Brains, it has recently been argued, are essentially prediction machines. They are bundles of cells that support perception and action by constantly attempting to match incoming sensory inputs with top-down expectations or predictions. This is achieved using a hierarchical generative model that aims to minimize prediction error within a bidirectional cascade of cortical processing. Such accounts offer a unifying model of perception and action, illuminate the functional role of attention, and may neatly capture the special contribution of cortical processing to adaptive success. This target article critically examines this “hierarchical prediction machine” approach, concluding that it offers the best clue yet to the shape of a unified science of mind and action.
Whatever next? Predictive brains, situated agents, and the future of cognitive science Andy Clark
Behavioral and Brain Sciences / Volume 36 / Issue 03 / June 2013, pp 181-204
James Glattfelder studies complexity: how an interconnected system -- say, a swarm of birds -- is more than the sum of its parts. And complexity theory, it turns out, can reveal a lot about how the economy works.
Imagine you receive a digital camera with a built-in memory card for your birthday. You bring it on a six-month trip to Africa where you won’t have access to a computer—so all the photos you want to keep must fit on that one memory card. When you first arrive you snap photos freely, and maybe even record some short videos. But after a month or so, the memory card starts filling up. Now you’re forced to be more judicious in deciding how to use that storage. You might take fewer pictures. You might decide to reduce the quality/resolution of the photos you do take in order to fit more. You’ll probably cut back on videos. Still, inevitably, you’ll hit capacity, at which point if you wish to take new photos you’ll have to delete old ones.
The maximum number of relationships we can realistically manage—the number that can fit on the memory card, as it were—is described as Dunbar’s Number, after evolutionary psychologist Robin Dunbar. But maybe it shouldn’t be. In the early nineties, Dunbar studied the social connections within groups of monkeys and apes. He theorized that the maximum size of their overall social group was limited by the small size of their neocortex. It requires brainpower to socialize with other animals, so it follows that the smaller the primate’s brain, the less efficient it is at socializing, and the fewer other primates it can befriend. He then extrapolated that humans have an especially large neocortex and so should be able to more efficiently socialize with a great number of humans. Based on our neocortex size, Dunbar calculated that humans should be able to maintain relationships with no more than roughly 150 people at a time. To cross-check the theory, he studied anthropological field reports and other clues from villages and tribes in the hunter-gatherer era. Sure enough, he found the size of surviving tribes tended to be about 150. And when he observed modern human societies, he found that many businesses and military groups organize their people into cliques of about 150. To wit: Dunbar’s Number of 150.