Uncertainty is another word for possibility, the condition that all education seeks to fit us for—or should, if it is truly education for deeper learning devoted to the expansion of human capacities. Much so-called educational reform, however, seeks to prepare students not for possibility, but for certainty, by mapping what are called “learning outcomes” to specific skills and “competencies” aimed at narrow and reductive models of “workforce preparation.” Following a neo-behaviorist paradigm, these educational strategies focus on the learner’s performance of specified tasks within well-defined problems. Such strategies are disastrous, I will argue, because education in a time of accelerating change must engage with unanticipated tasks and new domains of inquiry in a context of proliferating “wicked problems.” My presentation thus advocates four fundamental understandings that are necessary for any strategy of education for uncertainty. The first is an understanding of complexity: what it is, and how it differs from complication. The second is an understanding of what Brian Arthur calls the “nature” of technology. The third is an understanding of learning, including the vital and sometimes tense coexistence of networks and intimacy. The fourth is an understanding of understanding itself, the deepest of all learning outcomes, and yet also the learning outcome that dares not speak its name.