TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

ChatGPT-4 Outperforms Physicians in Clinical Study

6 点作者 hotdailys大约 1 年前

2 条评论

damnesian大约 1 年前
This is like saying a dictionary knows more words than a human author. The premise is highly conditional. LLMs are tools that answer questions in human language. They aren't physicians. So the only way they can outperform a human doctor at this point is "on paper" only.
Bostonian大约 1 年前
My daughter will start college in the fall as a pre-med. I wonder what the progress of LLMs will mean for how doctors are trained and what their job prospects will be. If only for cultural reasons, I think there will continue to be many doctors and nurses, but knowing how and when to use LLMs will likely become an important skill for health care workers. In the near future, when a patient is admitted the hospital, doctors may have access to an LLM fine-tuned with the patient's medical records.