{
  "id": 15877974,
  "title": "Handwashing changed medicine. AI could too, if hospitals stop treating it like traditional software",
  "url": "https://urgent.news/2026/10/21/handwashing-changed-medicine-ai-could-too-if-hospitals-stop-treating",
  "topic": "ai",
  "section": "AI",
  "published": "2026-10-21T10:33:00.000Z",
  "source": {
    "name": "Fast Company",
    "slug": "fast-company",
    "url": "https://www.fastcompany.com/91624590/handwashing-changed-medicine-ai-could-too-if-hospitals-stop-treating-it-like-traditional-software-ai-healthcare-technology"
  },
  "original_language": "en",
  "account": "In 1847, Dr. Ignaz Semmelweis introduced handwashing to a Vienna maternity clinic, drastically reducing patient deaths from 18% to 2% within two months. In healthcare, small changes can have a massive impact, as the system affects everyone and the stakes are life and death. AI has the potential to revolutionize healthcare, but institutions are failing to adapt to the moment. In the United States, a third of Americans have used AI for health information, with one in four using it in the past month. These tools serve as a form of unmet demand, as people seek more access to healthcare. Doctors are also adopting AI, but at an individual level. While this experimentation is normal, system-wide adoption is necessary for significant societal benefits. Many organizations struggle with AI, with 88% of AI pilots failing to reach production. Only about one in ten companies report productivity gains from AI. The main obstacle is the wrong mental model for AI. Healthcare is complex and data is fragmented, but progress is hindered by treating AI like traditional software. Leaders need to shift their mindset, seeing AI as something trained and grown, rather than written. Traditional software is deterministic, while AI is probabilistic and dependent on data. AI is built in loops, with most work happening before anything resembling a product exists. Traditional software management involves specifying the answer before verifying data can produce it. Developing AI systems requires different approaches, such as training on diverse data sets and avoiding shortcuts that limit model accuracy. A better approach, called \"wayfinding,\" involves navigating deep uncertainty without a known destination or route. This requires curiosity, humility, openness, and dedication, along with skills like psychological safety and problem selection. By adopting wayfinding, healthcare can effectively harness AI's potential while avoiding the pitfalls of traditional software management.",
  "summary": "In 1847, Ignaz Semmelweis asked the doctors in the Vienna maternity clinic where he worked to wash their hands in a chlorine solution before examining patients. Deaths on the ward fell from about 18% to about 2% within two months. In healthcare, small changes in practice can have disproportionately large effects, because the healthcare system affects everyone and the stakes are literally life and…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}