Leaders Who Model Curious AI Experimentation Build the Trust Required To Scale It Across The Organization
You can't ask 'how do we scale AI?' until you've actually understood what the value is from firsthand use.
Mike Cilla
Vice President of People Operations
Central Health
The workforce data on AI adoption tells a story of tension. BambooHR’s State of the Workforce 2026 report shows that 39% of companies cut headcount last year in part due to AI, 49% of leaders say AI is overhyped, and 54% of workers find AI disruptive to their daily work. Those numbers describe an environment where employees are bracing for replacement, leaders are skeptical of the tools they’re being asked to deploy, and the trust required to make AI adoption productive is eroding before most organizations have even built a strategy. The organizations building real momentum are the ones where leaders have used AI themselves, found genuine value in it, and can model that value visibly enough that their teams see the tool as something that strengthens their work rather than threatens their role.
Mike Cilla, Vice President of People Operations at Central Health, views the situation through a behavioral lens. He holds a master’s degree in industrial-organizational psychology and recently led a full digital HR transformation, including a large-scale UKG implementation that required coordinating across multiple business functions, stakeholder groups, and moving timelines. His personal AI practice shapes how he thinks about what leaders owe their teams before asking them to adopt certain tools.
“Every leader should be using AI personally before they try to scale it for an implementation purpose. You can’t ask ‘how do we scale AI?’ until you’ve actually understood what the value is from firsthand use,” he asserts. Though the principle sounds simple, the practice behind it reveals why most organizations haven’t gotten there yet.
Leader modeling matters more than policy
Research shows that employees whose direct leaders use AI are twice as likely to use it themselves. “Role modeling by the team’s leader is the number one indicator of effective AI usage on teams, so I’ve definitely tried to do that,” Cilla says. The implication is that AI adoption is a cultural signal before it’s a technology deployment. When employees see their leader using AI to prepare for meetings, track complex projects, draft communications, and think through problems, the tool stops looking like a replacement threat and starts looking like a capability upgrade.
Cilla’s own modeling has been deliberate. During the most demanding phase of his digital transformation project, he used ChatGPT as a working journal, executive coach, and project management partner. The tool tracked status across workstreams, surfaced what had fallen behind, and helped him prepare for the day during his commute. “I would open up the voice conversation option and say, ‘Alright, I’m driving into work. Where are we at? What happened last week? What do we need to get done this week?’ By the time I pulled in, I would already have a working game plan.” In this case, the value was not in the automation, but the preparation, context retention, and cognitive offloading that let him show up sharper for the human work the role required.
Curiosity over prompting skill
In most organizations, the conversation about AI skills centers on prompt engineering. Cilla argues that curiosity is the more consequential variable, and that the leaders who find the highest-value applications for AI are those who push it past their own assumptions about its limits. “The curiosity going into the use has allowed me to push AI tools in ways that unlock things I might not have known it could do,” he explains.
His recent migration from ChatGPT to Claude illustrates the point. Rather than starting over with a new model from scratch, he extracted three years of conversation history from OpenAI, migrated the data into Claude, and let the two models collaborate on building what emerged as an ‘AI Chief of Staff Operating Manual’ for his own leadership style. The output mapped how he frames problems, communicates, makes decisions, and engages or abandons challenges. “It blew my mind. I have this psychology background, and I was sitting here thinking, ‘This is better than any 360 evaluation I’ve had recently.”‘ He found the results so beneficial that he recently developed a structured version of the process for public use. Leadership OS takes three user-provided inputs, including your AI conversation history, and develops a personalized user manual for capitalizing on your strengths and using AI more effectively.
The broader lesson is that AI goes beyond a productivity tool and becomes a development tool when leaders bring enough curiosity to use it reflectively rather than transactionally. “If used correctly, AI allows us to become better humans,” Cilla notes.
Start with personal use, then scale with trust
The practical takeaway Cilla offers for HR and people leaders evaluating AI adoption is to reverse the typical sequence. Instead of starting with platform selection and policy, start with personal experimentation. Use AI as a thinking partner, a preparation tool, a coach, and a reflective mirror. Document what works and what does not. Let the team see the practice in action. “Using AI to get better at using AI is a major lever. I’m not sure how many people understand that,” Cilla says. “It has enabled me personally to see new value in it.” Ultimately, employees aren’t going to trust AI because a policy tells them to. They will, however, feel trust having watched their leader use it to show up better prepared, make sharper decisions, and still treat the human work as the part that matters most.
