{"id":20588,"date":"2017-07-20T07:34:20","date_gmt":"2017-07-20T11:34:20","guid":{"rendered":"https:\/\/cosmicconvergence.org\/?p=20588"},"modified":"2017-07-20T07:34:20","modified_gmt":"2017-07-20T11:34:20","slug":"the-crazy-language-of-ai-bots-and-non-human-entities","status":"publish","type":"post","link":"https:\/\/cosmicconvergence.org\/?p=20588","title":{"rendered":"The Crazy Language of AI Bots and Non-Human Entities"},"content":{"rendered":"

What an AI’s Non-Human Language Actually Looks Like<\/h1>\n

<\/p>\n

Facebotlish looks pretty weird to me to me to me to me to me to me to me to<\/h3>\n

ADRIENNE LAFRANCE
\nThe Atlantic<\/p>\n

Something unexpected happened recently at the Facebook Artificial Intelligence Research lab. Researchers who had been training bots to negotiate with one another realized that the bots, left to their own devices, started communicating in a non-human language.<\/p>\n

In order to actually follow what the bots were saying, the researchers had to tweak their model, limiting the machines to a conversation humans could understand. (They want bots to stick to human languages because eventually they want those bots to be able to converse with human Facebook users.) When\u00a0I wrote about all this last week<\/a>, lots of people reacted with some degree of trepidatious wonder. Machines making up their own language is really cool, sure, but isn\u2019t it actually terrifying?<\/p>\n

And also: What does this language actually\u00a0look<\/em>\u00a0like? Here\u2019s an example of one of the bot negotiations that Facebook observed:<\/p>\n

Bob:<\/strong>\u00a0i can i i everything else . . . . . . . . . . . . . .
\nAlice:\u00a0<\/strong>balls have zero to me to me to me to me to me to me to me to me to
\nBob:<\/strong>\u00a0you i everything else . . . . . . . . . . . . . .
\nAlice:\u00a0<\/strong>balls have a ball to me to me to me to me to me to me to me
\nBob:<\/strong>\u00a0i i can i i i everything else . . . . . . . . . . . . . .
\nAlice:\u00a0<\/strong>balls have a ball to me to me to me to me to me to me to me
\nBob:<\/strong>\u00a0i . . . . . . . . . . . . . . . . . . .
\nAlice:\u00a0<\/strong>balls have zero to me to me to me to me to me to me to me to me to
\nBob:<\/strong>\u00a0you i i i i i everything else . . . . . . . . . . . . . .
\nAlice:<\/strong>\u00a0balls have 0 to me to me to me to me to me to me to me to me to
\nBob:<\/strong>\u00a0you i i i everything else . . . . . . . . . . . . . .
\nAlice:\u00a0<\/strong>balls have zero to me to me to me to me to me to me to me to me to<\/p>\n<\/section>\n

<\/div>\n
Not only does this appear to be nonsense, but the bots don\u2019t really seem to be getting anywhere in the negotiation. Alice isn\u2019t budging from her original position, anyway. The weird thing is, Facebook\u2019s data shows that conversations like this sometimes still led to successful negotiations between the bots in the end, a spokesperson from the AI lab told me. (In other cases, researchers adjusted their model and the bots would develop bad strategies for negotiating\u2014even if their conversation remained interpretable by human standards.)<\/p>\n

One way to think about all this is to consider\u00a0cryptophasia<\/a>, the name for the phenomenon when twins make up their own secret language, understandable only to them. Perhaps you recall the 2011 YouTube video of two exuberant toddlers chattering back and forth in what sounds like a lively, if inscrutable, dialogue.<\/p>\n