We investigate collective phenomena with active particles, and particularly heterogeneous systems where there are more than one species of active particles. For example, we studied rotationally driven spinners of concave shape, where each spinner experiences a constant internal torque in either a clockwise or counterclockwise direction. Although the spinners are modeled as hard, otherwise non-interacting rigid bodies, we found that their active motion induces an effective interaction that favors rotation in the same direction. They thus phase separate/demix as if they were oil and water. We are now investigating systems with mixtures of self-propelled particles, imagine for example two species of fish, if they were close to one another, would they form two separate schools, would they mix, would they form one big school and under what conditions?