Well rick obviously knows the answer.
But yes, his "paradox" is sort of complex to disprove. Let me try:
Let's assume the water is already steam in the first place, you just make it warmer (like in a hot air balloon). Just so we don't have to discuss phase transitions.
We can also assume the steam to stay in a bubble and not mix with the air (you sort of did that already).
When you expand the steam, you are working against the air pressure at that level. Then the steam rises, because the pressure gets lower the higher you go (that's what's generating lift, the pressure below the steam is higher than above, so the steam gets pushed up). If you then contract the steam, you get energy back, but less than you put into it to expand.
The difference between energy used for expansion and energy gotten from contraction comes in the form of potential energy: The water can fall down.
Note that while this is not a perpetuum mobile, we don't necessarily have to be the ones who invest into expanding the steam (or heating the water): The sun does that, conveniently. Rain comes from exactly this "paradox", the sun is the energy input though. People have used it for thousands of years. Thank the sun for watermills!
Lemme give you another one:
Imagine a 747 is sitting on a conveyor belt, as wide and long as a runway. The conveyor belt is designed to exactly match the speed of the wheels, moving in the opposite direction (the plane is stationary relative to the ground around the treadmill). Can the plane take off?