Facebook to reveal advances in AI research

The world’s largest social media site reported 60 million new monthly active users in the third quarter, bringing its global users to 1.55 billion.

The social network’s new tech would enable blind people to interact with shared photos in a way that isn’t now possible.

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 especially helpful if you’re blind or can’t see the photo”, he wrote.

The company is now planning to pair this application with technology that can identify individual objects from photos and label them, such as a bat being held by a baseball player. Facebook drew on its image-processing software to create a system that learned to predict whether a stack of virtual blocks will tumble.

VR is the obvious first step towards this device development, and will of course include an Oculus headset, along with the Oculus Touch controllers, but the technology that Facebook is hoping to develop will include a lot more as well.

“Go” has more possible permutations than chess despite its simple rules, meaning it has so far been too hard to design a computer system that can beat humans.

Mr Schroepfer said that the firm “has a full-scale version fully constructed which will be undergoing flight tests very soon”. It has one of the world’s top deep learning experts, Yann LeCun, running its AI division; the eight-person team from machine-learning startup Wit.ai, which Facebook acquired in January, is running M. The company won’t say how many operators it’s using for M, but BuzzFeed found that Facebook is using outside services like TaskRabbit to complete a few of the requests.

The best human Go players succeed due to their ability to see visual patterns on the board.

“So much of the world is visual”. Science does not have better objective to serve and Facebook stands out in this. According to Facebook chief technical officer Mike Schroepfer, the company has plans to “effectively build a teleporter” by 2025. 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. The system now automatically offers to ask those two questions for a human agent at the click of a button, helping them respond more quickly.

“The reason this is exciting is it’s scalable”, he added.

Facebook’s M system is already connected to its MemNets project, he said, which helps it learn as users as questions and humans find, and plug in, the answers.

The AI is still in the works, and, as of right now, it does not get everything right.