Entropy, demons and statistical mechanics

Yesterday we discussed progress towards an understanding of causality in de Sitter space. The relation between thermodynamics and statistical mechanics - the latter being the study of an immense number of individual particles at the same time - has similar concerns that are still not fully ironed out.

Check out Philosopher of Science Katie Robinson's recent piece in Physics Today, "the Demons Haunting Thermodynamics". In it, Robinson not only recounts the famous demons of Loschmidt and Maxwell, but also related them to modern experiments attempting to manifest them. You might also enjoy the late Steven Weinberg's remarks about the associated troubles with quantum mechanics.

What's interesting about this field of study is the untangling of what is simply "too much data" : the positions and velocities of unfathomably many atoms and molecules in a macroscopic object, and what is "too little data" : beyond the fundamental resolution of the universe to know. For what it's worth, even Einstein couldn't keep this distinction straight.

Katie also presented the argument in a great talk, “In search of the holy grail: how to reduce the second law of thermodynamics.” You can see it on YouTube.