The use of artificial intelligence (AI) in medicine has seen exponential growth in recent years. While there are undeniable benefits to incorporating AI into healthcare, there is also a potential downside that must be carefully considered. A new viewpoint published in JAMA highlights the concern that over-reliance on AI may lead to a loss of clinical knowledge. In this article, we will explore the implications of this issue and discuss the importance of finding the right balance between AI and human expertise in medicine.
The authors of the viewpoint draw attention to the field of kidney pathology, where pathologists are relying on AI algorithms to analyze tissue specimens. While this streamlines the process and provides valuable data, it also raises concerns about the gradual loss of clinical knowledge. The authors argue that by moving the basic elements of pathology out of the pathologist’s view, there is a risk of diminishing the understanding of the kidney’s overall architecture. If pathologists are no longer required to evaluate the histology elements themselves, they may gradually lose the skill to do so.
Another key concern highlighted by the authors is the potential loss of problem-solving abilities in medicine. If AI models become heavily relied upon to solve medical problems and advance our understanding, there is a risk that human physicians will lose the capacity to think critically without the help of AI. The authors refer to a study that demonstrated how unsupervised AI models were able to identify tissue areas in kidney pathology that had not previously been recognized. However, little effort was made to understand these new findings. This lack of effort to comprehend newly identified constructs could lead to a rapid decrease in our knowledge of disease pathogenesis.
One aspect of AI that raises concerns is the concept of “black box computing.” This refers to situations where the AI system generates output without providing an explanation for how it arrived at its conclusions. The authors argue that if medicine reaches a stage where output is defined solely by the black box, our knowledge of disease mechanisms could be lost or forgotten. This scenario could result in a healthcare system solely focused on intervention strategies without a deep understanding of the underlying causes. Striking the right balance between AI and human input is crucial to prevent such a scenario.
While there are legitimate concerns regarding over-reliance on AI in medicine, it is worth acknowledging the significant potential of AI as a support tool in clinical decision-making. AI models have shown promise in assisting physicians and improving diagnosis. The authors of the viewpoint emphasize the need for physicians to recognize the importance of maintaining control over their profession when incorporating AI. It is crucial to find a balance that allows for meaningful progress in diagnosis and understanding of disease mechanisms while still preserving the expertise and critical thinking abilities of human physicians.
Artificial intelligence has the power to revolutionize healthcare, but it also poses risks if not approached with caution. The danger of over-reliance on AI in medicine lies in the potential loss of clinical knowledge and problem-solving abilities, as well as the risk of solely relying on black box computing. It is essential for physicians to carefully consider how to leverage the benefits of AI while preserving the vital role of human expertise. Striking the right balance will ensure the survival of the medical profession and meaningful progress in healthcare.
Leave a Reply