Diagnostic Orchestrator surpasses doctors in accuracy while lowering testing expenses.

Microsoft’s AI Diagnostic Orchestrator, built on OpenAI’s o3 model, correctly solved 85.5% of 304 challenging NEJM case studies versus 20% accuracy by physicians. The system reduced overall diagnostic costs compared to both doctors and standalone AI models by iteratively generating hypotheses and ordering targeted tests. Submitted for peer review, Microsoft positions it as a tool to enhance clinical decision-making alongside, rather than replace, medical professionals.

Read more:

https://cybernews.com/news/microsoft-ai-health-superintelligence/