Thanks for the paper. Brownian computers are a cool idea, they seem like exploitation of the ratcheting paradigm.
Entropy increasing is a illusion based on the specific selection of macroscopic observables / slices of configuration space that we use as inputs for entropy. There is no information lost as much as some species that live on certain observable sense space slices, becoming ignorant and unable to exploit the new patterns of information flow. Chaos and order are two sides of the same coin. There will be brownian computers, funny enough, in chaotic environments. Whereas there might be very compact and sharply defined high efficiency solitary entities in low energy environments.
Entropy increasing is a illusion based on the specific selection of macroscopic observables / slices of configuration space that we use as inputs for entropy. There is no information lost as much as some species that live on certain observable sense space slices, becoming ignorant and unable to exploit the new patterns of information flow. Chaos and order are two sides of the same coin. There will be brownian computers, funny enough, in chaotic environments. Whereas there might be very compact and sharply defined high efficiency solitary entities in low energy environments.
See https://arxiv.org/abs/1505.01125