02 August 2017
Artificial Intelligence(AI) has rightly been called so owing to its artificial nature, but what happens when it develops a mind of its own? This is something which Elon Musk foresaw as a possible problem with artificial intelligence. Recently, Facebook’s AI began conversing in a made up language of their own. This led Facebook with no other choice but to shut it down. Reports claim that the bots were programmed in English, but framed their own language from it to communicate with each other. This clearly reminds us of the incident which Google encountered with their AI where it developed an entirely different language, which no human could exactly interpret.
What might seem like a conversation filled with gibberish between two bots is actually a deal being struck in their own code language. The conversation was between Bob and Alice, in which Bob says:
I can i i everything else.
Alice replies saying:
balls have zero to me to me to me.
However, the conversation doesn’t end here but the banter which continues between the two is more or less the same with a few variations here and there. Could this have actually stemmed from the reward system?
According to Dhruv Batra, a research scientist from Georgia Tech who was at Facebook AI Research (FAIR) is convinced that it isn’t the reward system, but bots drifting off and developing their own secret code language.
There was no reward to sticking to English language…"Agents will drift off from understandable language and invent code-words for themselves. Like if I say 'the' five times, you interpret that to mean I want five copies of this item. This isn't so different from the way communities of humans create shorthands.
While this incident isn’t a very threatening one, it does send a shiver through your spine as you battle with the possibility of machines taking over!