We question the current framing of Artificial Intelligence (AI) under a mechanistic framework and how this produces blind spots preventing digitalization and AI from developing their potential to address the most pressing challenges of humanity. Some ideas are proposed for a due reframing around a complexity framework.