[newhouseb]

> In order to patent something, you have to reduce it to practice. That is to say, you have to actually make it work, in the real world.

This is false. I could patent the idea of using, say, a camera's sensor to recognize the species of a plant based on a leaf without any idea of how I would actually build such a thing (what algorithms would be required, etc). I could similarly patent the idea of tracking a user's eye movement to control a user interface, without the slightest clue of how to track a user's eye movement in realtime.

To take one small example - if you were to pull up patents related to virtual reality, the claims (which tend to be written in the mid-1980s) are all ridiculous, the dreamy technology simply didn't exist to build the proposed ideas (i.e. computers small enough to stick on people's heads), but they were still patented anyway.