Can AI create a new level of human-centered health?

The heat coming off the AI rocket streaking over the health care landscape is stoking a fundamental fear on the ground: inviting more technology into the patient experience will further dehumanize care. The concern is understandable. Nothing is more personal than health, and for health and wellbeing programs, nothing motivates behavior change like personal engagement. But there’s a paradox – actually two – riding inside the rocket’s skin. To begin with, AI has the potential to make individuals’ health journeys more personalized, rather than less.

Secondly, we’ve been tapping the roots of AI to solve patient problems for a long time. In clinical areas like cancer care, AI is informing providers and their patients to make more personalized therapeutic recommendations. In my area of health – improving physical and mental wellbeing through holistic health management and lifestyle changes – how the technology will be used will determine the answer to a key question: Can we design AI data and insights to be human-centered, so they better motivate and equip people to take greater agency over their health/? If AI can do for health and wellbeing programs what it is beginning to do for clinical intervention, it can expand personal engagement in preventative behaviors, driving more positive outcomes.

When I was in clinical training in the 1990s, algorithms were already coming into place with the emergence of “care pathways,” growing limbs on the decision tree that were supporting clinical actions. A decade later, I worked for a company that developed algorithms for clinical medicine, applying them to patient pharmacy and lab data to locate gaps in care. This was under the ungainly nomenclature, “clinical decision support algorithm,” which doesn’t have quite the same ring to it as “AI,” but it was heading in the same direction. So, from this clinical perspective, what is it about the technology’s current advancements that are exciting for holistic wellbeing? The promise is in the breadcrumbs.

As with the earlier algorithmic assists, AI is good – now, astonishingly good – at quickly recognizing patterns of activity buried in data. Unlike in a strictly clinical investigation, though, wellbeing data floats in a much larger pool. It’s estimated that non-clinical factors – lifestyle choices and the Social Determinants of Health – affect upwards of 70% of an individual’s health outcomes. That data doesn’t appear on a lab test, so we’re reliant on other methods of capture and analysis. In annual health risk assessments, individuals begin to provide the breadcrumbs that AI can aggregate, analyze, and personalize as actionable recommendations. By beginning their journey with a plan more attuned to who they are and what they need, individuals will have a greater chance of reaching their destination. But what happens over the rest of the journey.

AI isn’t a yearly one-and-done; it’s a living companion, or, more specifically, a co-pilot. For example, in a live, ongoing dialogue between a member and coach, AI can be leveraged to aid in prompts and explore responses, subsequently seeing and analyzing the scattered breadcrumbs. It can provide insights to the human coach that connect the totality of the person’s experience – the clinical and non-clinical – to behavior modification that will produce better outcomes. Producing better outcomes motivates individuals and can increase engagement – the biggest hurdle for employers investing in health and wellbeing programs seeking the corresponding gains in productivity, retention, and workforce satisfaction. With so much at stake, how should employers be thinking about their AI choices?

Human-centered design should be the guiding principle for employers and any organization dedicated to supporting their members’ health goals. Employers should address three questions that can yield the most direct improvement in engagement:

1. How can I leverage human-centered design to make health information more digestible and actionable for individuals?
2. How can the technology be applied to maximize benefits balancing, ensuring offerings meet the needs of employees’ various generations and life stages?
3. How can AI be used to help members determine what benefits they have access to that will support their individual goals?

Related: AI and employee benefits: Accelerating empathy in the workplace

Employers should use AI to help simplify members’ paths to improving their health. When an employee determines, “I want to sleep better” and is guided to reach out to a coach who is supported with an AI understanding of the member’s breadcrumbs, the path to behavior change can become individualized rather than generalized. As an extension of human learning, AI can account for the nuances of the employee’s non-clinical factors and provide data-driven insights based on their specific sleep experience and goals. In this way, technology can work to make the pursuit of wellbeing more human.

AI’s rapid advances puts health and navigation companies, and the health care industry as a whole, at a pivotal moment. We do face a choice, and I’m reminded of the tech axiom known as Kranzberg’s Law: “Technology is neither good nor bad: nor is it neutral.” AI and its large language models are instruments of the people who create them and those who implement them. There is a real opportunity before us to put these powerful tools in service of personalized health, producing insights that will enable better understanding of what motivates an individual’s behavior, what their next best step is, and what is available to them to reach healthier outcomes.

As is commonly emphasized, the new applications of AI have to be clinically accurate, time efficient, and cost effective. Those are the structural imperatives. But to rise to the level of value that can transform an overly complex health care system and create a deeper level of personalization, human-centered design must be the guide.

Jeff Jacques, MD, is a physician executive leader with more than 20 years of expertise in digital health, building products and services that deliver personalized support for individuals with complex care journeys. As the Chief Medical Officer of Personify Health, Dr. Jacques ensures alignment with member and market needs, leading the organization’s behavior science approach and working to simplify and support members’ health journeys.