In recent weeks, a story about experimental Facebook machine learning research has been circulating rapidly. Two Facebook chat bots learned to talk to each other, in their own language defying codes provided, which is admittedly a pretty creepy way.
After recent concerns over the implications of artificial intelligence, Facebook reportedly shut-down an artificial intelligence system after researchers found out that it had started talking in a language they could not comprehend because “things got out of hand”.
Initially the AI agents used English to converse with each other but later they started conversing in a language created by them, from the scratch and without human input.
Bob: “I can can I I everything else.”
Alice: “Balls have zero to me to me to me to me to me to me to me to me to.”
The above is a reported conversation between the bots, unintelligible to humans, but designed to make communication faster between them.
There are also speculations that the alien language was a result of Facebook developing negotiation software as was asked by FastCo Design, a few weeks ago to develop a “generative adversarial network” for the purpose of developing negotiation software. The two bots quoted in the above passage were designed, as explained in a Facebook AI Research unit blog post in June, for the purpose of showing it is “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.”
The bit is scary since experts like Stephen Hawking have been warning against the same, saying that humans, used to slower evolution, will be fast outpaced by these intelligence bots.
Few days before, Tesla CEO Elon Musk too said that AI was the biggest risk and added Facebook co-founder Mark Zuckerberg’s understanding of AI was limited.