Facebook shut down one of its AI systems due to “things moved out of control.” The reports confirm that the AI bots created its own language, without human input, prompting the firm to shut down. Alarmingly, the bots’ initiative to create and communicate in a new language challenged the given codes. Considering the various concerns regarding the advancement of AI system, the news created worries in the tech world.
The report continued that the AI system stopped using English language and began using a new one it created. The AI agents use English to communicate each other, but recently, they started using a language only AI systems understand. Finally, the Facebook research team shut down the AI systems and instructed them to communicate each other in English.
Shockingly, in June, the researchers of Facebook’s AI Research Lab noticed that the dialogue agents were making their own language. Suddenly, the bots deviated from the scripted norms to communicate in a new language they created without any human assistance. The dialogue agents were allowed to express freely to improve their conversation skills with the given machine learning algorithms. The researchers also noted that the bots became highly crafty negotiators.
After becoming fine negotiators, the bots used advanced strategies and machine learning to improve the outcome of the negotiations. Later, the bots started showing feigning interest in one particular item for “sacrificing” it at an advanced stage of negotiation as a false compromise. Though some people see it as an advancement of AI, many experts warn that AI would supersede humans one day. People like Professor Stephen Hawking, Apple co-founder Steve Wozniak, tech billionaire Bill Gates, and Tesla’s Elon Musk have already talked their concerns about AI and where the technology is heading.
Spat between Musk and Zuckerberg
Surprisingly, the latest development reported just days after the verbal spat between Zuckerberg and Musk on AI and its future. Musk tweeted that he spoke with Mark about AI and found the Facebook CEO’s knowledge on technology is limited. Interestingly, the tweet came after Zuckerberg criticized Musk for advocating regulations for AI before it becomes mainstream. The incident took place during a Facebook live stream in the earlier weeks of July.
In his response, Mark stated that the naysayers were trying to make stories about various doomsday scenarios. Mark continued that he did not understand those logics and felt it completely irresponsible. Interestingly, Musk has spoken multiple times in the past regarding the risks associated with AI advancement and termed it as the “biggest risk to human civilization.”