> we should have just gotten a dozen YOLO engineers and statisticians to run any permutation of dosage, duration, and vaccine candidate; much like how software

> we should have just gotten a dozen YOLO engineers and statisticians to run any permutation of dosage, duration, and vaccine candidate; much like how software companies run thousands of A/B tests at scale. I, for one, would try to run clinical trials the total opposite from software companies. "Fail fast, fail often" nor "Iterate fast, release often" really work for healthcare.