Both claims\u2014that AI has become compassionate and that physicians are becoming obsolete\u2014arise from the same evidence, but both are wrong. The algorithm did not become compassionate; however, the medical profession has drifted so far from the bedside that an AI language model (trained on pattern and probability) can now outperform physicians on the very quality that medicine assumed would remain uniquely human.1<\/a><\/sup> Something has gone terribly wrong in medicine.<\/a><\/p>\n\n\n\n The problem did not arise because AI crossed some final human threshold. Clinical work has been reorganized around tasks that pull clinicians away from direct patient care, but no single policy removed the physician from the bedside. The transformation happened fraction by fraction via documentation, performance metrics, prior authorization, billing, and inbox work (Figure<\/a>), but fractions compound. The administrative work accumulated regulation by regulation and click by click until the medical profession no longer recognized what it had lost. Over 4 decades, the practice of medicine buried physicians under this new process with an increased focus on indirect patient care. In ambulatory practice, physicians spend 49% of their office day on electronic health record and desk work and only 27% in direct clinical contact.2<\/a><\/sup><\/a><\/a><\/p>\n\n\n\n Figure. \u00a0Illustration of How Artificial Intelligence Can Assist Clinicians With Patient Care<\/p>\n\n\n\n Four decades of administrative accumulation, electronic health record (EHR) documentation, billing and medical coding, prior authorization, performance metrics, and inbox management have buried the patient encounter beneath layers of work, displacing the physician from the bedside (left). Artificial intelligence (AI) does not replace the physician but excavates these layers, restoring direct presence at the point of care (right). The return creates a governance loop: the physician who is present can exercise oversight, judgment, and integration over AI-augmented clinical processes, which are functions that cannot be performed from a distance. Throughput is maintained not by accelerating the work of the physician but by redistributing the work that never belonged at the bedside.<\/p>\n\n\n\n The same administrative burden structure also erodes the time of nurses, pharmacists, therapists, and others whose work likewise depends on their presence with patients. Burnout is one consequence. In a national survey,3<\/a><\/sup>physician burnout remained high from 2011 to 2023, peaking at 62.8% during the COVID-19 pandemic and decreasing to 45.2% in 2023 (a return to baseline but not resolution). Administrative displacement is not the sole cause, but it is among the most remediable, and it is the one AI addresses most directly.<\/a><\/p>\n\n\n\n The deeper injury is vocational. Administrative medicine has stripped physician autonomy, recast competence as compliance, and interposed the screen between the clinician and patient, severing the human relatedness that made practicing medicine feel like a calling.4<\/a><\/sup> What followed has not been mere inefficiency but the progressive extinction of the conditions under which medicine can be practiced. That is the context in which the literature on empathy should be interpreted.<\/a><\/p>\n\n\n\n AI chatbots were rated substantially higher on empathy than physicians in a meta-analysis1<\/a><\/sup> of 13 pooled studies (pooled standardized mean difference, 0.87 [95% CI, 0.54-1.20], P<\/em>\u2009<\u2009.001 favoring the AI chatbots), but the included studies were heterogeneous, publication bias could not be excluded, and all comparisons were text-based rather than bedside encounters. No AI chatbot has examined a patient, been able to interpret a patient\u2019s facial expressions, or carried out a plan of care through patient uncertainty. The highest empathy ratings occurred when AI-generated text was attributed to a physician in an experimental study.5<\/a><\/sup> People wanted to believe their physician wrote something so caring and the algorithm met their expectation, but the physician had not.<\/a><\/p>\n\n\n\n The prevailing prediction is that AI will finish what administrative medicine began and substitute the physician entirely, but that gets the technology exactly backward. AI cannot replace the physician, but it can remove the burden of what replaced the physician\u2019s clinical time, which we refer to as the excavation (Figure<\/a>). Documentation, coding, triage, and pattern recognition are essential but substitutable. A physical examination, a hand on the shoulder, and the willingness to stay with patients and listen to their fears are irreducible and nondelegable.