The article argues that the accuracy of medical AI should not be confused with efficiency. While AI systems may be accurate in tasks like radiology or pathology, this accuracy does not necessarily translate to efficiency gains. The development and integration of AI systems require significant human labor and resources, and ongoing maintenance and training are necessary. Moreover, AI systems can make mistakes, leading to potential harm to patients. The article emphasizes that the human side of technology implementation is crucial for efficiency, as the knowledge, skills, and trust of healthcare professionals influence the effectiveness of AI. Trust in AI systems can lead to overreliance, which may result in errors or inefficiencies. The article also highlights that the systemic effects of AI are complex and may not be fully understood until long after implementation. It calls for a more critical and nuanced discussion of AI's benefits, distinguishing between effectiveness and efficiency, and emphasizing the need for further research on trust and its relationship to efficiency. The authors conclude that it is important to remain conscious and critical about how AI is discussed in terms of its potential benefits, especially when referring to systemic changes based on single studies.The article argues that the accuracy of medical AI should not be confused with efficiency. While AI systems may be accurate in tasks like radiology or pathology, this accuracy does not necessarily translate to efficiency gains. The development and integration of AI systems require significant human labor and resources, and ongoing maintenance and training are necessary. Moreover, AI systems can make mistakes, leading to potential harm to patients. The article emphasizes that the human side of technology implementation is crucial for efficiency, as the knowledge, skills, and trust of healthcare professionals influence the effectiveness of AI. Trust in AI systems can lead to overreliance, which may result in errors or inefficiencies. The article also highlights that the systemic effects of AI are complex and may not be fully understood until long after implementation. It calls for a more critical and nuanced discussion of AI's benefits, distinguishing between effectiveness and efficiency, and emphasizing the need for further research on trust and its relationship to efficiency. The authors conclude that it is important to remain conscious and critical about how AI is discussed in terms of its potential benefits, especially when referring to systemic changes based on single studies.