Recent product announcements from major wearable companies showcased the polarization of views on the direction of well tech. During the past two weeks, Google has made significant moves to deepen its health AI ambitions, and Whoop has doubled down on positioning its device as a clinical-grade monitoring tool. Both companies represent serious, well-resourced bets on the future of consumer health. And watching these developments closely, I’ve found myself thinking about the strategic dichotomy they collectively reveal. Two camps appear to be forming in the fight to own the consumer health intelligence layer.

In one corner: AI-first platforms that promise always-on, personalized guidance at infinite scale. In the other: clinician-mediated models that drive users back toward face-to-face care, positioning the technology as a sophisticated referral mechanism, personalized to an individual’s needs. Both approaches have genuine merit. And yet both, taken alone, leave something important on the table.

The problem with AI-only

Let’s be honest about the limitations of deploying AI as a standalone health advisor. Large language models are genuinely impressive and genuinely fallible. In health contexts, hallucination isn’t an amusing quirk; it’s a patient safety issue. When an AI confidently recommends a supplement interaction, misinterprets a biomarker trend, or provides guidance that isn’t grounded in a person’s actual clinical picture, the margin for error approaches zero.

There’s also the lived-experience gap. AI systems are trained on data, not on the thousands of patient interactions that give a clinician the pattern recognition to know when something feels off, when the numbers say one thing but the person in front of them says another. That intuition matters, and it’s very hard to encode.

The problem with clinician-only

The clinician model has the opposite constraints. Clinicians are among our most valuable and most finite resources, even in a virtual context. In the U.S. alone, primary care physician shortages are projected to worsen significantly over the coming decade. The economics of scaling individualized health guidance through human providers don’t work, not at the volume the market demands, and not at a price point most consumers can sustain.

“Go see or talk to your doctor” is sometimes the right answer, but it’s not a scalable health intelligence strategy, especially when minor health abnormalities surface.

The question nobody’s asking

Here’s what I think gets missed in the framing of AI versus clinicians: The real opportunity isn’t to pick a winner between these two models. It’s to build infrastructure that lets wellness businesses deploy both, intelligently, at the right moments in a consumer’s health journey. That’s the thinking behind what InsideTracker has been building.

A third path: the health intelligence platform

InsideTracker recently unveiled Terra, a neurosymbolic AI platform designed specifically for wellness companies, including fitness OEMs and facilities, nutrition providers, corporate wellness programs and others. The premise is straightforward, even if the execution isn’t: Give businesses the tools to deliver hyper-personalized, biologically grounded health guidance at scale, without forcing a choice between chatbot convenience and clinical credibility.

Terra combines multimodal biologic data, including blood biomarkers, physiological signals, and genetic data, with over a decade of platform development and optimizing deterministic logic, more than 7,000 clinical studies in its evidence base, and over 10 billion data points refined across a real population. That’s not marketing language; it’s the infrastructure required to make health recommendations that are actually individualized rather than merely personalized-feeling.

The key design principle is flexibility. Wellness businesses have different needs, different customer relationships, and different clinical risk tolerances. Some have patients who are self-motivated and comfortable self-managing with AI-powered nudges alone, and others who need more hand-holding and a human connection, with a registered dietician for example, to stay engaged and motivated. Terra is modular precisely because there is no single right answer to how AI and human expertise should be weighted. That depends on the use case, the population, and the regulatory environment.

What “hyper-personalized” actually means

I want to spend a moment on this phrase because it gets overused to the point of meaninglessness. True hyper-personalization in health isn’t knowing that someone is a 42-year-old woman who exercises three times a week. It’s understanding that her ferritin levels have been trending downward for six months, that her sleep data suggests cortisol dysregulation, and that her dietary logs show a pattern inconsistent with her stated goals, and then generating guidance that responds to the intersection of all of those signals simultaneously, calibrated against population-level evidence.

That’s what biologic data makes possible. And that’s what Terra can do consistently at scale for wellness brands who want to deliver something meaningfully better than a generic app.

The competitive moment

For wellness businesses, the next 18 to 24 months will be defining. Consumer expectations around health guidance are being shaped right now by the products Google, Whoop, and others are putting into the market. The bar for what “personalized health” means is rising rapidly.

Companies that partner with the right health intelligence infrastructure now will be positioned to meet those expectations. Those that wait, or that try to build this capability from scratch, will find themselves competing against platforms that had a decade’s head start on the data and the science.

The consumer doesn’t have to choose between a chatbot and a clinician. But wellness businesses do have to choose their infrastructure partners. That decision is arriving faster than most people realize.

Jon Michaeli is a health care executive.