Roger Penrose is a brilliant thinker that offers some very interesting takes on consciousness and artificial intelligence. There is a great Lex Fridman YouTube interview of Penrose discussing where he stands on the computability of consciousness, so these quick notes originated from that interview and also from reading his book mentioned below.
In general, on Physics, Penrose mentions that Paul Dirac tacitly said that Quantum mechanics is a provisional theory. That we need still a theory that explains the collapse of the wave function. That Physics is essentially stuck. Quantum mechanics has to be modified, in an even-handed way (not just quantum gravity) to include both itself and gravity.
Regarding consciousness, he mentions that there are 3 key aspects of it: intelligence, awareness and understanding. One of the most approachable areas to understanding, he says, is that we do understand how consciousness can be “turned off”, as with anesthesiology. It is from there where his primary insights on the subject seem to come from.
In his 1989 book The Emperor’s New Mind he postulated that consciousness originates at the quantum level inside neurons, and although he admits his ideas are speculative it still sounds a bit too exotic. He also argued that human consciousness is non-algorithmic, and thus is not capable of being modeled by a computer. Nowadays we seem to be getting closer to AGI (Artificial General Intelligence), so most would say that that objection now appears to be incorrect.