Episode Description:
Once dismissed as Silicon Valley provocation, the question of 'will AI replace doctors', now feels a very real possibility. As large language models pass licensing exams and multimodal systems interpret scans at scale, the panel examines which parts of medicine are most exposed to automation and which remain deeply human.
Diagnostic specialties may shift first. Clinical knowledge can be scaled. But accountability, judgment under uncertainty, communication of bad news, and ethical trade offs resist simple optimisation. So what does the future physician role look like?
Key Topics:
- Vinod Khosla's prediction that AI would eliminate doctors
- AGI timelines and multimodal models
- Radiology, pathology, and diagnostic automation
- Why no doctor will practice without AI
- Psychiatry and the limits of avatar therapy
- High-stakes procedural care (intubation, intensive care)
- Accountability and liability barriers
- The "reward function" problem in healthcare AI
- Empathy vs optimisation
- Judgment, truth-telling, and delivering bad news