Artificial intelligence has made gigantic leaps in the past few years. Thanks to the advent of technologies like multicore processors, machine learning and neural networking, A.I. is able to mimic known human brain functions more closely than ever before. Facebook’s A.I. can recognize a human face and decide who it belongs to. Google’s A.I. can beat a world champion at an incredibly complex board game that’s been around for centuries. Amazon’s A.I. can figure out what products you may want for your home and help you manage your IoT devices. None of them, though, are anywhere close to being able to match the brainpower, learning skills and deductive capabilities of a human.
The White House Office Of Science And Technology Policy, looking for how to regulate A.I. and if such a move would even be necessary, held an event on Tuesday where experts in the tech and legal fields gathered to tackle the tough questions about A.I., such as the role of A.I. in fields like health care and transportation, how adept they may be, and what may need to be done on the lawmaking front to ensure that A.I. doesn’t become a dangerous or otherwise counterproductive technology. For the moment, however, the panel of experts there determined that A.I. is nowhere near human level in learning capability or flexibility, and there are a number of significant breakthroughs that will need to be made before they can hope to reach that level.
Specialists met in a similar manner and discussed similar questions back in 2009, driven more by public opinion than any legitimate questions that the A.I. technology of the time may have posed. While it’s entirely possible that the A.I. of the future could create a nearly human-free economy or otherwise revolutionize the world, feats like this are distinctly out of the pay grade of current A.I. tech. A Microsoft researcher, Kate Crawford, suggested adding in ethics to the basic training provided to future A.I. engineers to help ensure that Hollywood-esque A.I.-based horror stories never come to pass, no matter how advanced the technology gets. Tuesday’s meetup was only the first of four events that the White House has planned for A.I. researchers.