What the Failure of Wearables Can Teach Us About AI
A practitioner perspective by David Kirk, MD, Chief Medical Officer, Regard.
The argument
Wearables failed to move outcomes at scale because they collect data without clinical context - signal without synthesis. Healthcare AI must avoid the same trap by synthesizing longitudinal data inside clinical workflows instead of generating more raw data points.
Why we picked it
The HTC editorial desk curates practitioner perspectives that pass a simple test: would a health system executive, product leader, or investor act differently after reading it? This piece by David Kirk earns its place - it comes from the operating trenches of wearables & rpm, not the conference stage, and it takes a position specific enough to disagree with.
The bigger picture
The AI adoption question runs through every executive conversation in 2026. The evidence base is finally arriving: the largest controlled study of ambient documentation, published in JAMA, found clinicians saved 16 minutes of documentation time per day - but only a third of users hit the usage threshold for maximum benefit (our coverage). Meanwhile Rock Health has stopped labeling AI as a differentiator at all, because every company claims it - which is exactly why operating discipline of the kind described here separates winners from demo-ware.
Read it alongside our own reporting in Wearables & RPM, where the data behind this debate is covered daily on the HTC Wire.