Lately, I’ve been reflecting more deeply on the race to adopt artificial intelligence, especially in complex environments like healthcare. Here’s a story. When my father nearly died twice last year, I spent weeks in hospitals (actually, months). And I saw it firsthand:

• Fewer workers doing more tasks
• Tons of systems, but critical details are still lost in translation
• Burned-out staff are trying to do the impossible with limited time 

My father had lost his ability to speak, so every misstep in his clinical record had consequences. One note suggested he didn’t consent to a treatment plan, and suddenly the default course became: “send him home and hope for the best.” It took our family’s insights (about him), time, patience, and advocacy to sit with him, help him understand his options, and get his written consent. Then we had to align every provider, across every shift, to reorient the plan of care. What we experienced wasn’t necessarily neglect. It was the result of a system under pressure: fewer people expected to do more, with fewer resources, in less time. 

When time is lost, so is context. And without context, both patients and clinicians suffer. 

Artificial vs. Enterprise Augmented Intelligence 

In my opinion, the real innovation isn’t about AI. It’s about our ability to identify and respond to the assumptions baked into our data. Who and how do we determine what is true, and when it is true? 

I believe that Enterprise Augmented Intelligence, not artificial intelligence, is the future of healthcare innovation. It’s critical that leaders take a systems thinking approach, one that considers the big picture, accounts for the impact on both the workforce and patients, and integrates the human insights of those doing the work to understand the context behind the data in each environment.

So Much Data, So Little Clarity

Even in data-rich environments like hospitals, it’s still possible to miss the full picture. There are systems everywhere: monitors, dashboards, EMRs, but very few are designed to give staff more time with patients. Observations and face-to-face moments are where some of the most important insights live. Insights that no AI can glean. 

Today, the vast majority of enterprise data is unstructured, including notes, audio, text, and various forms of metadata. Yet many systems still rely on structured data, which is pre-labeled, flattened, and filtered. In doing so, we risk making confident decisions based on incomplete stories. The future of intelligent automation depends on our ability to amplify and improve the data we already have, not on collecting more and hoping for the best.

What’s Next

The next leap forward will belong to solutions that can:

1. Adapt quickly identify and act on small shifts that create meaningful value.

2. Integrate seamlessly with other enterprise systems, rather than operating as walled gardens.

3. Honor & respect the source of truth.

4. Illuminate & use smart automation and other strategies to reveal and improve context, not just compress it.

And do all of this without placing additional burden on people who are simply trying to survive or do great work.