Kuchling, Friston, Georgiev and Levin ([1]; hereafter KFGL) show how to construct, starting from generic classical physical assumptions, a model of a cell, or of any biological system, as an… Click to show full abstract
Kuchling, Friston, Georgiev and Levin ([1]; hereafter KFGL) show how to construct, starting from generic classical physical assumptions, a model of a cell, or of any biological system, as an inferential agent that acts to minimize Bayesian surprise. Key to this construction is the concept of a Markov blanket, defined in KFGL, §2.2.3 as the Cartesian product of the sets of “sensory” and “active” states of the cell / system. The existence of the Markov blanket assures the conditional independence of “external” and “internal” states that is required if the probabilities of external and internal states used in Bayes’ theorem are to be well-defined. As KFGL note, “Most fundamentally, we have assumed the existence of a
               
Click one of the above tabs to view related content.