Thursday, February 24, 2011
Will robots take over the world?
In London. To a UCL lunchtime lecture entitled ‘Will robots take over the world?’
Kathleen Richardson says no. She’s an anthropologist. Who’s better placed to answer such a question? An anthropologist or a roboticist? Surely the answer’s obvious! Or is it?
Another thought. Computers habitually do things they shouldn’t. So even if we programme them with Asimov’s 3 laws of robotics (of which number 1 is harm no humans) might they flout the laws?
Must see 2001 A Space Odyssey again. Last watched it in the 1970’s I suppose (it came out in 1968 – blimey!). Did HAL the rougue computer flout the Asimov laws? Or did HAL conform to the laws in ways the humans failed to foresee? Here's the offcial trailer to the film, 3:32 mins. For the famous sequence "open the pod bay doors HAL", jump to 1:45.
On the question of computers doing things they shouldn’t. You can say that, when you look into it, it’s always a human’s fault. But I don’t think that alters the case. Then of course if artificial intelligence is possible (which is disputed but say it is) then artificial malevolence is possible. And then there’s this: there are already malevolent robots, designed to be so. They’re called military robots. Yes they have built in protocols. But this brings us back to computers doing things they shouldn’t …
Lastly a bit of history. Čapek conceived the first robots in 1921, as humans, assembled from parts, like a Ford car. So if you’ve read my post Robot : a Czech Serf, I got that wrong. The metal man I've illustrated came a few years later. More of this when I’ve made sense of my notes.