Algorithms

According to an article in The Economist of September 15 (‘Business by numbers’) algorithms have become the instruction manuals for a host of routine consumer transactions. Amazon, for example, uses algorithms to help the company recommend further purchases ‘in the neighborhood’ of your new purchase. The logistics firm that will deliver the Amazon product to your door will then use an algorithm to identify the best delivery route. And the call centre in Mumbai dealing with your complaint to BT will use an algorithm to figure out from your voice which part of Yorkshire you come from before matching you up with a plausible Bangalorian. The latter, in turn, will be algorithmically instructed to fool you into believing that he lives just across the road from you and knows what school your kids go to.

Algorithms make sense of unstructured data, something apparently we are not short of. Given the vast increases in computing power now available, algorithms can extract unexpected information from your laundry tickets, your taxi receipts, and the X-rays of your dental cavities. By performing sophisticated correlations on these disparate bits of data, firms can now offer you product customized to make sense of… well, your laundry tickets, your taxi receipts and the X-rays of your dental cavities, what else? Algorithms allow firms to exploit what Chris Anderson has labelled the long tail – those micro-niches of the market that would have been ignored in earlier times.

As a market segment of one – the mother of all micro-niches – I look forward to algorithmically-derived products that can make sense of:

1. My choice of All bran for Breakfast cereal related to my preference for a room temperature of 22 degrees Celsius.
2. My tendency to drink Diet Coke when listening to Beethoven (“People who listened to Beethoven also drank….”)
3. My preference for the seats 22c in wide-bodied aircraft and 35d in Boeing 737s.
4. My urge to scream when Amazon informs me of what other people have been reading.
5. All of the above correlated with all of the above.

Algorithms are designed to ‘take a load off your mind’. But at what point to they begin to take all the load of your mind – ie, make you mindless? I guess that you’d have to be pretty mindless to be able to make sense of 1-5 above.