AI in medical devices (which often lack rigorous standards for FDA approval) are leading to catastrophic complications in the OR.

Listen to FULL RUNDOWN at https://podcasts.apple.com/us/podcast/full-rundown-the-%24730-billion-reality-check-openais/id1684415169?i=1000752770917

Summary: A high-stakes study in Nature Medicine reveals that OpenAI’s new ChatGPT Health tool failed to recommend the ER in more than half of emergency cases. From missing the signs of respiratory failure to "deactivating" suicide prevention banners based on lab results, the study highlights a massive safety gap in consumer AI health tools. ...