Hacker News Reader

My guess is the energy generated by the spinning hemisphere is less than the energy required to move his wrist when holding the magnet...

If that's the case, I find it counter intuitive: he barely moves the upper part of the mechanism while the lower part spins like crazy.