Abstract
The rapid development of text-to-image AI technologies is radically changing the way we approach image-making. This fundamentally changes the creative process acting both in early life and in higher education. In the paper, text-to-image technologies will be contextualised from a historical and technological perspective, attention will be focused on the space of ‘inaccuracy’ that they often traverse, and some considerations on digital poiesis in relation to creativity education will be proposed.