Much contemporary thought on technology in general, and translation technology in particular, is characterized by defeatism, determinism and a tendency towards universalism. The inexorable march of machine translation, we’re told, will turn us all into post-editors, while crowdsourcing will erode the professional basis of translation. But such comment does not pay enough attention to local differences, or the demands of specific languages and markets, and often little attempt is made to critique the practices that accompany technologization from a legal or ethical point of view. In this paper I consider how University programmes can help student translators prepare for a profession in which translation technologies may pervade, by helping them to develop not just technical skills, but also a high-level conceptual understanding of the technologies in question, and the critical ability that they will need to sustain careers in translation. My paper reviews a number of different translation studies responses to the challenges posed by technologization and especially by the rise of statistical machine translation (SMT). It draws on experience over the past four years of integrating SMT into the translation technology syllabus at Dublin City University, Ireland. It argues for a holistic treatment of technologies like SMT, one that involves translators at all stages of the translation workflow, and that takes account of the contexts in which technologies such as SMT are developed are applied.