Related articles
TL;DR
Mike Cilla
Central Health
Vice President of People Operations
Vice President of People Operations
The workforce data on AI adoption tells a story of tension. BambooHR’s State of the Workforce 2026 report shows that 39% of companies cut headcount last year in part due to AI, 49% of leaders say AI is overhyped, and 54% of workers find AI disruptive to their daily work. Those numbers describe an environment where employees are bracing for replacement, leaders are skeptical of the tools they’re being asked to deploy, and the trust required to make AI adoption productive is eroding before most organizations have even built a strategy. The organizations building real momentum are the ones where leaders have used AI themselves, found genuine value in it, and can model that value visibly enough that their teams see the tool as something that strengthens their work rather than threatens their role.
Mike Cilla, Vice President of People Operations at Central Health, views the situation through a behavioral lens. He holds a master’s degree in industrial-organizational psychology and recently led a full digital HR transformation, including a large-scale UKG implementation that required coordinating across multiple business functions, stakeholder groups, and moving timelines. His personal AI practice shapes how he thinks about what leaders owe their teams before asking them to adopt certain tools.
“Every leader should be using AI personally before they try to scale it for an implementation purpose. You can’t ask ‘how do we scale AI?’ until you’ve actually understood what the value is from firsthand use,” he asserts. Though the principle sounds simple, the practice behind it reveals why most organizations haven’t gotten there yet.
Leader modeling matters more than policy
Research shows that employees whose direct leaders use AI are twice as likely to use it themselves. “Role modeling by the team’s leader is the number one indicator of effective AI usage on teams, so I’ve definitely tried to do that,” Cilla says. The implication is that AI adoption is a cultural signal before it’s a technology deployment. When employees see their leader using AI to prepare for meetings, track complex projects, draft communications, and think through problems, the tool stops looking like a replacement threat and starts looking like a capability upgrade.
Cilla’s own modeling has been deliberate. During the most demanding phase of his digital transformation project, he used ChatGPT as a working journal, executive coach, and project management partner. The tool tracked status across workstreams, surfaced what had fallen behind, and helped him prepare for the day during his commute. “I would open up the voice conversation option and say, ‘Alright, I’m driving into work. Where are we at? What happened last week? What do we need to get done this week?’ By the time I pulled in, I would already have a working game plan.” In this case, the value was not in the automation, but the preparation, context retention, and cognitive offloading that let him show up sharper for the human work the role required.
Curiosity over prompting skill
In most organizations, the conversation about AI skills centers on prompt engineering. Cilla argues that curiosity is the more consequential variable, and that the leaders who find the highest-value applications for AI are those who push it past their own assumptions about its limits. “The curiosity going into the use has allowed me to push AI tools in ways that unlock things I might not have known it could do,” he explains.
His recent migration from ChatGPT to Claude illustrates the point. Rather than starting over with a new model from scratch, he extracted three years of conversation history from OpenAI, migrated the data into Claude, and let the two models collaborate on building what emerged as an ‘AI Chief of Staff Operating Manual’ for his own leadership style. The output mapped how he frames problems, communicates, makes decisions, and engages or abandons challenges. “It blew my mind. I have this psychology background, and I was sitting here thinking, ‘This is better than any 360 evaluation I’ve had recently.”‘ He found the results so beneficial that he recently developed a structured version of the process for public use. Leadership OS takes three user-provided inputs, including your AI conversation history, and develops a personalized user manual for capitalizing on your strengths and using AI more effectively.
The broader lesson is that AI goes beyond a productivity tool and becomes a development tool when leaders bring enough curiosity to use it reflectively rather than transactionally. “If used correctly, AI allows us to become better humans,” Cilla notes.
Start with personal use, then scale with trust
The practical takeaway Cilla offers for HR and people leaders evaluating AI adoption is to reverse the typical sequence. Instead of starting with platform selection and policy, start with personal experimentation. Use AI as a thinking partner, a preparation tool, a coach, and a reflective mirror. Document what works and what does not. Let the team see the practice in action. “Using AI to get better at using AI is a major lever. I’m not sure how many people understand that,” Cilla says. “It has enabled me personally to see new value in it.” Ultimately, employees aren’t going to trust AI because a policy tells them to. They will, however, feel trust having watched their leader use it to show up better prepared, make sharper decisions, and still treat the human work as the part that matters most.