Here’s another instance of bots defying human commands which might remind you of Terminator 2: Judgement Day, a blockbuster flick where a machine learning system called Skynet, seeks to supersede the human race and take over the world. As Facebook has now reportedly shut down an AI system after bots were spotted creating their own language, eventually disobeying human instructions. These bots were found to deviate from the standard norms and eventually started communicating in an entirely different language.
According to the report, Facebook had to terminate the Artificial Intelligence system that it was researching on because things got out control. As the system was instructed to communicate in English, instead it started using an entirely new language which lead Facebook to pull the plug on the system.
The report further notes that Facebook AI Research Lab (FAIR) also found that dialogue agents were busy creating their own language while researchers were trying to improve chatbots. Researchers suggest that these bots are apparently “incredibly crafty negotiators”.
While on the face of it, this appears to be a huge step forward for Artificial Intelligence Systems. That said, leading experts and scientists have expressed concerns over the rise of AI. Professor Stephen Hawking has raised fears that AI could supersede the human race, and eventually lead to their extinction. Even Elon Musk, CEO of Tesla dubs AI as our biggest existential threat. Even the likes of Bill Gates and ex-Apple co-founder Steve Wozniak have raised fears over the advancements in machine learning technology and machine learning systems.
Notably, the news comes just days after a verbal showdown between Tesla’s founder Elon Musk and Facebook CEO Zuckerberg, who exchanged some intense words over the future of AI in a debate. Musk has been constantly expressing his concerns over the advancements in AI, and calls it as the biggest risk we face as a civilization”.