Hi Peter. Interesting connection to Godel's theorem and the ability to prove consciousness.

Indeed, consciousness is subjective, so how can one prove it to others? Perhaps there is a way, but to do it, one would have to understand it, and one would have to have access to the "machine" underneath, which you mention. That's the key insight of your article I think.

Regarding AI, we don't know if it could be conscious, and as you point out, it would not be able to prove consciousness - unless it had access to the underlying consciousness "machine".

People often misunderstand AI. Neural networks are not software: they are modeled after our own neurons. We run them as software simulations, but there are real neural networks that use neuromorphic chips. Those don't contain software. Perhaps those will have access to the consciousness "machine" - but there is no way to know! - at least until we understand what the consciousness machine is and how it works.