Tad presents an overview of the major issues stopping molecular machine development. There are three areas of progress for molecular machines – discovery, control of individual machines, and design behavior for large numbers of machines. The discovery process uses simulations to visualize machines we cannot yet build, and to use machine learning to design and fabricate conceptual machines. We have poor intuition for molecular machines with regard to thermal motion and viscous damping, making individual control difficult. For group control, swarm behavior is the dominant solution, and can be programmed or exist passively as a product of the nanobot design.