Can’t remember the author and title of that book that fascinated you as a child, or even just a couple of years ago? Stump the bookseller to the rescue. Tell them what you can remember and, with any luck, they’ll get the answer back to you courtesy of a crowdsourced readership. The New York Times Magazine had a little write-up in their 19 March 2017 issue.

This is almost the opposite of algorithm-based recommendation systems: here you crowdsource the dark corners of readers’ memories in order to identify a book about which you’ve forgotten everything except for the couple of niggling incidents or characters you can’t get out of your mind. Pay your $4 and Loganberry Book’s site, Stump the bookseller will list your query where the multitudes can add precision.

I suppose this sort of thing isn’t worth programming up as an AI app. If only we were blessed with memory systems of the sort of precision that such an app might need to work with (though then of course we’d probably remember the title). Unfortunately the nature of human memory is such that what we remember is often actually wrong in some slight but critical detail buried within its cloudy vagueness. Computer programs like precision: human minds deal in cloudy nuance much better.

PMS 342

The one book whose title I might like to be reminded of is one of which I can remember nothing except that it had a cloth case in a dark green, sort of like PMS 342, was printed in black & white, had a few line illustrations, was in a smallish format, about 5″ x 7″ maybe, and also contained disfiguring drawings added by its youthful, and doubtless slightly bored, owner. Unfortunately that’d stump any crowd of bookseller’s friends. And besides, if I did know the title, what’d I do about it? Maybe check whether the book really was all that boring. I think that in the post-WWII period, when luxuries were less common, one would feel guilty for not enjoying a book on which someone had spent money.

Advertisements