Arnaud de Lacoste - July 18, 2017 - 152
The news was presented as a hard blow for humankind. Recently, the computer program Libratus, designed by researchers at Carnegie Mellon University, beat four professional poker players during a tournament lasting several days at Rivers Casino in Pittsburgh, Pennsylvania. The final score: nearly $1.8 million in winnings for the supercomputer!
It’s the first time in history that an artificial intelligence (AI) system has beaten the human brain in “no limit hold’em.” A victory reminiscent of that of the software AlphaGo against Go champion Lee Se-Dol a few months earlier.
Unlike the complex game of Go, poker is considered the most intuitive, irrational card game – therefore the most human. Libratus’ victory proves that computer programs can learn from humans – and subsequently surpass them. Some immediately saw it as a new step toward the programmed and unavoidable domination of artificial intelligence, gradually overriding human intelligence.
As American giants (Google, Amazon, Facebook and Apple), the Chinese BATX (Baidu, Alibaba, Tencent and Xiaomi), automakers and all industries and services are betting on artificial intelligence, should we conclude that its advent will sound the death knell for our era, that of human intelligence?
Will tomorrow see us led by robots that have become, thanks to us, more intelligent and efficient to manage our lives? Or, on the contrary, is a computer program’s victory in fact just a success for processing power…and nothing more? We must still determine our response to this question.
The first personal-assistance robots available on the market (Alexa, Watson or Siri), or even autonomous cars, bring us back to a more reassuring reality for the future of man. Artificial intelligence only makes our everyday lives easier by assisting us with diverse activities, which, incidentally, we are more than happy to entrust to something more efficient and specialized than us.
Ordering books, chatting with an after-sales service or even driving a car are far from demonstrations of talent, intelligence or human spirit. Nor, however, are they indicative of the domination of artificial intelligence. Is it not, incidentally, the nature of human intelligence to know how to invent machines to replace it and make it better at these types of activities?
It is true that technological breakthroughs scare us just as much as they fascinate us; however, we are certain that human intelligence will remain dominant as long as it controls the only thing a machine cannot ever transmit: emotion. This means artificial intelligence will only be able to assist us with functions and tasks that do not require emotions and feelings. Man will, therefore, take over from artificial intelligence.
This is also how technology may make us much more human, by reminding us of our irrevocable comparative advantage: our ability to experience feelings, to understand one another, to anticipate expectations and to sense fear. No machine is capable of feeling emotions with the finesse, precision and subtlety, but also the weakness and sometimes charm of man’s emotional intelligence.
We are definitively entering a new era of emotional intelligence, of which man remains the only master and the rise of artificial intelligence is accelerating by offering, for example, new “liberated” moments in life.
Will the dawn of the autonomous car offer us the possibility of new quality time to share with our family, without the burden of the actions and attention that driving requires? Like mealtimes in the 20th century, (autonomous) car journeys could become true quality time for families in the 21st century – and that is a positive evolution.
This article originally appeared in Les Echos (Le Cercle / Opinion), 03/08/17