It can read the labels and the position of the book using its image processing and optical character recognition software," the professor said. Once the book is located, it has to grasp it and take it off the bookshelf, which is not a simple as it might seem. For this, the team had to develop special fingertips like a nails, with one nail longer than the other.
Professor Pobil said it was a"real possibility" that teams of robots could, in about five years' time, realistically perform searching and fetching tasks. They could even mill around doing their work at night, working on library inventories, or identifying missing books, or mapping libraries.
I don't see this impacting the role of the librarian but the work of hundreds of work-study students. My question is how will the OCR treat non-English text? And will this intervention of the AI give us newer ways of cataloguing and stacking? Like towers of books reaching skyhigh...
And as to my reference to Historians....
2. Take the algorithm that runs Google News which aggregates information without any human intervention.
+ 2. Add to it Amazon's A9 engine which has hundreds of thousands of books scanned.
= Wait patiently for the AI to improve for the next 5 years.
4. Voilá. A Historian Engine that can process through any set of academic data and spit out a rational, logical arrangement of facts and analysis. No?
Ok. I know I am reaching here but given that most mechanical jobs have or will soon be lost to robotic industry, how long before the Ivory Tower comes under seige? Do the publishing industry really needs a Ph.D. to come out with A Quick Guide to Irish History when a program can arrange the facts into a simple narrative?
We already have the AI to piss off republicans.