Gary Marcus (NYU) thinks that field programmable gate arrays might be our best bet for thinking about neural computation (NYTimes):
FIELD programmable gate arrays consist of a large number of “logic block” programs that can be configured, and reconfigured, individually, to do a wide range of tasks. One logic block might do arithmetic, another signal processing, and yet another look things up in a table. The computation of the whole is a function of how the individual parts are configured. Much of the logic can be executed in parallel, much like what happens in a brain.Although my colleagues and I don’t literally think that the brain is a field programmable gate array, our suggestion is that the brain might similarly consist of highly orchestrated sets of fundamental building blocks, such as “computational primitives” for constructing sequences, retrieving information from memory, and routing information between different locations in the brain. Identifying those building blocks, we believe, could be the Rosetta stone that unlocks the brain.To put this differently, it is unlikely that we will ever be able to directly connect the language of neurons and synapses to the diversity of human behavior, as many neuroscientists seem to hope. The chasm between brains and behavior is just too vast. […]
If neurons are akin to computer hardware, and behaviors are akin to the actions that a computer performs, computation is likely to be the glue that binds the two.