Title: Comment on the perspective article "Thermodynamic uncertainty relations constrain non-equilibrium fluctuations"

Abstract: In a recent perspective article, Horowitz and Gingrich discuss thermodynamic uncertainty relations that have been derived using "stochastic thermodynamics", a theory based on a hypothesis known as local detailed balance. The authors examined the foundations of this theory in their "Box 1: A brief primer on local detailed balance and stochastic thermodynamics", where a kinetic scheme for transport of particles between two reservoirs is presented. Horowitz and Gingrich arrive at a relationship between the probability to cycle through the states in one order vs. the probability to cycle in the reverse order. This relation bears on the extremely important question of what governs directionality when a system is driven away from equilibrium by contact with multiple reservoirs that are not in equilibrium with one another. In the original version of their paper the relation given for this ratio was obviously wrong and contrary to the second law of thermodynamics. Based on our private communications and on the recent paper authored by my colleagues and myself, Horowitz and Gingrich accepted the necessity of correcting the error in their Box 1. Unfortunately, in making the correction, the authors introduced an equally serious, if less transparent, mistake, and continue to base their theory on the thermodynamically impossible idea that transitions are mediated by only one of the two reservoirs. In this comment we illustrate how the principle of microscopic reversibility reveals that the true origin of directional cycling amongst a network of states is kinetic asymmetry. This understanding is important in guiding synthesis of molecular machines and other devices designed to exploit transport or catalysis to drive non-equilibrium processes.