It seems to me that if you have a bar magnet, rotating in a horizontal plane about its centre so that an observer to (say) the west felt the effects of each pole in alternation as it spins, it would be possible to draw energy from the system in at least 2 ways:
1) a loop of wire leading to a load could be arranged to collect energy and dissipate it in the load (this is how electricity is generated commercially), or
2) a mechanical device composed of a spring and a piece of steel could be made to oscillate back and forward under the applied magnetic field and work would be done.
In either case, the work done by the mechanism, be it electric or mechanical, would be felt by the spinning magnet, and to the extent that energy (work) is supplied to the mechanism it would be lost by the rotating magnet.
Please notice that the air (and empty space) is a a very poor conductor of magnetic flux, so that these effects are usually useful only over very small distances.