<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Medicine · Joshua Reese</title>
    <link>https://epistemicstatus.com/medicine/</link>
    <description>Notes, experiments, readings, and occasional essays.</description>
    <language>en-us</language>
    <atom:link href="https://epistemicstatus.com/medicine/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Anatomy, Physiology, and the Physician</title>
      <link>https://epistemicstatus.com/anatomy-physiology-and-the-physician/</link>
      <pubDate>Mon, 10 Aug 2026 00:00:00 +0000</pubDate>
      <guid isPermaLink="true">https://epistemicstatus.com/anatomy-physiology-and-the-physician/</guid>
      <category>note</category><category>physiology</category><category>clinical reasoning</category><category>AI</category><description>&lt;p&gt;&lt;em&gt;Status: exploratory&lt;/em&gt;&lt;/p&gt;&lt;p&gt;Practicing physicians should make anatomic and physiologic reasoning the backbone of clinical thought. Guidelines, algorithms, and pattern recognition are indispensable, but they should sit on top of an internal model of how the human body is organized and how its systems behave—not substitute for one. That depth of anatomy, physiology, pathophysiology, and the ability to reason from first principles through an unfamiliar or rapidly changing clinical problem is precisely what the medical training pathway is meant to develop, and it should be one of the things that distinguishes physicians from clinicians trained through shorter pathways such as APP programs. The distinction matters most when the easy algorithm stops working: when several organ systems interact, when the patient does not resemble the population behind a guideline, or when the available recommendation simply does not make physiologic sense.&lt;/p&gt;
&lt;p&gt;The same habit of mind is a defense against other forms of borrowed authority. Guidelines can be overzealous, poorly fitted to an individual patient, or simply stale; expert consensus can harden into practice before the underlying reasoning deserves that confidence. Increasingly, AI adds another layer. A model can produce fluent, plausible clinical advice that is nevertheless internally incoherent, anatomically impossible, or physiologically unsound. A physician who has outsourced the underlying model of the body is poorly positioned to recognize that failure. Anatomy and physiology therefore matter not merely as preclinical hurdles but as durable tools for evaluating guidelines, consultants, algorithms, and AI output against reality.&lt;/p&gt;
&lt;p&gt;This is a placeholder for a much larger project and a substantial effort on my part: an attempt to say more precisely what physiological reasoning should mean in contemporary medical practice, how well physicians actually live up to that ideal, what distinguishes physician training when it works as intended, and why those questions become more—not less—important as increasingly capable AI enters medicine.&lt;/p&gt;
</description>
    </item>
  </channel>
</rss>
