The Autonomy-Alignment Problem in Open-Ended Learning Robots: Formalising the Purpose Framework | Publicación