KausFiles

I Brake for Hayek?

We welcome our robot overlords, but which robot overlords? Hero Overwriter Jack Baruth explains how the debate between individual autonomy and central control is being replayed as techies try to figure out how to manage Google Cars. Except this time there are zombies (phantom individuals that may be created by hackers).

Baruth sides with the (Preprogrammed) Invisible Hand:

If you have ten thousand cars going through an intersection every day, the chances are high that one of them will misbehave, but the others should be able to compensate. A malicious “intersection controller”, on the other hand, could just stack up dead bodies like cordwood until its registers overflowed.

Hmmm. I can see where a mistake by a centralized controller would do a lot more damage than a mistake made by one Google Car. But if you were a malicious hacker … well, if you knew how the masses of Google Cars were programmed couldn’t you still make them all go over a cliff, lemming-style? Or use their commonality to subvert them all at once in other ways? Individual Google Cars are unlikely to be as differentiated as individual humans, nor as independent and resilient. That seems like a weakness (in the metaphor, and in real life). …

Update: Glenn Reynolds volunteers a tried-and-true method of avoiding a vulnerable “monoculture.” Google Cars can’t do that to themselves! Though if they did, you could stage it in Islip, New York and sell tickets. …