A university in China has launched an artificial intelligence (AI) hospital called “Agent Hospital,” where virtual doctors can treat up to 3,000 patients a day.
Tsinghua University, the so-called Harvard of China, is attempting to revolutionize healthcare by transforming the way doctors diagnose and treat patients.
The hospital is all virtual – all doctors and nurses are driven by large language models (LLMs) that autonomously interact with each other. There are 14 different doctors and four nurses – a virtual medical staff that works together to consult, examine, diagnose, and treat virtual patients.
As per Global Times, AI doctors can treat 10,000 in just a few days. It would take a human doctor a couple of years to see that many patients.
The leading physician of the hospital, MedAgent-Zero, stores successful treatments and is learning from misdiagnoses.
Yes, the AI doctors aren’t seeing human patients. However, this project is believed to have a revolutionary effect on the health industry.
Would the AI hospital town bring practical benefits to the real healthcare industry? The answer is yes.
Scientists behind the project say it will enhance medical training. Students can treat virtual patients without fear of causing real harm.
The AI hospital can simulate and, therefore, help future doctors prepare for various scenarios, such as the spread of new infectious diseases.
As per Global Times, the AI hospital is “nearing readiness for practical application.”
An intriguing project
AI strategist Emanuelis Norbutas expressed cautious optimism about the potential implementation of similar projects in the healthcare industry.
“There are many questions to answer before we need to worry about AI replacing human doctors. However, if models like these are set up correctly, with robust learning cycles and proper implementation, they could soon serve as digital assistants for physicians, providing well-informed guidance when final decisions are made,” Norbutas said.
He believes that such examples highlight one of AI’s greatest strengths: the democratization of information and learning.
“Medicine is a highly specialized field, and any technology that improves the learning process, expands the expertise of human doctors and in that way increases their numbers, deserves serious attention,” said Norbutas.
He added that the project is even more intriguing from a technical standpoint. “What’s innovative here is the use of multi-agent large language models that can autonomously collaborate to diagnose and treat virtual patients in real time. Trained on extensive, field-specific datasets, they appear able to mimic a wide range of medical scenarios and continuously refine diagnostic accuracy through feedback mechanisms.”
Your email address will not be published. Required fields are markedmarked