Thermodynamic constraints on information processing in living systems

Life can be understood as a non-equilibrium process driven by information handling. It is therefore of great importance to understand the constraints that thermodynamics imposes on the efficiency of information processing in a noisy environment. Recent progress in tackling this problem has been made by the use of generalizations of the second law to take into account information, within the general scheme of stochastic thermodynamics. This allows to derive a general lower bound for error correction that generally relates it with dissipation. I shall review these recent developments and illustrate the results by an application to kinetic proofreading, discussing the corresponding speed-dissipation-accuracy trade-offs.