The paper "On the Limits of Artificial Intelligence (AI) in Education" by Neil Selwyn addresses the growing hype and hyperbole surrounding AI in education, highlighting critical issues that need to be considered in future discussions. The author argues that the current enthusiasm for AI in education is driven by financial interests, political support, and technological advancements, but this enthusiasm often lacks a balanced and reasoned approach. Key concerns include the limited statistical modeling capabilities of educational processes, the potential for AI to perpetuate social harms for marginalized students, the ecological and environmental costs of data-intensive AI, and the need to slow down and recalibrate discussions around AI in education. Selwyn emphasizes the importance of recognizing the limitations of AI, which are primarily statistical and computational. He points out that AI systems are designed to address specific tasks and operate within predefined boundaries, making them brittle and context-specific. The paper also discusses the social harms of AI, such as algorithmic discrimination, quality-of-service issues, and the adverse impact on social relations in educational settings. Additionally, the paper highlights the environmental burden of AI, including the high carbon emissions and resource consumption associated with its development and use. The author calls for educators to take control of the agenda surrounding AI in education, engaging in open and sustained dialogue that addresses power dynamics, resistance, and the potential for reimagining education AI along more equitable and beneficial lines. He suggests that educators should challenge the hype and focus on the practical outcomes of AI technology, advocating for the development of AI that aligns with 'green-tech' principles and addresses the pressing environmental and social issues of our time.The paper "On the Limits of Artificial Intelligence (AI) in Education" by Neil Selwyn addresses the growing hype and hyperbole surrounding AI in education, highlighting critical issues that need to be considered in future discussions. The author argues that the current enthusiasm for AI in education is driven by financial interests, political support, and technological advancements, but this enthusiasm often lacks a balanced and reasoned approach. Key concerns include the limited statistical modeling capabilities of educational processes, the potential for AI to perpetuate social harms for marginalized students, the ecological and environmental costs of data-intensive AI, and the need to slow down and recalibrate discussions around AI in education. Selwyn emphasizes the importance of recognizing the limitations of AI, which are primarily statistical and computational. He points out that AI systems are designed to address specific tasks and operate within predefined boundaries, making them brittle and context-specific. The paper also discusses the social harms of AI, such as algorithmic discrimination, quality-of-service issues, and the adverse impact on social relations in educational settings. Additionally, the paper highlights the environmental burden of AI, including the high carbon emissions and resource consumption associated with its development and use. The author calls for educators to take control of the agenda surrounding AI in education, engaging in open and sustained dialogue that addresses power dynamics, resistance, and the potential for reimagining education AI along more equitable and beneficial lines. He suggests that educators should challenge the hype and focus on the practical outcomes of AI technology, advocating for the development of AI that aligns with 'green-tech' principles and addresses the pressing environmental and social issues of our time.