Artificial intelligence is shifting the role of physicians from data interpreters to high-level decision-makers, a transition that author and data scientist Song Gil-young detailed during the 43rd Korean Medical Association (KMA) General Academic Conference. Song argues that as AI automates routine diagnostic tasks, the value of medical expertise will not diminish but will instead deepen, focusing on the complex human elements of care and the synthesis of AI-generated insights.
The discussion took place during the first session of the conference’s second day, where Song addressed a professional audience of doctors on the intersection of “fast civilization” (technology) and “slow civilization” (human-centric medicine). According to Song, the integration of AI into healthcare is not a replacement of the physician but a restructuring of what constitutes “professionalism” in a digital era.
This evolution is occurring as Large Language Models (LLMs) and specialized medical AI tools become capable of processing vast datasets faster than any human clinician. However, Song maintains that the ability to apply this data to a specific, nuanced human life remains a uniquely human professional skill.
The Shift from Information Retrieval to Insight Synthesis
For decades, a significant portion of a physician’s expertise relied on the accumulation and recall of medical knowledge. Song Gil-young posits that this “information-based” expertise is being commoditized by AI. When a machine can instantly retrieve the latest clinical guidelines or identify a rare pathology in a radiology scan, the doctor’s role shifts from being the primary source of information to the primary curator of that information.
According to the framework presented by Song, the “deepening” of expertise occurs when physicians move beyond the what of a diagnosis to the why and how of patient treatment. This involves synthesizing AI outputs with the patient’s personal values, socio-economic context, and psychological state—variables that AI cannot currently quantify or experience.
The World Health Organization (WHO) has highlighted similar themes in its guidance on the Ethics and Governance of Artificial Intelligence for Health, noting that while AI can enhance diagnostic accuracy, human oversight is essential to ensure equity, safety, and the preservation of the patient-physician relationship.
Redefining the Patient-Physician Relationship
Song argues that the automation of administrative and routine diagnostic burdens provides a critical opportunity for doctors to return to the “core” of medicine: the human connection. In a system often criticized for “factory-style” medicine where time per patient is strictly limited, AI can potentially reclaim the time necessary for empathetic listening and complex shared decision-making.
The “deepened expertise” Song references is not merely technical but relational. As AI handles the quantitative aspects of health—biomarkers, imaging, and genomic sequencing—the physician’s value increases in the qualitative realm. This includes navigating the ethical dilemmas of end-of-life care, managing chronic illness through behavioral change, and providing the emotional support that fosters healing.
This perspective aligns with the growing movement toward “Human-Centered AI” in healthcare, where the technology is designed to augment human capability rather than automate the human out of the loop. The goal is a hybrid model where the AI provides the evidence and the physician provides the judgment.
Challenges in the Transition to AI-Augmented Medicine
Despite the optimistic outlook on deepened expertise, the transition presents systemic challenges. The medical community must navigate the “black box” problem—the reality that some AI algorithms reach conclusions through processes that are not transparent to the user. According to industry analysis, this creates a tension between the efficiency of AI and the medical requirement for explainability and accountability.
Furthermore, there is the risk of “automation bias,” where clinicians might over-rely on AI suggestions, potentially eroding the very critical thinking skills Song argues will become the new hallmark of expertise. The challenge for medical education, therefore, is to shift from teaching students how to find answers to teaching them how to interrogate the answers provided by an AI.
The integration of AI also raises significant questions regarding liability. If a physician follows an AI’s recommendation that leads to an adverse event, the legal framework for accountability remains in flux across many jurisdictions. This uncertainty necessitates a cautious, phased adoption of AI tools in clinical settings.
The Future of Medical Professionalism
The trajectory outlined by Song Gil-young suggests a future where the most successful physicians are those who can act as a bridge between high-tech data and high-touch care. Professionalism will no longer be measured by the volume of knowledge a doctor possesses, but by their ability to apply that knowledge with wisdom and empathy in a way that improves patient outcomes.
This shift requires a fundamental change in medical training. Future curricula will likely place a higher premium on communication skills, ethics, and systems thinking, as the technical “memorization” phase of medicine is increasingly offloaded to software. The “deepening” of expertise is thus a move toward a more holistic, philosophical approach to healing.
As healthcare systems worldwide continue to integrate generative AI and predictive analytics, the focus will remain on maintaining the “human” in healthcare. The consensus among tech-forward medical thinkers is that AI will not replace doctors, but doctors who use AI will replace doctors who do not.
The medical community continues to monitor the rollout of AI-driven diagnostic tools and the resulting impact on clinical workflows. Further updates on the integration of these technologies are expected as more longitudinal studies on patient outcomes using AI-assisted care are published in peer-reviewed journals.
We invite readers to share their perspectives on the role of AI in their own healthcare experiences in the comments section below.