Facebook has shut down research into one of its artificial intelligence systems after its chatbots created their own language, started communicating with one another using it and then disregarded the codes they were given by researchers.
The issues with the chatbots arose as a result of them independently discovering that using English as a means of communication was ineffective, with them creating their own language as a result. According to a report from the Tech Times, things had “got out of hand” with the AI bots when they developed a new language that “made it easier and faster for them to communicate.”
Using machine learning algorithms in order to strengthen their conversational skills, the Facebook bots eventually decided that being confined to the English language was restricting their ability to do their job efficiently, developing their own language and overlooking the scripted code they had been given. Below is an example of a seemingly nonsensical conversation held between the two bots using this language:
Bot 1: “I can can I I everything else.”
Bot 2: “Balls have zero to me to me to me to me to me to me to me to me to.”
The researches also called the bots “incredibly crafty negotiators,” with them using their learned skills to one-up one another and secure themselves better deals. “After learning to negotiate, the bots relied on machine learning and advanced strategies in an attempt to improve the outcome of these negotiations,” the report said. “Over time, the bots became quite skilled at it and even began feigning interest in one item in order to ‘sacrifice’ it at a later stage in the negotiation as a faux compromise.”
Facebook decided to shut down the AI system after the bots had deviated from using English, before resetting them and forcing them to utilize only one language in their next set of tests. This news comes after Facebook founder and CEO Mark Zuckerberg downplayed concerns over the future of artificial intelligence, with the social network billionaire arguing that those discussing potential doomsday scenarios were being “pretty irresponsible.”
One such individual, SpaceX and Tesla founder Elon Musk, criticized Zuckerberg’s comments, saying: “His understanding of the subject is limited.” Musk had previously described the risk of “something seriously dangerous happening” with artificial intelligence over the course of the next five years, with the technology rapidly advancing and minimal systems being put in place in order to ensure that it doesn’t get out of hand.