The Gardellab at the University of Chicago has realized an active liquid crystal where the active stresses could be modulated spatiotemporally through selective illumination with blue light. In the image, the region of high activity is enclosed within the red rectangle. The de Pablo, Dinner and Vitelli groups and collaborators, have constructed both hydrodynamic simulations and machine learning algorithms that faithfully predict the dynamics of liquid crystal with structured activity. The structured activity is used to guide the motion of defects (indicated by yellow arrowhead) and boundaries of low and high activity. Defects are constrained to regions of high activity.