Noise Robustness and Conscious Supremacy

Abstract

It is of interest whether artificial intelligence systems can emulate conscious information processing of the human brain. Here I discuss the aspects of conscious computation leading to artificial consciousness. One of the salient differences between biological neural networks and digital computers is the ubiquitous presence of noise in the former. I outline how noise robustness could correlate with conscious supremacy, where the brain exhibits computational capabilities excelling over classical computers, particularly in the presence of noise. Studying noise robustness and conscious supremacy in biological neural networks would help develop artificial intelligence systems of enhanced computational capabilities, with possible emergence of artificial consciousness, and implications for AI alignment.