Chat follow-up: Humanoid robots

The live chat of the O’Reilly webinar that Christopher delivered on 18 April 2013 had some great questions, but not all of them made it out of the chat room and onto the air. I’m taking a short break from the release of the sci-fi survey to answer some of those questions.

Q: Adrian Warman asks: Humanoid robots (android) are not as efficient mechanically, yet we ‘prefer’ them (C3PO v. R2D2?) Will our preferences always override efficiency?


A: I think it depends on the context of use. Humans are good at humans. So, when robots have social functions, it’s best that they appear humanoid, while avoiding the Uncanny Valley (or see page 183 in the book for more). They should stick to the Canny Rise, to coin a term. When they need to do other, non-social things for us, like build cars, or vaccuum our floors, or mine for rare earths in asteroid belts, they should be fit to task.

Giving credit where credit is due, this is exactly the case with the Star Wars robots. C-3PO’s a protocol droid, for “human-cyborg relations” and R2-D2 is ostensibly an astromech maintenance droid. Their appearances match their functions.

Leave a Reply