Some of the best minds in biophysics have been divided on the question of how to model the workings of the human brain. Mainstream research has concentrated on developing complex neural networks to perform the cognitive processes that we call consciousness. However, others, including Roger Penrose at Oxford, have argued that consciousness is in fact a quantum effect and can only reasonably be simulated by a quantum computer.
The big issue in quantum consciousness is coherence time: how long can a superposition of quantum states in the brain persist? For the brain to be a true quantum mechanical system, the coherence time must be at least 1 s. Some say that it is, while others, including Stephen Hawking, say that it isn't. Now, Mark Tegmark of Princeton has weighed in with new calculations of coherence times for two cognitive models.
The first model describes the thought process as "neuron firing", where brain impulses are caused by neurons pumping out sodium ions and absorbing potassium ions. In the quantum picture, a neuron can exist in a superposition of firing and non-firing states.
In Penrose's alternative model, consciousness resides in microtubules hollow cylinders of protein that act as "scaffolding" to maintain the shape of brain cells. Using techniques from string theory, thoughts can be modelled as the collapse of a quantum superposition of electrical excitations propagating in a microtubule.
Whether the model involves neuron firing of charged ions or propagating excitations in microtubules, Tegmark calculated that the influence of nearby ions, by collision or coulomb interaction, will destroy the coherence of the superposed states on very short timescales (10-20 s for neurons and 10-13 s for microtubules) too short for a quantum system.
"There is nothing fundamentally quantum mechanical about the cognitive process in the brain", said Tegmark. For the neural network community "it's business as usual".
AIP