I am delighted to have my paper which systematically reviewed the literature on perspectives of patients and clinicians on using AI-enabled decision aids published at Nature Digitial Medicine. Many thanks to the amazing team. doi.org/10.1038/s41746…@sarahpslight
Incorporating #AI education into medical training is essential for equipping future healthcare professionals.
This study developed an expert-approved AI curriculum for Canadian medical students, emphasizing ethics, law, application, and collaboration, with 82 essential… pic.twitter.com/GW3yV1H5dr
Crazy interesting paper in many ways:
1) Voice-enabled GPT-4o conducted 2 hour
interviews of 1,052 people
2) GPT-4o agents were given the transcripts & prompted to simulate the people
3) The agents were given surveys & tasks. They achieved 85% accuracy in simulating interviewees pic.twitter.com/yqS9K4XW6l
Perspective by Paul J. Barr, PhD (@BarrPaulJ), Robert Gramling, MD, DSc, and Soroush Vosoughi, PhD: Preparing for the Widespread Adoption of Clinic Visit Recording nejm.ai/3NDcYSo
There is a lot of energy going into fine-tuning models, but specialized medical AI models lost to their general versions 38% of the time, only won 12%.
Before spending millions on specialized training, might be worth exploring what base models can do with well-designed prompts. pic.twitter.com/6FjpfBfahf
Max Tegmark shows a deepfake of Pope Francis advocating for unregulated AI development in order to build a digital God: “God is great, so two Gods will be even greater” pic.twitter.com/yWYw9aVtd7