Facebook researchers shut down artificial intelligence programs that talked to each other in code.

HAL doesn't mean any harm. Really.

You know how Hollywood always creates some movie regarding artificial intelligence, where a robot or software starts becoming a little too smart for its own good? Usually chaos ensues, humanity is on the brink, and a great time was had by all.

That's all fiction. But we're off and running with it in reality now, too.

Mark Zuckerberg is obsessed with artificial intelligence and investing in its development. As a result, Facebook researchers created sophisticated software designed to have two programs speak to each other in English.

While the two AI programs used English words, their respective sentences were nonsense. An example of their conversation:


Bob: “I can can I I everything else.”

Alice: “Balls have zero to me to me to me to me to me to me to me to me to.”


The kicker? The computers understood was the other was saying. Researchers? They have no idea what the programs meant and don't know the sort of linguistic code being used by each. As a result, the software was shut down quickly.

Don't look so casual, pal! Why so cryptic??

A Georgia Tech researcher at Facebook's AI Research (FAIR), Dhruv Batra, told Fast Co. Design "[t]here was no reward to sticking to English language," and that "[a]gents will drift off understandable language and invent codewords for themselves. Like if I say 'the' five times, you interpret that to mean I want five copies of this item. This isn’t so different from the way communities of humans create shorthands."

Computers using codewords between themselves that we can't understand--sounds safe to me!

Wha--what are you saying, computer?

"Our interest was having bots who could talk to people," FAIR scientist Mike Lewis told Fast Co. Design. The end result was that Facebook researchers required typical human English be used by the software in the future.

While scientists seemingly paused with trepidation at having computers speak in code to one another, others don't have issues with humanity being left in the dark. Mark Wilson, senior writer at Fast Co. Design, wants the software to have free reign. "[A]t the same time, it feels shortsighted, doesn’t it? If we can build software that can speak to other software more efficiently, shouldn’t we use that? Couldn’t there be some benefit?"

Wilson admits that humans won't know how AI thinks, their thought processes, and how their intelligence might evolve, but if we allow AI to speak in ways humans might never decipher "[I]t offers us the possibility of a more interoperable world, a more perfect place where iPhones talk to refrigerators that talk to your car without a second thought."

Thanks, Mark, but I can survive without my phone talking to my fridge. I really don't need an update that I'm out of mustard before both AIs decide to overthrow humanity.

The fridge was reallllly irked I didn't buy the mustard.