New technologies for recording neural activity, controlling neural activity, or building brain circuits, may be capable someday of serving in therapeutic roles for improving the health of human patients - enabling the restoration of lost senses, the control of aberrant or pathological neural dynamics, and the augmentation of cognition and empathy, through prosthetic means. High throughput molecular and physiological analysis methods may also open up new diagnostic possibilities. We are inventing new noninvasive methods for targetedly controlling brain dynamics in living human subjects, and also exploring novel ways of reading activity from the brain in noninvasive fashion. We are assessing, often in collaborations with other groups, the translational possibilities opened up by our technologies, exploring the safety and efficacy of our technologies in multiple animal models, in order to discover potential applications of our tools to various clinically relevant scenarios. New kinds of "brain co-processor" may be possible which can work efficaciously with the brain to augment its computational abilities, e.g. in the context of cognitive, emotional, sensory, or motor disability.