<\/a><\/p>\n\n\n\n The essential excavation work was not a waste in that it reduced errors and built the data infrastructure on which modern clinical research depends. However, these tasks continued to be assigned to humans long after AI could assist or absorb them. When health systems merely shift the work to a larger team, the administrative burden simply moves from one clinician to another. AI is the first tool that can absorb much of the administrative burden rather than simply redistribute it, freeing every clinician for the work each was trained to do. This is not aspirational. Large language models already match or exceed resident physicians in diagnostic accuracy under time pressure, which is precisely the condition in which human reasoning degrades.6<\/a><\/sup> A physician whose cognitive reserves are spent on documentation has less capacity for the unpredictable moment in patient care. AI does not replace physician judgment, it preserves it.<\/a><\/p>\n\n\n\n The physician at the bedside must remain accountable for governing the AI system now operating, together with the rest of the clinical team. An absent physician cannot audit an algorithm, but a present one can. The risk is not AI itself but AI integrated without clinician governance or redesign of the work around it. AI creates a fork in the road. One path leads back to the bedside and the other leads further from the bedside (ie, faster documentation and higher throughput with the same absent clinician now more efficient at being absent). The technology does not choose, but the profession does.<\/a><\/p>\n\n\n\n AI deployed for throughput rather than physician presence generates new summaries to review, new inbox burden, and new layers of downstream bureaucracy, pushing clinicians further from patient care. All of which is not better care, but a faster, thinner, less accountable patient-physician encounter. The choice belongs to clinicians, educators, and health systems, yet too often it is being made in their absence. For clinicians, automated charting could become merely accelerated volume. For educators, physician training could shift toward what no AI model can do by encouraging close observation, embodied examination, tolerance of uncertainty, and active listening beyond the patient\u2019s chief concern to the life beneath it. For health systems, the lever is not the technology but what the organization chooses to value. Among more than 20\u202f000 clinicians surveyed, feeling valued by one\u2019s organization was the strongest mitigator of burnout, reducing the odds by 78%.7<\/a><\/sup> When the metrics are volume and revenue alone, clinicians receive the message that their presence does not matter. Time with patients, continuity of relationships, and patient experience of being heard must carry weight in whatever counts as productive work.<\/a><\/p>\n\n\n\n AI does not humanize medicine, but it can remove one of the most durable excuses for failing to do so. Each administrative task the algorithm absorbs is time the profession can reclaim for the clinical work that made medicine a vocation. The bedside was always the foundation, but in building everything above it we simply forgot that we were standing on it. AI is not the end of the physician\u2014it is the return of every reason to be one.<\/p>\n","protected":false},"excerpt":{"rendered":" Viewpoint AI in Medicine Artificial Intelligence […]<\/p>\n","protected":false},"author":3,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":[],"categories":[24,23],"tags":[],"_links":{"self":[{"href":"https:\/\/csccm.org.cn\/index.php?rest_route=\/wp\/v2\/posts\/30845"}],"collection":[{"href":"https:\/\/csccm.org.cn\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/csccm.org.cn\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/csccm.org.cn\/index.php?rest_route=\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/csccm.org.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=30845"}],"version-history":[{"count":2,"href":"https:\/\/csccm.org.cn\/index.php?rest_route=\/wp\/v2\/posts\/30845\/revisions"}],"predecessor-version":[{"id":31288,"href":"https:\/\/csccm.org.cn\/index.php?rest_route=\/wp\/v2\/posts\/30845\/revisions\/31288"}],"wp:attachment":[{"href":"https:\/\/csccm.org.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=30845"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/csccm.org.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=30845"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/csccm.org.cn\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=30845"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}![\"\"\/](\"https:\/\/cdn.jamanetwork.com\/ama\/content_public\/journal\/jama\/939842\/jvp260027f1_1778787871.17331.png?Expires=1785932927&Signature=tBDyF1SCepYpq2HcQTaUFQlir8XCLunv7WZoR1kLz69aoaKM2tQ5rInP3IsZwTy4VY0VOKL-1BdlKSEnnDR3IDuQaLMwnXy-ydFWCc2OgPEJNIZSU9qnJ0ZONouflWUPNrHgmHk6LDkmRlmqwEQiW8XxOzSYpOxR6IYBJrWBoZFL074Ko69keaw~11U8CNAat6gbVwq-UDYCSrKntFeI18dpXuN7-n-0UdkyXV~~e-ays~G-ns14gZ0AZkkDRp6-LcxpMHeo68Zyt3LQ7ATw9VNFg92mAnoBohAEEfBYKEZ0P-4kV~BlmF9Yy-BxzpqCuwauDVcUcfq4ADXWPmHJ3A__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA\")
<\/figure>\n\n\n\n