Microsoft says AI system better than doctors at diagnosing complex health conditions | Artificial intelligence (AI)

Microsoft has revealed details of the artificial intelligence system that performs better than human doctors in complex health diagnoses, creating a “path to medical cancellation”.

The company’s artificial intelligence unit, which it leads Britain Pioneer Pioneer Mustafa SuleimanHe developed a system that imitates a committee of expert doctors who take cases “in diagnostic and intellectually demanding.”

Microsoft said that when you are associated with the advanced O3 Ai model, his approach is “solving” more than eight of 10 case studies chosen specifically to challenge the diagnosis, “Microsoft said. When these status studies were tried on practicing doctors – who were unable to reach colleagues, textbooks or chat – the accuracy rate was out of 10.

Microsoft said it’s also a cheaper option than using human doctors because it was more efficient in demand tests.

Although the potential cost savings of its research are highlighted, Microsoft has reduced the effects of the job, saying that it believes that artificial intelligence will complete the roles of doctors instead of replacing them.

“Their clinical roles are much broader than just a diagnosis. They need to move in mystery and build confidence with patients and their families in a way that has not been prepared artificial intelligence to do,” the company wrote in a blog that announces the research, which is presented to review the peer.

However, the use of the “Path to the Supertelligence Medical” logo raises the possibility of radical change in the health care market. While artificial general intelligence (AGI) refers to systems that correspond to human cognitive capabilities at any specific task, excellent cancellation is an equal term that refers to a system that transcends human intellectual performance in all fields.

In explaining the logical basis behind the research, Microsoft has raised doubts about the ability of artificial intelligence to be exceptionally good in the US medical licensing exam, a major test for a medical license in the United States. She said that multiple selection tests prefer to save answers about the deep understanding of the topic, which can help “exaggerate” the efficiency of the artificial intelligence model.

Microsoft said it is developing a system, such as the real world doctor, takes step-by-step measures-like asking specific questions and requesting diagnostic tests-to reach a final diagnosis. For example, a patient who suffers from coughing and fever may require blood and chest tests before the doctor reaches the diagnosis of pneumonia.

The new Microsoft approach uses complex case studies from the New England magazine (NEJM).

Solomon’s team turned more than 300 of these studies into “interactive cases challenges” that he used to test its approach. Microsoft’s approach to the current artificial intelligence models use, including those produced by the Chatgpt, Openai, Mark Zuccerberg’s Meta, Anthropic, Elon Musk’s Grok and Google’s Gemini.

Then Microsoft used the AI ​​system similar to factors called “Orchestrator Diagnosis” to work with a specific model on the tests that can be requested and what may be the diagnosis. Orchestrator actually imitates a doctors, which then comes with diagnosis.

Microsoft said that when the O3 O3 APENAI advanced O3 model, Microsoft said it has “solved” more than eight of 10 NEJM case studies – compared to a success rate out of 10 for human doctors, “Microsoft said.

Microsoft said her approach was able to use the “breadth and depth of experience”, which exceeded individual doctors because it may extend from multiple medical specialties.

He added: “Expanding the scope of this level of thinking-and what is behind it-has the ability to reshape health care. Amnesty International can enable patients to manage the routine aspects of care and prepare doctors with advanced support for the decision for complex cases.”

Microsoft admitted that her work is not ready for clinical use. More tests are needed on “Orchestrator” to assess their performance on more common symptoms, for example.