"The idea that the brain is computational has from the outset been central to neuroscience. The brain is a problem-solving machine operating according to the principles of mathematical logic, and stores memories, and processes information, as a computer. But despite the advance of AI, an increasing number challenge this assumption. Computers use transistors to build elementary logic gates, enabling them to store files exactly, in 0s and 1s. They are precise and perform identically on each occasion. Human brains are biological, the neurons do not operate as simple logic gates but have thousands of inputs and their output is dependent on past activity and its current internal state. Remove a computer’s processor, and it breaks. But humans can survive with only one brain hemisphere. Fundamentally, brains think and enable perception and consciousness while computers do not.
Is it a mistake to see the brain as computational and do we need to abandon the analogy? Are computers at root dumb machines with little in common with the sophistication of the brains of living things? Or have computers uncovered the essential character of all thought?"
