richards, a very rational guy, wrote an excellent article about why we shouldn’t fear our robot overlords. I’ll share a small sample with you here but strongly suggest you read the whole thing.
In a test round of “Jeopardy!,” for instance, the host gave this answer: “Barack’s Andean pack animals.” Watson came up with the right question almost instantly: “What is Obama’s llamas?” We’re getting a glimmer of the day when a computer could pass the “Turing Test,” that is, when an interrogating judge won’t be able to distinguish between a computer and a human being hidden behind a curtain.
Artificial intelligence gives lots of people the creeps. When I tell friends and family about Watson, most of them think of Terminator or The Matrix. They see Watson’s victory as a portent of some future cataclysm, when machines will take over the world and reduce human beings to slavery. Maybe everyone I interact with has become a Luddite, but that seems unlikely. I live in Seattle, after all.
As it happens, this fear of technology by the tech-savvy is quite common. In 1998, inventor and futurist Ray Kurzweil described the coming age of “spiritual machines” at a Telecosm Conference sponsored by George Gilder and Forbes Magazine. Kurzweil’s vision of man-machine hybrids, conscious computers, and human beings casting off our fleshy hardware for something more permanent elicited a variety of responses, including one by Bill Joy of Sun Microsystems. Joy penned a famous piece for Wired magazine in which he called for government to limit research on the so-called “GNR” technologies (genetics, nanotechnology, and robotics). These were the most ethically troubling technologies because, in Joy’s opinion, they were most likely to open Pandora’s box. Joy, who had enjoyed decades of unfettered research and entrepreneurial creativity, had now fingered the true enemy of humanity: the free market.
Talk about an overreaction. Still, part of the blame must rest with AI enthusiasts, who aren’t always careful to keep separate issues, well, separate. Too often, they indulge in utopian dreams, make unjustifiable logical leaps, and smuggle in questionable philosophical assumptions. As a result, they not only invite dystopian reactions, they prevent ordinary people from welcoming rather than fearing our technological future.
Yes, I know, until just now you thought GNR stood for Guns and Roses. Which, sadly, may serve to reinforce the point here today.
One of the problems with the whole idea of robot overlords is that robots are, basically, computers. And computers are limited by the fact that they do not have quantum thinking capabilities. They are either on or off, yes or no. That is, until now. Alex Knapp, another really smart guy who works at Forbes, says that some scientists seem to have cleared that hurdle.
One of the primary goals of quantum computing research is the development of a consistent “quantum speedup” — a process that, in MIT Professor Scott Aaronson’s words, means to “solve some actual computational problem faster using quantum coherence.” In order to achieve such a speedup, it’s necessary to take advantage of the ability of qubits (the basic unit of information in quantum computing) to exhibit quantum entanglement. Quantum entanglement allows qubits to exhibit multiple states — enabling faster calculations than traditional bits, which can only exhibit one state at a time. Such entanglement has been demonstrated on a small scale in superconducting circuits by the Schoelkopf Lab at Yale, which last year published a paper demonstrating three qubit entanglement.
What’s needed to build on this work is a much bigger scale of entangled qubits. And that scale may be possible soon, thanks to some important work by physicist Olivier Pfister and his team at the University of Virginia. Their research, which was published in Physical Review Letters describes the team’s ability to entangle cluster states of Qmodes. Qmodes are part of a quantum computing architecture whereby the normal modes of light are actually used as qubits to perform quantum computing operations.
In this set of experiments, the Qmodes were generated as lasers emitted by a optical parametric oscillator. The qmodes were forced by the oscilaltor to create what’s know as an optical frequency comb. This resulted in a series of Qmodes that were separated by known frequencies, and related to each other based on their phase. Using this method, Pfister and his team were able to entangle 15 cluster states of 4 Qmodes each, for a total of 60. The team ascertained that all 60 Qmodes were equally entangled.
This is an exciting step forward in quantum computing, but there are a couple of caveats. First of all, this is miles from the thousands of entangled qubits necessary to achieve quantum speedups. This seems like a pretty scalable solution, but that remains to be demonstrated. Moreover, although the authors state that “[t]here is no known fundamental impossibility to the implementation of quantum computing with Qmodes”, there are some special challenges when it comes to entangling qubits optically as opposed to entangling them in a superconductor or other quantum computing method. So it may turn out that this is scalable, but not economical or practical. There’s still a lot of work to do.
No, qubits are not anything like Q-bert. The fact that you thought of that would, again, seem to reinforce the point.
But robots will need more than just the ability to process data if they are going to overthrow the world. Or, more likely, just ask us to lie down and have our tummies rubbed while they do the real work. The nice people at The Telegraph (UK) tell us not to worry. Extremely functional robots are just waiting for their new super brains.
The event (Shanghai International Conference on Robotics and Automation in China), hosted by the Institute of Electrical and Electronics Engineers (IEEE), attracted more than 1,700 engineers, academics and businessmen from around the globe to show off inventions.
On display were robots who were able to perform range of tasks, from writing calligraphy to serving food.
Chinese engineers showed a deep-sea remotely-operated vehicle that could dive into depths of 3,500 metres (11,483 feet) to collect samples and announced plans to use robots to assist in space exploration to Mars.
The president of IEEE Robotics and Automation Society, Kazuhiro Kosuge, said he saw central to the role of robots was improving human society.
“The best robot is perhaps a robot that can serve us like a human does. To do so, the robot has to know what you want, how you want to be helped and how you want to be assisted,” said Kosuge.
“The robot has to estimate what you are trying to do. So we have to develop a lot of technology with which we can communicate to the robot, and so that it can communicate with the [human]. That is the most challenging issue we have to solve from now.”
Oh joy. When, in history, has a slave class of people not revolted? That would be “never” for those of you who slept through third grade. Yet isn’t that exactly what these people are trying to create? An electronic underclass designed to serve man.
In Munich they are developing robots that can cook and serve a meal. In New Jersey, far from the brain dead rantings of JWOWW and her ilk, scientists have created a sex bot that could, with minor alterations, run Human Resources for any large company.
And, no, that is not a proper definition of irony.
Well, wait, actually it is.
And if you think they’ll need to keep humans around for entertainment or sport, the participants at this year’s World Robocup Soccer Championships say you’re wrong.
Once they get those super brains they won’t need us at all.
And if they don’t need us, what’s the point of keeping us around? Evolution would seem to demand that we go the way of the Dodo.
Listen to Bill McCormick on WBIG AM 1280, every Thursday morning around 9:10!