Command and Control (C2) is the practice of directing teams of autonomous units, regardless if the units are directed by human decision making or are unmanned. Humans are adaptive and their behaviors are recognizable to their superiors who were educated in a similar manner. This paper describes the sparse supervisory control that must be exercised over teams of highly autonomous units, and considers what it means for a commander to supervise autonomous un-manned systems (AUS) that employ machine learning and cooperative autonomy. Commanders must decide whether to trust behaviors they have never seen before, and developing that trust may require several strategies. This paper describes some possible strategies in an effort to explain the challenges that must be solved.