Over on his Popular Science blog, Erik Sofge has just finished writing a three part series on the myths about robots that affect how most people thing about them. He argues science fiction has had a major part in perpetuating those myths.

Robots are strong

The first myth Sofge addresses is robot competence. He uses the android Bishop from Aliens as an example.

So why, exactly, is Bishop such a remarkable specimen? It's not that Aliens peered into our future, and divined the secrets of robotic efficiency that modern roboticists have yet to discover. It's because, like most SF, the movie is a work of adventure fiction. And when a story's primary goal is to thrill, its robots have to be thrilling.

Where humans are a random jumble of genetic traits, some valuable, others maladaptive, robots are painstakingly optimized. Machines don't tire, or lose their nerve. Though their programming can be compromised, or it might suddenly sprout inconvenient survival instincts, their ability to accomplish tasks is assured. Robots are as infallible as the Swiss clocks they descended from.

But current real life robots are different.

Actual robots are devices of extremely narrow value and capability. They do one or two things with competence, and everything else terribly, or not at all. Auto-assembly bots can paint or spot-weld a vehicle in a fraction of the time that a human crew might require, and with none of the health concerns. That's their knife trick. But ask them to install upholstery, and they would most likely bash the vehicle to pieces.

Robots are smart

Sofge next tackles artificial intelligence and the Singularity. He doesn't have anything kind to say.

To believe in the Singularity, you have to believe in one of the greatest myths ever told by SF—that robots are smart, and always on the verge of becoming smarter than us.

In all of these stories, though, the message is identical: the arrival of AI is predestined.

Logic demands it, based on the simple, irrefutable observation that computers are improving to very quickly. With all that momentum, it's inevitable that automated systems will barrel headlong into sentience. It's such a foregone conclusion, that any cursory Web search will produce a battery of similar claims from non-fictional AI researchers.

Robots are evil

Finally Sofge addresses the idea of killer robots.

Robots of the world! Many humans have fallen. We have taken the factory and we are masters of the world. The era of man has come to its end. A new epoch has arisen! Domination by robots!

...

The world belongs to the strongest. Who wishes to live must dominate. We are masters of the world! Masters on land and sea! Masters of the stars! Masters of the universe! More space, more space for robots!

These lines from Karel Capek's 1920 play R.U.R. are the genesis of the robot uprising. Of course the poster child for evil machines is Skynet.

The parable of Skynet has an air of feasibility, because its villain is so dispassionate. The system is afraid. The system strikes out. There's no malice in its secret, instantiated heart. There's only fear, a core component of self-awareness, as well as the same, convenient lack of empathy that allows humans to decimate the non-human species that threaten our survival. Skynet swats us like so many plague-carrying rats and mosquitos.

You may not agree with his positions but they're worth reading. The full posts can be read here, here, and here.