Sorry to burst your bubble, but the story circulating the Internet about Facebook AI going off the rails is a load of sensationalist garbage. Sites and news sources that are spreading this should all be ashamed of themselves.
Here’s the Real Deal
1. The bots came up with a shorthand method of communication, based on English. Human languages are not always the most efficient for conversations in a mathematical system. Chatbots and AI for the most part are rule-based glorified calculators, as are all computer programs. They’re not about to take over the world, and are a far cry from doing so.
2. The language was human readable, but diverging from English was not part of what the researchers wanted to accomplish. The researchers were looking to have the chatbots to communicate in plain English.
3. They have shut down the chatbots and restarted them before because of #2. It’s a known problem, and this was an expected possible outcome.
4. It’s not new, or alarming. It shouldn’t scare anyone, like all of the headlines suggest.
Listen, we understand that news sources want to make some money off of advertising. Apparently the media either has absolutely no understanding of what happened, or they can’t report on something without completely misleading people with an anxiety inducing tagline.
Short Story: Robots and AI are not rising up to harm you in the near future. When our robot overlords do come, they most certainly will not be coming in the form of a Facebook AI chatbot.ƒ