When a Good Theory Meets a Bad Idealization: The Failure of the Thermodynamics of Computation
Summary
John D. Norton's essay critiques the thermodynamics of computation, arguing that Landauer's principle rests on flawed idealizations that ignore molecular fluctuations. Using thought experiments and historical debates (Maxwell's demon, Szilard's engine), it presents a no-go result: isothermal reversible processes cannot be achieved at microscopic scales without dissipation, calling for a more careful treatment of entropy, equilibrium, and phase-space reasoning.