In 2010, Microsoft's Kinect is a motion-sensing camera that requires no control pad. You are, as the marketing goes, the controller. In the beginning, though, things weren't so fancy.
According to a big write-up on the history of Kinect on Wired, in 2007, Xbox boss Don Mattrick had called for a "reimagining of the way we interact" with consoles. Something beyond the decades-old video game standard of using a device in your hand to control something on a TV screen. Combining various different areas of Microsoft research, from speech recognition to facial recognition, that call to arms would eventually become Kinect.
What's interesting, though, is that while some parts of Kinect had to be drawn from parts of Microsoft completely unrelated to the Xbox - the facial recognition came from the company's China offices, for example - the Xbox team's own R&D outfit in Redmond, Washington had already been working on "exotic gyroscopic and accelerometer-based controller prototypes".
In other words, their research wasn't the result of Mattricks decision. They'd been designing them well before that. Which begs the question; had Kinect not come together as the result of all these various pieces of Microsoft technology, would Microsoft had gone down the same road as Sony with its Move peripheral, and make their motion controller a physical device? Who knows, maybe the Kinect decision killed off just such a device...
Kinect for Xbox 360: The inside story of Microsoft's secret 'Project Natal' [Wired]