In physical/chemical systems metastability and local _minima_ (of an energy function) are the same. But for cases where the thing being optimized is not energy, and where the universe isn't driven to the optimal state, do they need to be the same?