Machines contain the breadth of human knowledge, yet they have the common sense of a newborn. The problem is that computers don’t act enough like toddlers. Yann LeCun, director of artificial intelligence research at Facebook, demonstrates this by standing a pen on the table and then holding his phone in front of it. He performs a sleight of hand, and when he picks the phone up?ta-da! The pen is gone. It?s a trick that?ll elicit a gasp from any one-year-old child, but today’s cutting-edge artificial intelligence software?and most months-old babies?can?t appreciate that the disappearing act isn?t normal. ?Before they?re a few months old, you play this trick on them, and they don?t care,? says LeCun, a 54-year-old father of three. ?After a few months, they figure out this is not normal.?
One reason to love computers is that, unlike many kids, they do as they?re told. Just about everything a computer is capable of was put there by a person, and they’ve rarely been able to discover new techniques or learn on their own. Instead, computers rely on scenarios created by software programmers: If this happens, then do that. Unless it’s explicitly told that pens aren’t supposed to disappear into thin air, a computer just goes with it. The big piece missing in the crusade for the thinking machine is to give computers a memory that works like the grey gunk in our own heads. An AI with something resembling brain memory would be able to discern the highlights of what it sees, and use the information to shape its understanding of things over time. To do that, the world’s top researchers are rethinking how machines store information, and they’re turning to neuroscience for inspiration.
This change in thinking has spurred an AI arms race among technology companies such as Facebook, Google, and China?s Baidu. They?re spending billions of dollars to create machines that may one day possess common sense and to help create software that responds more naturally to users? requests and requires less hand-holding. A facsimile of biological memory, the theory goes, should let AI not only spot patterns in the world, but reason about them with the logic we associate with young children. They?re doing this by pairing brain-aping bits of software, known as neural networks, with the ability to store longer sequences of information, inspired by the long-term memory component of our brain called the hippocampus. This combination allows for an implicit understanding of the world to get ?fried in? to the patterns computers detect from moment to moment, says Jason Weston, an AI researcher at Facebook. On June 9, Facebook plans to publish a research paper detailing a system that can chew through several million pieces of data, remember the key points, and answer complicated questions about them. A system like this might let a person one day ask Facebook to find photos of themselves wearing pink at a friend’s birthday party, or ask broader, fuzzier questions, like whether they seemed happier than usual last year, or appeared to spend more time with friends.
Read more at?BLOOMBERG