The Kindle doesn't have a spine, and it turns pages by the press of a paddle rather than flip of a paper, so it is even easier to scan by machine than a hard book is.
There are machines that accept a stack of (unbound) paper and OCR it all. There's no "batch mode" for Kindle (apart from cracking the DRM, but that's a separate issue).
And a duck isn't an orange. His point was that an OCR machine would be easier to make for the kindle--just have it press the paddles. His point was not that an existing OCR machine for books would work on a kindle.
And all the while are we arguing about creating a digital copy of digital content. Even thinking about OCR here is like priting a screenshot and scanning it back in to send it via mail. Possible? Sure. Useful? Not at all..
