Eliezer wrote:
> There are only two logically consistent positions: CCC and PPP.
>
NNN might also be proposed, but one would hope it turns out to be false.
I would expect that it could be possible, for example, to produce and simulate consciousness computationally, but not explain it. To explain something you must often go outside the system. If consciousness is an emergent phenomenon, it might only be explainable/understandable in N, even though producable and simulatable in C (Hofstader?). But then this is CCN, which you have attributed to Chalmers.
Eliezer also wrote:
My position is simple: No Zombies. Only single-letter codes are logically consistent. Every facet of consciousness we can notice, that we can *have a need to explain*, obviously influences our physical actions. If there's any quality of consciousness that doesn't influence our actions, nobody has ever written a paper about it.
Here, you are stating that if we (consciousnesses) can explain consciousness, and consciousness can be produced+simulated computationally, then an explanation of consciousness can be computational. The flaw as I see it is that consciousness is not explainable (so far) by us, or any other consciousness that we know of (?). Perhaps it can never be. We may be able to be aware of our consciousness (necessary to thought?), yet be entirely unable to fully explain that quality (and what is a partial explanation but an isomorphism of the unexplainable?).
By the way, why PPP rather than CCC?
Emlccn