
What's Your Problem? Predicting Human Health with AI
6 snips
Jul 6, 2023 Unlearn AI’s Charles Fisher is building models that simulate patients’ future health, from Alzheimer’s to ALS. Hear how “digital twins” could help clinical trials recruit fewer people, and why turning those tools into personalized predictions for doctors is a much bigger leap. He also imagines one model spanning diseases, plus shares a lesson from refereeing kids’ hockey.
Chapters
Transcript
Episode notes
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Introduction
00:00 • 4min
Machine Learning and Generative AI in Medicine
03:32 • 2min
The Importance of Simulating Health Outcomes
05:10 • 2min
How to Assess Causality in Clinical Trials
07:40 • 2min
How to Build a Model for Alzheimer's Disease
10:05 • 2min
How to Predict What Will Happen Next in a Clinical Trial
12:07 • 3min
The Relationship Between Predicted Outcomes and Observed Efficacy of Placebo Treatment
14:56 • 3min
The Effect of 25% Less Patients in a Phase Three Trial
17:33 • 2min
The Importance of Understanding the Health Outcomes of Different Treatment Options
19:28 • 2min
The Importance of Confidence in Health Models
21:20 • 2min
The Future of AI in Medicine
23:27 • 2min
How to Build a Language Model That Understands How Language Works
25:26 • 2min
How to Map Health Data to the Same Representation
27:23 • 2min
The Future of Machine Learning
29:06 • 3min
The Evolution of AI
32:17 • 3min
Lucky Land Slots - Get Lucky Just About Anywhere
35:11 • 2min
