In recent weeks, some alarmist statements were relayed by the media about the potential dangers of artificial intelligence. Statements has taken seriously since they emanate not unbridled imagination SF novelists or a few charlatans in evil in the media. Elon Musk and Stephen Hawking evoke nothing less than the end of the human species. Others, like Eric Horvitz believe that its impact will be positive and a source of progress for humanity. Stanford researchers will try to arbitrate between these two visions.
Military robotics is that the emerged part of the problem
Never a scenario to the Terminator, with an artificial intelligence, the Skynet, which turns against his creator to destroy it with an army of robots, has never been also frequently mentioned in the press. That’s also brilliant minds that Stephen Hawking or Elon Musk (per elsewhere investor in Vicarious, one of the startups in advanced in the development of artificial intelligence!) evoke in turn such a scenario of Revelation gives actually think. It is true, military artificial intelligence is a real ethical problem. Drones strike daily in the Middle East, in Afghanistan. Of the terrestrial armed robots begin to arrive in arsenals, but for the time being it is the man who still press the button at the time of shoot. Discussions at the United Nations focuses on a possible ban of autonomous military robots in their decisions.
If the risk that such robots are someday released one day or another on battlefields is real, the risks posed by the rise of artificial intelligence and software in our daily lives more broadly raises more question. The study by the AAAI (Association for the Advancement of Artificial Intelligence), directed by Eric Horvitz, Co-Director of Microsoft Research sought a evolve the dangers of artificial intelligence in the long term. In addition to these military applications, their study points a number of potentially dangerous applications: domestic robots, autonomous cars, are the most obvious, but an intelligence artificial ‘malicious’ could well empty your Bank via Tools e-commerce account, or even intelligent agents who will be posing as humans for purposes of extortion. Finally, researchers suggest the risk of addiction people with virtual being. A phenomenon already portrayed by Spike Jonze in his film “Her” in 2013 and as it starts to be the case in Japan where the idea of a marriage between a human and a robot is already beginning to be mentioned.
No simple solution to the rise inevitable artificial intelligences
How to accompany the rise of artificial intelligence in our societies? The answer is far from simple. Implementation of rules inspired by the 3 rules of Asimov in the robots and software agents will only solve part of the problem because criminals and dictators to keep well to implement this type of electronic ‘scruples’ in their achievements. This first study was conducted from 2008 to 2009 and strength and to see that nothing has yet really changed on this plan both on the side of researchers, industrialists or regulation. Eric Horvitz has just relacer an initiative, the ‘ One Hundred Year Study on Artificial Intelligence . (AI100) “with Stanford University. An extensive program of study that will seek to assess how artificial intelligence will evolve in the long term, what will be its impact on democracy and our freedom, but also on the economy, the military balance, the crime.
The reflection is launched, it will probably never end.
Translation : Bing Translator
“One-Hundred Year Study of Artificial Intelligence: Reflections and Framing”, Eric Horvitz, 2014
“Study to Examine Effects of Artificial Intelligence”, The New York Times, December 15, 2014
“Stephen Hawking warns artificial intelligence could end mankind”, BBC, December 2, 2014
“5 Very Smart People Who Think Artificial Intelligence Could Bring the Apocalypse”, Time, December 2, 2014
“Why Elon Musk is scared of artificial intelligence — and Terminators”, Washington Post, Novembre 18, 2014