(New account, but I know the rules, so don't auto-downvote. Read. Thanks!) Strangely enough, magnets do weaken over time, and we can show it from conservation

(New account, but I know the rules, so don't auto-downvote. Read. Thanks!) Strangely enough, magnets do weaken over time, and we can show it from conservation of energy. If a magnet and paperclip are placed nearby on a desk, they will zip towards each other. This momentum has to come from somewhere. It cannot come from the paperclip, because paperclips do not attract each other, so it must come from the magnet. The magnet must therefore somehow store energy, and when its magnetic field causes motion, that motion must somehow consume the magnet's energy, even if the magnet's magnetic field doesn't appear to change. Commercially, magnets are produced by heating above a certain temperature[0] and then coercing the heated metal's magnetic fields into alignment. Since this process consumes energy, you could imagine that the equivalent amount of energy expendable by the magnet over its lifetime is bounded by that process. This might seem silly, but it has some corollaries which we can directly experimentally observe. For example, a nail gun correctly oriented relative to the Earth's magnetic field will temporarily magnetize ferromagnetic (steel, usually) nails while the nail is in motion, by the same sort of domain-alignment process involved in permanent magnetization. (Please do not do that experiment without taking proper safety precautions. Maybe we can talk Adam Savage into doing it for us.) In terms of calculations, we can use the "demagnetizing field"[1] but I think that it is not a very good name. Nonetheless, it does show that any time that there is a strong magnet, there is an automatic effect which will slowly slowly slowly weaken it. More generally, any maintenance done on a machine violates the closed-system assumption. You'd have to include the cost of maintenance if you wanted to show that you're violating thermodynamics. Because magnets are so long-lived, they might seem perpetual, but they are not. They are finite energy stores and they do run out eventually. [0] https://en.wikipedia.org/wiki/Curie_temperature https://en.wikipedia.org/wiki/Curie_temperature [1] https://en.wikipedia.org/wiki/Demagnetizing_field https://en.wikipedia.org/wiki/Demagnetizing_field