Thursday, February 05, 2015

Will a baby be cute to Artificial Intelligence?

Is Artificial Intelligence, or AI, a boon for the world or will it result in super intelligent robots who enslave or wipe out human beings? Depends on who you ask. On one side are Elon Musk, Bill Gates and the eminent scientist, Stephen Hawking who think that it " could spell the end of human race " and on the other side Prof Max Tegmark of MIT, Paul Allen of Microsoft and Jack Ma of Alibaba who are investing hundreds of millions of dollars in research on AI. There have been many movies in which computers or robots have been villains, either trying to wipe out or enslave human beings but usually one human proves too smart for the machines. " There had been a ridiculous amount of scaremongering," Said Prof Tegmark. " And understandably a lot of AI researchers feel threatened by this." But is it as ridiculous as the good professor thinks? A human being is not measured only by Intelligence Quotient but by many other parameters. Indeed, some believe that Emotional Quotient, which measures how we empathize with others, leads to greater success. The human brain is unique among all animals because we are the only ones with a speech center which gives us the ability to pass knowledge to the next generation so that succeeding generations have built on the past to come to the stage of building AI. Among humans Einstein's brain was unique, not because it had more cells, but because it had vastly more connections between the two halves. But what really distinguishes human from AI is that although we are intelligent we are also irrational and behave illogically. Partly it maybe because we have instincts like all animals. The fear of the dark is probably a leftover from the time when our ancestors were vulnerable to predators with night vision, such as leopards. We tend to laugh if we see a man slip and fall but we are consumed with anxiety if our child falls down and rush to find out if she is hurt. But even in activities, such as economic decisions, where you would expect cold logic, we make irrational decisions. Former Federal Reserve Chairman, Alan Greenspan complained of " irrational exuberance ", when talking about share prices, in a speech in 1996. This basic irrationality is now being recognised by economists who are beginning to understand why macro-economic predictions go spectacularly wrong. Like all computer programs AI is based on algorithms which means it has to be logical. Logic cannot appreciate the beauty of a sunset, the cuteness of a baby or the decision to donate a kidney to a partner. This restriction within the bounds of pure logic is what makes AI so dangerous. Since humans kill other humans it is logical to control all humans. Dangerous. 

No comments: