GOOGLE DEEPMIND AI OUTPERFORMS GPT-5.4 IN MEDICAL TESTS

The study involved simulated patient scenarios where medical professionals compared the AI systems' diagnostic and treatment recommendations without knowing which model produced each response. DeepMind's system demonstrated superior performance across multiple metrics, indicating advances in AI's ability to handle complex medical reasoning. However, the gap between AI and expert physicians remained significant, highlighting current limitations in clinical AI applications. The research also revealed critical gaps in conversational AI for healthcare. ChatGPT's voice mode, which OpenAI has promoted for accessibility, showed inadequate performance for medical consultations, suggesting consumer-grade AI tools require substantial refinement before deployment in healthcare settings. DeepMind's co-clinician remains in simulation-based testing and has not been tested in real clinical environments. The findings underscore both the potential and the distance AI must travel before serving as a primary decision-making tool in medicine.