The problem with Phi is that it equates consciousness with a particular functionalism: networks / graphs. This base assumption then leads to the value range associated with Phi. The lower the connectivity, the lower Phi will be, and thus the lower the consciousness. If consciousness can arise from the interaction or processing of information, then we have to accept the numerous proofs of computational equivalence in information processing. We have to understand that there isnt something ontologically unique about a graph or network computation as opposed to another algorithm, unless Phi is asserting that a specific type of functionalism (and not information processing) is giving rise to the phenomenon — which isnt the assertion. The same error is made of those who hold the belief that parallelism would be somehow significant to the phenomenology of consciousness in an artificial system. #strongai #ai #consciousness
Posted on: Mon, 04 Nov 2013 21:33:46 +0000
Trending Topics
Recently Viewed Topics
© 2015