Braking Technology: Amazon’s Recommendation Engine

By Kathryn 

I imagine that, for John Eklund, writing this essay felt like stretching so thoroughly, every joint applauded with a ‘pop.’ From Wal-Mart’s much-superior forecasting skills to the loathsomeness of book clubs, Eklund rattles off everything about the book world that’s ever irked him.

Top on his list: Amazon (and Amazon-styled) recommendations, which he argues are v. v. “bad for reading”:

I don’t really have a problem with small-scale targeted book marketing. If I’ve bought six volumes of Irish poetry over the past year, and my local bookshop has been clever enough to quietly track these purchases, is it really such an imposition to receive a postcard alerting me to a new Seamus Heany collection? But the “customer-recommends” algorithm removes the pesky human from the interaction. And it does the exact opposite of what it claims to do: far from expanding my reading horizon, it contracts it. It doesn’t show me new worlds, it tries to duplicate as closely as possible the reading world I’m stuck in. When I’m offered “more like this” I want to scream NO! Not more like that. More like something else entirely, more like some other reader I’m nothing like, more like some new and different experience.

So … I’m guessing he doesn’t like gift cards?