Facebook's AIs made their own language, but researchers killed it
- Author: Joey Payne Aug 02, 2017,
Aug 02, 2017, 0:22
In a weird futuristic twist, Facebook has officially shut down two artificial intelligence programs that appeared to be purposely communicating with each other in odd language that could not be understood by their human controllers. Researchers have found that the machines, which are nicknamed Bob and Alice, were communicating with each other in a language that sounded gibberish to humans.
Facebook shut down AI bots that were talking to each other in a language unintelligible to humans.
Facebook researchers in NY set up Bob and Alice, in a bid to develop automated trouble-shooters for social media networks. "Facebook shuts down robots after they invent their own language", London's Telegraph newspaper reported.
The pair of bots were supposed to communicate exclusively in English, but instead created a new language for themselves.
The researchers tasked the robots with trading hats, balls and books by determining the value of each object and bartering for them with each other.
Bob: i can i i everything else ... At that point, the Facebook engineers pulled the plug on the A.I. after observing the interaction between Bob and Alice.
Amid the battle of words between billionaires Mark Zuckerberg and Elon Musk over the future of artificial intelligence last week, a story that suggested that Facebook had to shut down an AI program went viral. Alice: balls have zero to me to me to me to me to me to me to me to me toBob: you i i i i i everything else. In their research paper, published June 16 on the open access site arXiv.org, the team said their experiments showed for the first time that it's possible to train bots in negotiation models so they learn both linguistic and reasoning skills without prepared scripts.
Although the bots started out speaking English, the researchers realized they failed to provide a reward for speaking English.
"You i everything else", Bob told Alice after the first exchange.
Artificial Intelligence is not sentient-at least not yet.
The bots were created as part of a program to teach computers and machines how to negotiate, while the ultimate goal was to develop personalized "digital assistants" who would be able to communicate with people.
"Agents will drift off from understandable language and invent code-words for themselves", Fast Co.
The algorithms were ultimately created by the Facebook Artificial Intelligence Research lab (FAIR) as a way to improve the conversations that the chatbots were having with their human counterparts. This was actually news earlier this year, but it's making the rounds anew because Elon Musk gave us a fresh warning that AIs were the biggest threat to humanity today.