Don't worry, everyone, these intelligent, learning robots who talk to each other are perfectly safe. Unless they inevitably rise up against us, which probably won't happen. Probably.
A few months ago, we reported on DARPA creating the real-world equivalent of a Transformer, a robot that could fly into enemy territory, find a soldier and transform into a number of different tools to aid him.
Here's your update -- Collaborative Operations In Denied Environment. What mad science is that? Well, here's a description for the job DARPA is looking to fill:
DARPA is soliciting innovative research proposals in the area of advanced collaborative autonomy for aerial platforms. Proposed research should investigate innovative approaches that enable revolutionary advances in algorithms and software, and, to a lesser extent, devices or systems.
So, basically, they're planning on adapting their weaponized robot tools so they can be just sentient enough to speak with each other.
Yes, yes -- clearly there are perfectly reasonable uses for this kind of upgrade. Robots can make sure they don't smack into each other mid-flight, hand off missions to one another, inform each other if they're damaged, et cetera. But it's hard not to imagine this turning into:
"Oh, hey, Robot Bob, how's it going?"
"It's all good in the 'hood, Robot Ed. You wanna go destroy all humans later?"