New Learning’s Updates

Reverse Engineering Neuroscience

So, computer scientists like to use brain metaphors to support their aspirations to artificial intelligence ("neural nets," "Google Brain," and the like). But what happens when we use the methods of neuroscience to interpret the workings of the simplest of computers? We know how the computers work, so this might be a good test of the scientific validity of the methods of neuroscience. The result? Neuroscience methods mostly get things very wrong ... so claim resesearchers Eric Jonas and Konrad Paul Kording, as reported in The Economist and PLOS Computational Biology.

  • Richard Vines