2. Gestures

Current generation touch-screen devices already have rudimentary gestures. In fact, even the Apple Newton, one of the first personal digital assistants, supported gestures. If you circled text while writing on the Newton, the circled word would then be "selected." That's a gesture. Interestingly, multi-touch amplifies the power of gestures by an order of magnitude. For example, you can put two fingers on the left and right side of a photograph, then use the gesture of moving your fingers apart to instantly enlarge the picture.

3. Physics

Second-generation UIs have folders, trash cans and documents that represent physical objects. But they don't act like physical objects. They don't move like they have weight, mass and momentum. When you slide a folder across your Windows desktop, it doesn't slow down gradually, but stops the instant you release the mouse button. When you crash an icon against other desktop objects, they don't scatter like bowling pins. If they did, your mind would more readily accept them as real objects. Here's an example of gestures combined with physics.

4. 3D