Systems Research系统研究Oct 24, 20262026年10月24日Note短评

Topping a diagnosis benchmark is not being a safer diagnostician.诊断基准上拿第一，不等于就是更安全的诊断者。

A bright, sunlit hospital hall where staff in scrubs, seen only from behind, pin a huge gold first-place prize rosette onto a blank white board amid falling confetti, celebrating; to the right an empty hospital gurney with a folded blanket and a stethoscope sits unattended. Overlaid title: 'Aced the test. Never met the patient.' with a line of Chinese beneath.明亮阳光的医院走廊里，几位穿手术服、只见背影的医护人员在彩带中把一枚巨大的金色一等奖奖章别到一块空白白板上庆祝；右侧一张铺着叠好毯子、放着听诊器的病床空无一人、无人照看。叠加Image: WAKENEX图片：WAKENEX

An AI scored four times better than doctors on a diagnosis test, and won a far narrower contest than the headline. On October 22, 2026, Microsoft published a study in Science: its MAI Diagnostic Orchestrator, running on GPT-5, reached the right answer on 83% of hard New England Journal of Medicine case records, against 36% for generalist physicians who were allowed textbooks and the internet but not colleagues or specialists.

Read what the test measured. The cases are written case records, deliberately chosen to be hard, solved as text, with no real patient, no imaging and no uncertainty about what is wrong, scored by another model at the level of naming the disease. Microsoft's own orchestration added only about four points over GPT-5 alone, so most of the gap is the setup, not the system.

Microsoft says the real test plainly: a live clinical trial, announced with Beth Israel Deaconess and not yet run. So the 83% answers whether it can solve a curated diagnostic puzzle, not whether it is a safer diagnostician for the patient in front of you. We'd revise this if a prospective trial shows the same edge on real patient outcomes, not just disease-level accuracy.

Next time you read "AI beats experts at X," do one thing first: find what the benchmark measured, and ask whether anyone has tested it on the real thing yet. This is research, not medical advice.

一个 AI 在诊断测试上比医生准四倍，赢下的却是比标题窄得多的一场比赛。2026 年 10 月 22 日，微软在《科学》发表研究：它的 MAI 诊断编排器（运行在 GPT-5 上）在最难的《新英格兰医学杂志》病例档案上有 83% 答对，而被允许查教科书和网络、却不能找同事或专科医生的全科医生只有 36%。

但要看这个测试到底测了什么。这些是写好的书面病例，被特意挑成疑难杂症，以纯文本作答——没有真实病人，没有影像，没有“到底哪里出了问题”的不确定，而且由另一个模型在“说出病名”这一层打分。微软自家的编排相对单用 GPT-5 只多了约四个百分点，所以大部分差距来自这套考法，而不是这套系统。

微软自己也把真正的检验讲得很清楚：一场活体临床试验——已宣布与 Beth Israel Deaconess 合作，但尚未开展。所以这个 83% 回答的是“它能不能解开一道精选的诊断谜题”，而不是“它对你面前这个病人是不是更安全的诊断者”。如果一场前瞻性试验显示它在真实病人结局上（而不只是说出病名这一层）有同样的优势，我们就会修正这个判断。

下次你读到“AI 在某件事上超过专家”，先做一件事：弄清这个基准到底测了什么，再问有没有人在真实场景里验证过。研究，不构成医疗建议。

Built on the WAKENEX concept:所用的 WAKENEX 概念： Match the feedback to the question反馈分工：不要用一种指标替代其他问题的答案

WAKENEX / NOTESWAKENEX / 短评

Add WAKENEX as a preferred source on Google在 Google 把 WAKENEX 设为首选来源

READ NEXT接着读

OpenAI halved Terra's price, and the discount moved the cost onto you.OpenAI 把 GPT-6.1 Terra 的价格砍了一半，而这笔折扣悄悄把一项成本转移到了你身上。

Mochatrade sells Indian retail leveraged price exposure to US stocks.Mochatrade 卖给印度散户的，是对美股价格的带杠杆敞口。

NetSuite now lets you swap the AI model, not the permission gate.甲骨文旗下 NetSuite 的新说辞——“自带 AI”——是一道挪了位置的护城河，而不是一堵倒下的墙。