Recently, the social media giant Facebook found in its tests that its own AI bots were started talking in a strange and apparently erroneous language. Hence, before losing the control completely the social network giant Facebook has decided to shut down the system.
Facebook division of artificial intelligence development (Facebook Artificial Intelligence Researchers) created a system devoted to see how they could interact in colloquies.
The results of the tests of the system began showing that the “bots” were talking in a strange and apparently erroneous language. However, this was not an error, since the system had created its own language. Faced with this type of interaction on its own Facebook has decided to deactivate it, as the technology portal “The Next Web ” points out.
The two “bots”, called Bob and Alice, used random and nonsense words, so the researchers thought at first that the AI system was suffering a failure. However, after an analysis they noticed that in these words there was a pattern, so the “bots” had created their own language because they had considered that in this way it was much more direct than the preferred language.
The problem for an AI system is that if it decides to ignore the language through which it has been programmed, it implies a difficulty for the development of neural networks and the subsequent technology that their manufacture can assume.
Given the difficulty of communication between the system and the human being, the social media giant Facebook has decided to shut down the system before losing control completely. However, the company of Mark Zuckerberg is not the first one that has to deal with a problem of this caliber, since the AI system of the tech giant Google also suffered setbacks in the language.
Mark Zuckerberg vs Elon Musk
The controversy over the future threat posed by artificial intelligence led a few days ago to face Facebook founder Mark Zuckerberg and Elon Musk, the man behind Tesla and Space X. While Musk warns that if he is not careful in the development of artificial intelligence may in the future pose a danger to the human being, Zuckerberg rebuked him that there was nothing to fear and that the fears were a delay to progress.
“I have a pretty sharp opinion on this,” Zuckerberg said during a live broadcast. “I am optimistic. I think you can build things to make a better world. But, in addition, I am especially optimistic as far as artificial intelligence is concerned.” later he added, “and I believe that there are those who are pessimistic and try to give hype to all these apocalyptic scenarios. I just do not get it”.
Given these statements, Elon Musk decided to intervene and address the visible face of Facebook, what then tweeted “I talked to Mark about this. Their understanding on the subject is limited.”