AI programming is a long and complex process often made more difficult by shifting design demands and personnel turnover. Some of the difficulty, however, is of the AI team's own making. Worse yet, they often fall into the same traps over and over. Why do they keep making the same mistakes when developing AI? Often, the issues aren't technical at all, but spill out of the mindset, conscious or unconscious, that they bring to their projects. In this lecture, a professional AI firefighter shares her insight on the problems that teams keep experiencing and how to avoid them before they happen.