I will present recent work on using gerative AI and multiagent reinforcement learning to study and control fluid flows. I will also diuscuss limitations of AI in simulations of complex fluid flows and discuss possible fusions of AI and classical methods for flow modelimng, simulation and control.