When AI Produces More Information Than a Team Can Use

I use AI every day. It helps me organize information, explore a problem, draft ideas, and spot patterns I might want to examine more closely. But I’ve learned that giving a team more information is not the same as helping it do better work.

I came to operations through sleep medicine. Over the years, I’ve worked across clinical care, technology, teams, and revenue. Much of that work has involved building systems around people who already had a lot to manage. The technology mattered, but the work only moved when people knew what to do, who owned the next step, and how they would know whether it was working.

That’s why I’m cautious when a company treats AI as the solution before it understands the operating process around it.

Information needs a path into the work

In a healthcare engagement I’ve been supporting, AI-assisted runbooks and reports could provide a great deal of information about the organization’s infrastructure. The challenge wasn’t simply getting more detail. The team needed a clear way to interpret that information, decide what mattered, and turn it into action.

A runbook can describe a system. A report can surface a pattern. Neither one, by itself, tells a team who should review it, what decision needs to be made, when to escalate a problem, or how to follow through. Without those operating steps, the information can pile up while people are left wondering what they’re supposed to do with it.

That distinction has come up in other work, too. In The Problem You See Isn’t Always the Problem, I wrote about starting with the issue underneath the one that first gets reported. AI can help surface possibilities, but someone still has to understand the work well enough to ask the right question and test what the output means.

Experience and AI should work together

I also see people hesitate to use AI because they believe their own knowledge and experience should be enough. I understand that instinct. Experience matters, especially when a decision affects patients, employees, customers, or a business’s ability to operate. AI can be wrong, and a polished answer is not proof that the answer fits the situation.

But I don’t think protecting experience means avoiding the tools altogether. I think experienced people can use AI to examine information, consider options, and reduce repetitive work while still applying their judgment. As AI becomes part of how teams prepare, analyze, and document work, refusing to learn where it can help may leave experienced people with fewer ways to contribute.

I’ve seen the same tension in CRM work. People often dislike entering information because the system asks them to document activity without making the next step easier. In The CRM Problem Isn’t the CRM, I describe why adoption depends on how the system fits into the work. AI doesn’t remove that requirement. If anything, it makes the process around the system more important.

Start with the work, then choose the tool

When I look at an AI opportunity, I start with the work that needs to happen. What decision is the team trying to make? What information does it need? Who is responsible for reviewing the output? What happens when the answer is unclear or something falls outside the usual process? How will the team know the change is helping?

Once those questions have answers, AI has a clearer role. It might help summarize information, draft documentation, identify a trend, or prepare a person to make a decision. The team still needs a process for checking the work, acting on it, and learning from the result.

AI can make a capable process faster and easier to use. It can also produce more reports, recommendations, and runbooks than a team knows how to manage. The difference is the operating structure around it: clear responsibilities, useful workflows, sound judgment, and a way to follow through.

I don’t start by asking which AI tool a company should buy. I start by asking how the work actually happens and what needs to change. Once that is clear, it’s much easier to decide where AI belongs and whether it is helping.