Sublime

So Bezos suggested that the personalization team develop a much simpler system, one that made recommendations based on books that customers had already bought. Eric Benson took about two weeks to construct a preliminary version that grouped together customers who had similar purchasing histories and then found books that appealed to the people in e

The symbolists’ master algorithm is inverse deduction, the connectionists’ is backpropagation, the evolutionaries’ is genetic programming, the Bayesians’ is Bayesian inference, and the analogizers’ is

probabilities attach themselves to the descriptions of events and not to events themselves.

Netflix came to a similar conclusion for improving its recommendation algorithm. Decoding movies’ traits to figure out what you like was very complex and less accurate than simply analogizing you to many other customers with similar viewing histories. Instead of predicting what you might like, they examine who you are like, and the complexity is ca

peer recognition

As knowledge grows, scientists specialize ever more narrowly, but no one is able to put the pieces together because there are far too many pieces. Scientists collaborate, but language is a very slow medium of communication. Scientists try to keep up with others’ research, but the volume of publications is so high that they fall farther and farther

case. If people were independent, each making decisions in isolation, societies would indeed be predictable, because all those random decisions would add up to a fairly constant average. But when people interact, larger assemblies can be less predictable than smaller ones, not more.