Andy Clark in How to Knit Your Own Markov Blanket: Resisting the Second Law with Metamorphic Minds writes:
“In a revealing twist on his original (2016) demon-description, Hohwy (Hohwy 2017) further shows that to the extent that all this is true, the prediction-error minimizing agent is in fact manipulating the demon (the intelligent vat) by giving action commands whose fulfillment requires the demon to compliantly alter the apparent flows of energy across her (apparent) sensory surfaces.”
My question is, why would a demon care to be pushed around by us in this manner? Why not just show us the truth of our illusion? If he’s a demon, and demons are evil, I would see that as instead the most evil thing to do, for humans hate the truth, and they hate it because they fear it.
Show us the truth, not an illusion, and you’ve done us evil.
This is because we are quite safe in our delusions, and it makes me think instead that philosophy has gotten it wrong all along…
The demon is actually an angel.
Clark, A. (2017). How to knit your own Markov blanket: Resisting the second
law with metamorphic minds. In T. Metzinger & W. Wiese
(Eds.) Philosophy and Predictive Processing.