Practice leaders are motivated to use artificial intelligence (AI) to automate and improve efficiency. Labor costs are rising, reimbursement is tightening, and practices are under pressure to do more with less. As leaders, we have a financial responsibility to aggressively automate repetitive, logic-based tasks wherever it makes operational sense. But if we define AI's return on investment (ROI) only by the reduction of resources and human capital, we may overlook a more impactful strategy. The greater opportunity lies in using AI to improve our human potential by elevating decision quality, exposing assumptions, reducing unnecessary cognitive load, and creating more space for work that requires human judgment and connection.
I strongly believe in using AI to create efficiency and reduce costs. But I’ve also found that the most valuable effect AI is having on me personally is making me a more impactful thinker. Artificial intelligence can challenge my assumptions, pressure-test high-stakes decisions, expose gaps in my reasoning, and help me spot problems from angles I might otherwise miss. I’ve found the highest return on my use of AI isn't time saved; it’s discovering stronger strategies and new ways to work.
We have seen a version of this before. Practices that approached electronic health records as electronic versions of paper charts often reproduced old inefficiencies in a new system. The practices that gained the most value redesigned the work around what the technology made possible. Artificial intelligence requires the same discipline: Rather than simply adding a tool to how we already work, it presents an opportunity to reconsider the work itself.
Improve the Decision, Not Just the Output
I have learned to leverage AI to challenge rather than validate. One example is when I recently worked with AI to redesign an employee's incentive structure. I modeled different strategies, pressed for feedback on the potential influence on behavior, and pressure-tested how the framework could create unintended consequences. Working with AI helped me uncover a simpler structure that moved the reward closer to the behavior we wanted. The new model made the outcome visible and tangible, making the incentive easier to understand and more motivating.
Once we defined the structure, I modeled aggressive, moderate, and modest-success scenarios so the employee could see the potential financial impact over time. That led to a deeper conversation about the specific levers within the employee’s control and how those behaviors connected to the practice's broader goals. The ROI of this exercise came from our interaction and her understanding of how she could become more influential, impactful, and rewarded in our organization.
The greater purpose of leveraging AI is to help us ask better questions before committing the organization to a decision when judgment carries the greatest consequence. For practice leaders and physician owners, the cost of a poor decision can dwarf the value of the administrative time automation saves.
To apply this concept, I start by giving AI the problem, context, and constraints. Then I ask it to interview me, one question at a time, to help me clarify the issue. I ask it to challenge my assumptions, identify potential outcomes, and develop counterarguments. Artificial intelligence can easily reinforce embedded assumptions in our questions, and it is tempting to buy in. The opportunity is to create friction by asking AI for critique, pushback, and alternate points of view.
Move Repeatable Thinking Into the Process
The same discipline can reveal weaknesses in our systems. I experienced this while examining our revenue cycle. Every practice needs defined controls from appointment scheduling through claims submission and payment posting, with clear ownership at each step. I believed my team and I understood those controls. When I used AI to document our workflow and challenge each handoff, I realized that too much of our revenue cycle depended on individual knowledge rather than documented standard operating procedures (SOPs).
What began as a mapping exercise exposed vulnerabilities in our practice—thin areas in our SOPs, where ownership was assumed rather than explicit. As a result, our team better understood our business, and we created a comprehensive, living training resource that has improved the accountability of everyone involved.
A second type of system opportunity is complex work where human involvement doesn’t improve the output. Our cataract surgery quoting process is a good example. Producing an accurate quote required understanding insurance responsibility, cataract package fees, facility and practice fee allocation, and how those calculations changed depending on where the procedure was performed. That created significant cognitive load on our patient concierge team during important, front-facing financial conversations.
Rather than defaulting to more training, I worked with AI to design a custom-coded quoting tool with logic that calculates patient responsibility based on the team’s selections. It reduced complexity, simplified the workflow, and increased our patient concierge team’s confidence in educating the patient. This allowed our team to shift their attention from creating the quote to building the connection and trust that patients need to make a confident surgical decision.
Automate the Right Thinking to Develop Better Leaders
At See Vision, we live by a simple principle: Your brain is too valuable to spend your best energy on work that should be built into a process. On the other hand, leaders should be equally careful not to over-automate and sacrifice the experiences and judgment people need to grow. The leadership challenge is deciding which thinking should be systematized vs what should remain with a person. Artificial intelligence does not develop leaders; their growth comes from how we teach them to think about their work, where they can create the highest value, and how to use AI to improve it.
We reinforce this mindset through our leadership meetings, where we workshop these ideas. Our team presents problems they’re trying to solve, along with how they are using AI to test assumptions, explore alternatives, and challenge their critical thinking. Our goal is for our leaders to leverage AI to become more disciplined in how they think about work by asking stronger questions and improving their decision-making.
Our team understands that the simplest phase of AI adoption is to remove people from processes. However, our culture is built on the belief that health care is a relationship business. The second, potentially more impactful, phase is applying critical thinking to enhance patient relationships and build trust by reducing friction and protecting the human elements of our interactions.
Innovation Without Discipline Is Risk
An AI-forward practice requires compliance guardrails. The deeper AI becomes embedded in our cultures, the more important governance becomes. Practices should seek technical guidance to understand how their workspace is configured, whether tools are enterprise-grade and covered by appropriate business associate agreements, where data is stored, who has access, and how long it is retained. Work with legal counsel to develop an internal AI policy that creates clear boundaries around protected health information and defines approved tools, as well as safe and unsafe uses. If you haven’t created a policy and evangelized it, assume that your staff is likely already experimenting with free tools on personal devices and possibly on the company network.
Responsible AI use also requires a strong trust-but-verify mindset. Artificial intelligence can be confidently wrong. It does not replace knowledge, clinical judgment, accountability, or human decision-making. Teams need to frame questions carefully, verify “factual” claims, challenge outputs, and ask for pushback. The human user remains the editor in chief. Training cannot be one-and-done because the capabilities, risks, and use cases continue to evolve.
The ROI We Choose
The most obvious way to use AI is to save time, energy, and resources. The greater return lies in what leaders choose to do with the capacity they create, and how we develop and strengthen our teams’ leaders. We can use AI simply to do the same work with fewer people and resources, or we can reinvest those resources in better decisions, stronger leaders, deeper expertise, and more meaningful patient conversations.
The greater ROI of AI lies with our people. Efficiency matters, and we should eliminate waste wherever possible. But the return is amplified when technology helps us move repetitive work into better systems, improve our critical thinking, and preserve humanity for the moments when judgment and connection matter most.
The early question of “What work can AI eliminate?” should now shift to “How can AI help us develop stronger, more capable, and impactful teams and organizations?” If we embed that question into our cultures, we will accomplish something more valuable than automation. We will redesign the work.
