If machines are to be more useful, they’ll need the same kind of common-sense understanding, says Mike Schroepfer, Facebook’s chief technology officer. While people can see stripes and shapes, the computer only recognizes pixels, making it more hard to distinguish what is in the images.
Advertisement
The Faceboffins have also scaled up a technology that helps neural networks develop “a short-term memory” and answer questions as a human might. This is something you and I do naturally – for example, none of us had to go to a university to learn that a pen will fall to the ground if you push it off your desk – and it’s how humans do most of their learning. The user can ask questions like ‘Is there a man in the photo?, Is he smiling?’ and the system will answer them.
Earlier this year, we showed a few of our work on natural language understanding – specifically, a system called Memory Networks (MemNets) that can read and then answer questions about short texts.
The advancements were outlined in a blog post by Facebook CTO Mike Schroepfer.
Schroepfer said the Facebook AI Research team will present a new paper at AI conference NIPS 2015 next month detailing its state-of-the-art object-detection system, which segments images 30 percent quicker than previous industry benchmarks, using 10 times less training data.
Facebook isn’t the only tech company working on these problems. It has developed a system created to automatically interpret visual contents for those users by spelling out the basics of a photo.
Striving to keep up with the increasing demands of delivering users’ News Feeds, Facebook is pressing hard to advance artificial intelligence.
Or, users could ask the systems questions about a photo, such as if there’s a baby in a photo, where the baby is standing, and what it’s doing.
Advertisement
According to the report, Facebook plans on the Oculus being able to trick your senses as well into thinking the VR world you’re seeing is real, one the replicates and mimic around their real world, as well as letting anyone create anything to they want in 3D. One of the things we’ve built to help do this is an AI player for the board game Go. That’s mainly, Schroepfer said, because Go involves an extremely complex branching system, where even a small number of moves have more than a hundred thousand possible outcomes. Rather, Schroepfer suggested, the true promise of Facebook’s AI efforts may well be in providing every one of its users with an always-on personal digital assistant. However, to handle complicated requests and tasks, the system needs more advances in machine language, vision, prediction and planning. “But with the right AI system, we could deploy that for the entire planet”. This is a really big challenge, and we’re just getting started.
