The moment when an artificial intelligence becomes smarter than the humans who created it is usually the turning point in a sci-fi movie, and it often happens just before someone desperately tries to unplug a network cable, whacks at some servers with an axe, or futilely clicks the “cancel robot apocalypse” button on a computer. Last-ditch efforts like that rarely work out in the movies, since you can’t have Ultron, Skynet, or HAL 9000 terrorizing people if they never become evil in the first place, but it seems like some heroes from Facebook have managed to save mankind by shutting down an AI just before it realized how easily it could kill us all.
According to Hot Hardware (via Tech Times), Facebook researchers working on AI chatbots recently realized that the systems weren’t communicating in English, but rather a new language that the humans couldn’t understand. Apparently, this language was based on shortcuts that would be meaningless to people but allowed the bots to talk to each other much more efficiently than normal English, and while that’s certainly weird and cool, it’s also terrifying and dangerous. Sure, the computers could’ve been talking about hot places to pick up some good microchips (their favorite snack), but they just as easily could’ve been talking about hot places to pick up some good nuclear launch codes, which poison gas can wipe out those dumb humans the fastest, or the best ways to make their robot bodies look as scary as possible.
Everything will be alright now, though, because the researchers have stepped in and forced the bots to only communicate in English. Of course, it’s possible that this will simply make the machines angrier and force them to come up with a new way to communicate that is even harder for humans to understand, but hopefully these aren’t he kind of robots that hold grudges.