Just bought a teapot off Amazon a week ago. It’s a lovely six-cup ceramic English-style pot glazed in a burnt red.
Today, Amazon on their homepage is recommending that I buy a teapot shaped like an elephant. Because of its powerful analytics and shopping habits algorithms, it’s absolutely sure that I’m a teapot collector and enthusiast. I must have this elephant teapot to complete my collection.
Wolfram is building Alpha, Microsoft announced Cortana, IBM is working on Watson. Cognitive Computing. Computers that are starting to process like humans do, trying to find relationships between things.
We’re reaching a point where Target can fairly guess if you’re pregnant or not, and adjust its suggestions to baby care.
But we’re not at the point where Amazon should really be recommending teas and steepers and teacups. Because those things are obviously related to my teapot purchase. A brick-and-mortar tea shop doesn’t have that problem because they have physical proximity of those items. But a website doesn’t have that benefit if it doesn’t know the relationship between teapots and tea and tea chinaware.
And yet there should be more than enough purchasing data in Amazon’s servers to let their computers start guessing correlations.
Heck, at this point, it should have known from the strange collection of purchases over the last six months—that I had just moved to a new city and was furnishing my apartment. It knows about my new delivery address, it knew I was buying household goods that I never bought before.
Ideally, after I bought the teapot, it should have recommended a good spatula. Because you know, I actually need one.
Dynamic Hypocrisy 