Post by anwenwilson on Aug 3, 2017 9:20:01 GMT
Relax, there is no rogue artificial intelligence system shut down by Facebook because it was trying to avoid communicating with humans.
In June, Facebook's Artificial Intelligence Research unit started an experiment attempting to train bots to negotiate.
"Similar to how people have differing goals, run into conflicts, and then negotiate to come to an agreed-upon compromise, the researchers have shown that it’s possible for dialog agents with differing goals (implemented as end-to-end-trained neural networks) to engage in start-to-finish negotiations with other bots or people while arriving at common decisions or outcomes," reads a June blog post from researchers.
In June, Facebook's Artificial Intelligence Research unit started an experiment attempting to train bots to negotiate.
"Similar to how people have differing goals, run into conflicts, and then negotiate to come to an agreed-upon compromise, the researchers have shown that it’s possible for dialog agents with differing goals (implemented as end-to-end-trained neural networks) to engage in start-to-finish negotiations with other bots or people while arriving at common decisions or outcomes," reads a June blog post from researchers.
Researchers created negotiating scenarios involving balls, hats and books, with bots holding set values for each item (for example, one bot might value balls more than books).
As Fast Company reports, during the negotiating process, the AI developed their own way of communicating. The report says researchers shut down the conversation once it became too tough to decipher.Of course, the idea of AI developing their own language conjures up thoughts of Skynet, the infamous sentient AI made popular by The Terminator movie franchise.
As Fast Company reports, during the negotiating process, the AI developed their own way of communicating. The report says researchers shut down the conversation once it became too tough to decipher.Of course, the idea of AI developing their own language conjures up thoughts of Skynet, the infamous sentient AI made popular by The Terminator movie franchise.
In a Facebook post published early Tuesday, Dhruv Batra, one of the researchers on the project, says there's no reason to freak out.
"While the idea of AI agents inventing their own language may sound alarming/unexpected to people outside the field, it is a well-established sub-field of AI, with publications dating back decades," writes Batra.
With the use of AI on the rise, concerns about the technology might evolve have cropped up, most notably from Tesla and SpaceX CEO Elon Musk, who has called for governments to consider regulations on AI.
Not everyone is focused on AI's doomsday potential. In a recent live chat, Facebook CEO Mark Zuckerberg said he is more upbeat about the technology.
With the use of AI on the rise, concerns about the technology might evolve have cropped up, most notably from Tesla and SpaceX CEO Elon Musk, who has called for governments to consider regulations on AI.
Not everyone is focused on AI's doomsday potential. In a recent live chat, Facebook CEO Mark Zuckerberg said he is more upbeat about the technology.