The rapid advancement of Generative Artificial Intelligence (GenAI) sparked both excitement and concerns in the educational community and a renewed interest in the field of Artificial Intelligence in Education (AIED). While academic discourse has acknowledged both the opportunities and challenges of AIED, recent discussions appear increasingly polarised, with stakeholders often positioning themselves either in strong support of or opposition to initiatives under the AIED umbrella. This oversimplification overlooks the critical nuance that not all concerns surrounding AIED are equivalent. For instance, the implications of using conversational agents in classroom settings differ significantly from those of intelligent systems that sequence learning materials or analyse learner data to support pedagogical decisions. Without such nuance, the pedagogical potential of these technologies might lead to what might be described as "throwing the baby out with the bathwater." In this contribution, we synthesise some of the existing discourse across diverse contexts. Our aim is to offer a multifaceted perspective that neither dismisses legitimate concerns nor underestimates the transformative potential of AI in educational practice.