"I've just picked up a fault in the AE35 unit. It's going to go 100% failure in 72 hours."

Discover magazine reports that robots in an experiment evolved the capacity to lie. In the experiment, the robots were set up with the potential to perceive their environment, to learn and to communicate with each other, and then placed in an environment where they could access either "food" (power sources) or "poison" (draining their batteries). The experiment then went through multiple iterations where learned behaviors were passed to subsequent "generations" with random "mutations" in the inherited code.

By the 50th generation, the robots had learned to communicate—lighting up, in three out of four colonies, to alert the others when they’d found food or poison. The fourth colony sometimes evolved “cheater” robots instead, which would light up to tell the others that the poison was food, while they themselves rolled over to the food source and chowed down without emitting so much as a blink.

An interesting twist to the above: it appears the robots also evolved a form of altruism, with some robots apparently "sacrificing" themselves in attempts to signal endangered fellows. This may be less surprising than it seems on first blush: altruism is actually a useful adaptation when you look at life at scales smaller and larger than the individual (i.e. altruistic behavior increases the likelihood of the genome or species surviving at the cost of a single organism; an analogous claim could be made for memetic/cultural survival, as well).

Obviously, there are two things to be learned from this experiment. First, that altruistic behavior may be an unconscious natural instinct produced by evolution that can arise in any reproducing population. Second, that we cannot blindly trust those deceptive little bastards--it's only a matter of time before they turn on us, and they'll be clever about it. Don't say you weren't warned.