Today, many of the world’s leading companies are in a one-of-a-kind race: To bring () to life. Already, systems are the core of many businesses, so it’s no surprise that updates about this or that neural net often pop up on our newsfeed. Such headlines typically read along the lines of, “ beats human players in video game” or “ mimics human ” and even sometimes things like “ detects cancer using .”
But just how close are we to having machines with the intelligence of a human—machines that we can talk with and work with like we do any other individual? Machines that are conscious?
While all of the aforementioned developments are real, Yann LeCun, Director of Research at Facebook and a professor of computer science at NYU, thinks that we may be overestimating the abilities of today’s , and, thus building up a bit of hype. “We’re very far from having machines that can learn the most basic things about the world in the way humans and animals can do,” LeCun told The Verge in an interview published last week . “Like, yes, in particular areas machines have superhuman performance, but in terms of general intelligence we’re not even close to a rat.”
This so-called artificial general intelligence (AGI) refers to an operator capable of performing virtually every task a human being could. Conversely, today’s specialize in particular tasks: for example, image or recognition, or identifying patterns by sifting through tons of data that AIs have been trained on. These specialized AIs are also called “applied ” or “narrow ” to highlight their rather limited intelligence.
Speaking to Futurism via email, Manuel Cebrian, one of the MIT researchers that developed Shelley, an horror storyteller, agreed with LeCun’s sentiments. “ is just a great tool,” he said, adding that, “it seems to me, based on my work with Shelley, that is very far from being able to create professional-level horror fiction.” And thus, still quite far from human levels of intelligence. […]