This is my final note and covers more chapters than the previous ones. Anil Seth proceeds to make a case for how our brain works by using Bayesian inference to make predictions of what both external and internal sensory inputs are going to be as well as actively influencing our subjective experience in order to make those predictions become real. In this way he explains how every part of our conscious experience ranging from external stimuli resulting in the perception of physical objects to our internal emotional states (the conscious experience of fear arises from rising blood pressure and other changes the body makes in response to a danger instead of the other way around for example) and even just the fundamental feeling of a self existing may be a "controlled/controlling hallucination" driven by this process. He also attempts to tie all of this into Karl Friston's Free Energy Principle which he gives a brief explanation of and views as consistent with his theory of consciousness.

Next the author gives his views on free will and an explanation for how the experience of free will fits into his theory. He attempts to sidestep the question of whether free will is an illusion by saying his theory works whether the universe is deterministic or not but also rejects a "ghost free will" that would somehow not be bound by the physical laws of the universe and would be making decisions independent of causality. He acknowledges that responsibility of our actions is a real question but argues for it being enough to establish responsibility that the physical processes in our brains are in control. He does not discuss his reasoning for this at length but it seems like some type of a compatibilist view.

I do think Seth makes a compelling argument that our decisions are still "ours" because they are ones that only we and the particular animal-machine we are could have made even if we could not really have done otherwise and they follow directly from the causality of everything that has come before and happened to us. Whether that establishes responsibility for them is another question.

At this point we are pretty much done with the central thesis of the book and the author uses the remaining pages for a discussion of what his views on animal (other than human) consciousness and AI consciousness is. Seth makes a case for most mammals probably being conscious based on the similarity of their brains to ours, behavioral observations etc. He also mentions birds as likely conscious and suggest a response to experiencing pain as a possible marker that an animal is conscious. He gives more examples of things that suggest different animals ranging from fish to ants could be conscious but that there is no conclusive evidence.

I thought he presented an excellent case for the dangers of both anthropomorphism and also anthropocentrism when trying to assess the consciousness of animals. This ties into his assertion that intelligence is a property orthogonal to consciousness and that anthropocentrism has led into us confusing the two. He criticizes AI researchers that make the implicit assumption that an intelligent machine would necessarily also be conscious as a side effect. He suggests that even an intellect superior to our own could be unconscious and that consciousness might be a property that has arisen as a consequence of how life functions and has evolved. He also ponders the difficulties we will face in the future with humans either having to deal with our intuition of attributing consciousness to machines that aren't or the ethical treatment of conscious machines that are very different from ourselves.