Quantum memory for light is one of the first significant inventions of this millennium. This technology has now been tested with important quantum information primitives: single photons, squeezed and entangled states. A big question remains though: what will remain of an *arbitrary* optical state after it has been stored and retrieved? Can it still be used for the purpose it was created with? In my talk, I will address our group's effort to answer this question and to implement quantum memory for exotic states of light.