Overall, this is a nice piece and we believe the criticism is valid: a model can never be the modelled. In fact, models have deluded (e.g., geocentric solar system) and other models have fully blocked progress (e.g., Cartesian dualism).

These sentences are erroneous however:

“There are neurons whose task is to produce dopamine and flood the brain with it. This dopamine affects the way other neurons behave, and this is a crucial mechanism of reinforcement learning in the human brain.”

Dopamine doesn’t work like this: No flooding occurs. A presynaptic neurone releases it to a postsynaptic; at least according to accepted neurology.

Finally, placing hope that consciousness will be ‘found’ in electrical currents or quanta or whatever, seems as unwise as Hinton thinking it would be ‘found’ through nodes or Descartes in the pneumatic models of nervous systems in his day. If there is an ‘it’ of consciousness, it will be observed in the actions of persons not squishes and squirts of biology, in brains or elsewhere. More likely though, the concept of consciousness is, at best, a label for acts not some magical/mysterious thing causing those acts. When we stop reifying the concept, the magical bunk is revealed. A series of post we recently began addresses the issue. The overview is available here: besci.substack.com/p/bb…