The space for a soul-of-the-gaps is becoming exceedingly thin. Machine intelligence can increasingly display all the phenomenological traits we associate with a mind. If not already, then in the near future AIs will pass the Turing test. Machine intelligence currently lacks the chemicals sloshing about in the bloodstream and the interoception in the viscera that entail emotional states, but there is nothing in principle to prevent us making an AI into a form of state machine with different behaviours according to the level of a "hormone".

As for the wetware-based alternative, we have known for centuries that subjective states can be radically altered by taking various chemicals. We can track the neural correlates for moving a limb. We can track the neural correlates for intending to move a limb. We also know, by the way, that such intentions can be detected before the conscious perception of a will to act. We can stimulate the brain to perceive an "out-of-body" experience. We can increasingly use electronics to place visual perceptions in a living brain, or track the activity of the visual cortex to see what the living brain is perceiving. I see no reason to doubt that the model of an object in the neurons of the visual cortex is the conscious perception thereof.

There is very little left for a "soul" to do, no evidence for its existence and no mechanism by which it would influence nervous activity. It also flies in the face of all we do know. Face it: Consciousness inheres in neurology, and there is a very strong case for those like Dennett and Damasio who regard the problem as solved.