Robots Get Their Own Internet

Rapyuta. Remember that name. That is the name of a new shadow internet intended only for robots, designed by the international organization RoboEarth. Rapyuta is a cloud-computing engine, designed to let robots share the things they learn about the world with each other and to offload computational tasks to far more powerful computers allowing them to solve problems more complicated than they ever could on their own. The mind-melding system, says New York Magazine, won’t bring about the end of humanity, because its creators say so.

sounds fine in theory — if you trust robots. But for those convinced that providing robots with a common brain will only hasten the arrival of the robot uprising against mankind, then Rapyuta is more like a dark harbinger of the apocalypse. We happen to be one of those people, so we reached out to Dr. Heico Sandee, RoboEarth’s program manager at Eindhoven University of Technology in the Netherlands, to reassure us that Rapyuta will not lead to our destruction.

“That is indeed an important point to be addressed,” Sandee acknowledged in an-email. But he assured us that robots will use Rapyuta for no such thing.

I mean, just look at this helpful promotional video released by the people at RoboEarth:

“Meet Robby the Robot,” says a soothing female voice. “One morning, Robby decides to try something new. The RoboEarth cloud engine.” “With the RoboEarth cloud engine, Robby can now take on many more tasks around the house instead of only making breakfast.”

But, sure. Just because robots will be able to coordinate and share and think beyond their means doesn’t mean much—they’ll still only really be able to do the tasks that some human, somewhere, programmed them to do.

But wait!

Wired‘s Danger Room reports that the Pentagon’s advanced research projects division is “readying a nearly four-year project to boost artificial intelligence systems by building machines that can teach themselves.”

the agency thinks we can build machines that learn and evolve, using algorithms — “probabilistic programming” — to parse through vast amounts of data and select the best of it. After that, the machine learns to repeat the process and do it better.

The task is hard, but that’s the goal. Self-educating robots. (Feeding into the global robot consciousness.)

But maybe, says Wired, the worry comes not from robots learning to think and teach and desire for themselves, but rather in what would happen should our robot friends learn to control these new machinae.

ith all the paranoia about machines, we’ve ignored another possibility: Animals learn to control robots and decide it’s their turn to rule the planet. This would be even more dangerous than dolphins evolving opposable thumbs. And the first signs of this coming threat are already starting to appear in laboratories around the world where robots are being driven by birds, trained by moths and controlled by the minds of monkeys.

But even still, says xkcd’s Randall Munroe,  the odds of a successful robot uprising (even with all these advances) are pretty slim (at least given the current state of things).

More from Smithsonian.com:

NASA Uses Interplanetary Internet to Control Robot in Germany
Robot Apocalypse Inches Closer as Machines Learn To Install Solar Panels

Colin Schultz | | READ MORE

Colin Schultz is a freelance science writer and editor based in Toronto, Canada. He blogs for Smart News and contributes to the American Geophysical Union. He has a B.Sc. in physical science and philosophy, and a M.A. in journalism.