Thermodynamics is the cockroach of scientific theories - while new revolutions come and go, it stays squarely put. Or is it ? Do new-age thermal machines using quantum effects actually go beyond where Carnot dreamt ? 

A large chunk of my Ph.D. research was spent poring over this question. Together with collaborators I studied thermodynamics of systems with memory and saw indeed, this memory helps to partially reverse the flow of entropy.  We also saw quantum effects can indeed impact the performance of autonomous thermal machines like refrigerators, but alas, iron laws of thermodynamics seem to hold firm.