More Americans than ever before are tracking their health using wearable devices like fitness bands and smartwatches – a movement dubbed the “quantified self.” This generates a lot of personal health data, but will all this data help improve the health of our country? A recent article and interview from mHealthNews help answer that question with a resounding… “maybe?”

The main problem is that wearable users tend to be young, upper-middle class and fit – and these are not the patients we need to monitor. Just as Apple launched its Watch with testimonials from a marathon runner, most of the avid quantified selves out there are athletes, joggers, or chronic dieters. And studies have shown that pedometers aren’t helping get people disinclined towards exercise to start moving.

The patients that most need monitoring, as Sentrian CEO Dean Sawyer wrote in a recent article, are the “millions of chronically ill whose care and treatment accounts for some 70 percent of U.S. healthcare spending, or more than $2 trillion a year.”

On top of the fact that we’re collecting data about the wrong demographic, the data being collected (heart rate, steps walked) is too narrow to understand population health. Oh, and there’s also the problem that physicians don’t trust data from unregulated sources and that they don’t have established methods for incorporating it into patient care.

These barriers are the reason Jack Kriendler (an MD and entrepreneur who co-founded Sentrian with Sawyer) recently concluded: “I actually think that “quantified self” has failed to make an impact in medicine.”

But the gloomy tone of these experts’ outlook on quantified self is a bit misleading, since they both also seem to have hope for the relevance of wearable health monitors.  As Sawyer’s article explains, this lies in a transition from the “quantified self” to a “collectivized us.” In other words making the personal health data movement relevant means shifting focus from just understanding our own daily habits to large-scale information gathering and population health management using data analytics.

While he doesn’t discuss specific technology or initiatives, it turns out that the shift from ‘me’ to ‘us’ is more about machine learning than a utopian vision of caring for others. The idea is that individual data logs – including caregiver observations and patient feedback – are already being compiled – to fuel population-scale machine learning that will notify a clinical team when a patient requires attention. So while doctors might not know what to do with patient-generated health data, computers do (IBM’s Watson seems like a good candidate).

This type of application would require a shift in who uses wearables, exactly what they monitor, and how we handle data security – but at least the quantified self has helped to make them cheap. By adopting activity monitors for personal use, first adopters have helped dramatically reduced costs for hardware and software. As Kriendler puts it, “Activity monitors, when I used to use them on Formula One drivers, cost tens of thousands of dollars. Now they cost tens of dollars and they’re almost as good as the ones we stick on athletes. It’s quite extraordinary really.”

For wearables to really deliver on their promise, Kriendler says they will need to prove that they have real value for physicians, providers, and patients – something they have so far failed to do. While he’s working on it, it’s not clear from the interview if Kriendler thinks the quantified self will morph into something greater. Sawyer’s article, on the other hand, concludes with an optimistic conviction: “The revolution in health has begun. It’s no longer about you, me, her and him. It’s about us.”