Beyond Kinect: Microsoft's vision for next-gen interfaces

Face recognition a key component

Devices in the home need to be good at voice recognition, too. Blake says that recognition technology is in quite good shape.

"Your machine learning, like in Kinect, has to adapt across the population, it's already got to encompass the range of people. The second thing is that, as already takes place with speech recognition, that there's some further fine tuning as you use the system."

Blake adds that although they don't deal with end products, his researchers do think about real-world applications because it keeps them grounded: "One of the things we're excited about is in the medical area. Let's say you're in the operating theatre and you've got some complex 3D data you want to look at as part of the operation.

"At the moment it's difficult for the surgeon to disengage and move over to a computer… for one thing if you started tapping on the keyboard you'd have to scrub up again. So it would be great to have some kind of a gesture and voice driven system to get this information up.

"We're already talking to medics… it's not that we're about to produce a commercial system but that we need to understand in some depth how they would value the technology."

The display is the limit

Blake adds that display technology is a limitation, but it is something that's advancing. "We have a group here in the lab that particularly keeps its eye on that, such as Second Light. I think it's incredibly exciting and an interesting way to [explore] 3D data.

• SecondLight explained: MS Surface on steroids

"You can imagine taking the secondary display device and look through [the data] – you could take anatomical slices. It's very hard to explore 3D data – our eyes are not 3D, they see surfaces rather than volumes. Thinking of ways to interact with volumes is quite challenging."

Blake is understandably wary of pinpointing definitive shifts trends in future technology. "It's always hard to know how users will react to the technology. [Over the] next 5-10 years, I expect we'll get some unexpected failures and some unexpected successes… some things might turn out to be a technologist's dream but it doesn't hack it with the general public."

One of our scientists here is an anthropologist by training. He's very keen on the idea of starting from the user end if you like; running trials in people's home's of experimental technology and seeing how people interact with it. I think that's a great idea, as it allows you to fail early.

Blake says there are so many concepts that have failed in the eyes of users that technologists would have chosen every time. "There are all kinds of stories, texting is a classic example of a technology… it was just put into mobile phones and yet it has become a pervasive technology." No wonder he'd rather not make too many predictions.