AI in healthtech needs to solve the interpretation problem, not just the data problem

For years, one of the longstanding challenges in health technology was access to data.

We needed better ways to measure physical activity, sleep, heart rate and other indicators of health. Wearables have made enormous progress in solving that problem. Today, people can generate an extraordinary amount of information about their bodies everyday.

But as health technology becomes better at collecting data, a new problem is emerging.

We have more health data than we know what to do with.

A wearable can tell you that your sleep was shorter than usual, your heart rate changed overnight or your recovery is different from your usual pattern. But what does that actually mean? Should you change your behaviour because of it?

The next frontier of AI in healthtech should not simply be collecting more data or producing more sophisticated dashboards. It should be helping people interpret the information they already have.

From making a nation active to understanding personal health

Singapore offers an interesting example of what technology can achieve when it is designed around behaviour change. 

Through initiatives such as the National Steps Challenge, supported by the Healthy 365 platform, technology has helped make healthier behaviours more visible and accessible at scale. Since 2015, Actxa has supported this journey as one of its wearable technology suppliers for the National Steps Challenge. 

But the further we move into personalised health, the more complicated the picture becomes.

Health does not happen in isolated categories. Sleep affects recovery, exercise affects recovery and stress can influence sleep and activity. What we eat and how we move can influence our metabolic health. Looking at each metric independently can leave consumers with a collection of numbers without a clear understanding of how they relate to one another.

This is where I believe the next opportunity for AI lies. Technology does not change behaviour simply because it collects data. It becomes useful when the information is translated into something people understand and have a reason to act on.

As wearable technology evolves, we are moving beyond steps and activity into sleep, recovery, heart rate variability, stress and metabolic health. The challenge is no longer simply getting people to collect health data but helping them understand what those different signals mean together.

The journey from making a nation more active to helping individuals better understand their health is the next step in digital health.

From measurement to meaning

The first generation of consumer wearables largely focused on measurement.

How many steps did you take? How long did you sleep? What was your heart rate? How active were you today?

These metrics have helped make health more visible and encouraged people to become more aware of their habits. But individual measurements rarely tell the whole story.

Sleep, exercise, stress and metabolic health are interconnected. A demanding workout can affect recovery. Poor sleep can influence how we feel and perform the following day. Stress can affect both sleep and activity. Everyday choices can have consequences across multiple aspects of our health.

Looking at each metric in isolation can therefore provide an incomplete picture.

This is where AI has an opportunity to make a meaningful difference.

Instead of simply reporting that a number has changed, AI can potentially help identify patterns across different signals, put them into context and highlight changes that may be more meaningful to the individual.

The distinction is important: Data tells us what happened; interpretation helps us understand what it might mean.

The most useful benchmark may be your own baseline

There is also an opportunity to move beyond a one-size-fits-all approach to personal health.

General health guidelines and population-level benchmarks will always have an important role to play. But individuals have different routines, bodies and starting points.

For consumer health technology, understanding what is normal for an individual can therefore be just as useful as comparing them against a generic benchmark.

What is normal for me?

What has changed?

Has this happened before?

Could several seemingly unrelated changes actually be connected?

These are questions that become increasingly difficult to answer as the amount of available health data grows.

But interpretation also comes with responsibility. Health is not an area where AI should be rewarded simply for sounding confident. Consumer health technologies need to be clear about what an insight can and cannot tell someone, and about the distinction between wellness guidance and medical diagnosis. The goal should be to provide useful context without creating false certainty or unnecessary anxiety.

This shift from measurement towards interpretation is also shaping the next chapter of Actxa.

We are taking our experience in population health into a more direct-to-consumer model, where the relationship with health technology becomes more personal. Our upcoming Spark Band is one example of this shift, bringing together activity, metabolic health and other everyday health signals.

The important question, however, is not how many signals a device can capture. It is whether those signals can help someone make sense of their own health.

The objective is not simply to add another set of metrics to an already crowded dashboard but to make personal health information easier to understand in the context of an individual's everyday patterns.

That distinction will become increasingly important as wearable technology becomes more sophisticated.

The future is not another dashboard

The healthtech industry has spent years asking how much more we can measure. The more important question now is how much better we can help people understand what those measurements mean.

The most useful health technology of the future may not be the device with the greatest number of sensors or the longest list of metrics.

It may be the one that connects the signals, understands the individual's baseline and presents the information in a way that is genuinely useful.

For AI to have a meaningful role in healthtech, it needs to move beyond simply recognising patterns. It needs to help translate those patterns into information people can understand, while being responsible about the limits of what that information can tell us.

Ultimately, AI's biggest opportunity in healthtech is not to give people more data. It is to turn measurement into understanding, and understanding into action.

That means helping people answer three simple questions:

What is happening?

What might it mean?

What can I do next?

If we can get those three things right, AI could move consumer health technology from simply tracking our bodies to helping us understand them.

And that is where the real value of health AI begins.