It seems possible to extract energy from the ambient air without temperature gradients (or at least with very small local gradients that will always be available). Being curious about this, I checked out some other threads and did not really find the specific question. (I did find some interesting posts about Maxwell's Demon, etc.) Anyway, this is the concept: The ambient thermal energy created by the Sun, the Earth's core, gravity, etc., are in abundance and there are decent ways of extracting it. According to the posts, however, it appears that thermodynamic theory suggest that extraction of thermal energy requires a temperature gradient. This seems counterintuitive. It seems that the molecules are flying around waiting to be put to work and constantly being re-energized by the aforementioned external sources. Statistically, it makes sense that one would not get too far trying to do anything useful since they are likely to exert the same pressure in all directions on any object. However, couldn't one imagine a very thin pizoelectric device sticking out of a box and being randomly strained in various random directions by the air molecules? Wouldn't this solve the statistical issue and thereby "get around" the thermodynamic theory?