Google Now Indexes Scanned Documents (Scribd acquisition to follow shortly)

I am a mathematics student and sometimes need to look up papers. A lot of these papers are in journals not available online and have references to more of such papers. This can sometimes make even understanding a relatively simple proof that could be explain in a book in a few pages a gruesome all-day ordeal where 3/4th of the time is spent looking up references.

It would be really freaking cool if Google (Scholar?) could scan these articles into PDFs, then attempt to OCR the references and see if Google has indexed those, then embed hyperlinks to the references. This is technically feasible, and I imagine it would really make a lot of researchers very happy.

>> Scribd acquisition to follow shortly

Really.