Indeed, Novid did have promise. Po's work should be celebrated.

The problem is that Novid depends on self-report. Self-report is not reliable, or else we'd have seen success around this kind of self-reported contact tracing, which does not model reality accurately at scale.

The problem is that a surveillance network is perfect for this kind of healthcare application, where the balance of power is toward the invisible surveiller and does not depend on the surveilled being compliant.