
31 Aug
2011
31 Aug
'11
9:58 p.m.
i sincerely hope michael pulls another rabbit out of his hat...
Not exactly sure what would be pulled by who where. But Google is making progress towards "reasonable" 100% human-free OCR translation of the body text of books, not including "the hard parts" such as TOC, Index, etc. and then only on relatively recent pubs such as circa 1900s. One could imagine a system that accepted such "good enough" OCRs and put them online immediately, and which allowed real-world readers to propose fixes as they read the book. You could still require PG volunteers to review the proposed fixes before accepting them.