Will Drover is Professor of Entrepreneurship & Innovation and Department Chair at the Neeley School of Business, Texas Christian University. He serves as Founding Director of Neeley AI Forward and as the Dean’s Advisor on AI. His work on AI adoption has appeared in *MIT Sloan Management Review*, with coverage in the *Wall Street Journal* and *Los Angeles Times*. Drover teaches graduate courses on applied AI, leads executive education on AI strategy and leadership, and holds ownership stakes in early-stage AI and robotics ventures.
Are you really talking to your employees about AI?
I recently ran an AI strategy session for a leadership team that had followed the standard playbook: training programs, access to multiple AI tools, and a no-code platform. Then I asked a simple question: How many of you have built something with AI that actually changed how work gets done?
One hand went up.
Most people in the room considered themselves AI users. Almost none saw themselves as builders. No personal agents. No custom assistants. No reusable workflows.
That distinction is critical. Using AI for one-off tasks delivers a temporary productivity boost. Building turns that boost into a durable tool or workflow that can be reused and scaled. Leaders often track how many employees use AI. The more revealing metric is how many are building with it. Call it the builder activation gap: the distance between the many people who *could* build with AI and the few who do.
In executive education sessions and applied AI courses for working professionals, I see the same pattern repeatedly. Nearly anyone who can describe what they want in plain English can now create a working assistant, app, or automation without writing code. Yet relatively few are building anything that matters.
A couple of years ago, Caroline Davis, Chief of Staff at Capital Factory, saw herself as an AI user, not a builder. When I asked her in an applied AI course who had ever created a working tool with the technology, her hand stayed down.
Today, many recurring parts of her job run through tools she built herself. The centerpiece is an agent named Sunny, after her daughter. It connects to her email, calendar, Airtable CRM, and Google Sheets, drawing on roughly a dozen documented workflows that prepare briefs, track fundraising, onboard new investors, and more. Tasks that once took hours now take 10 to 15 minutes. Several automations run on a schedule, finishing work before she even asks. The workflows are versioned, reused, and improved rather than discarded after a single use. Sunny also coordinates with other agents. Davis has even applied the same approach outside work, using AI to rebuild a photography business she ran a decade earlier, including its website.
It started small. In the course, she used natural language to build her first working AI assistant. What followed, she says, was less about technical mastery and more about confidence. That first experience “built up my acumen as a whole and gave me the confidence to test out stronger AI models.” She stopped treating AI as a reactive chat helper and began treating it as a system she could design to handle recurring work. One build led to the next, and eventually to the library powering Sunny. Along the way, her self-perception shifted.
She still downplays the expertise. “I don’t think I’m a power user by any means,” she told me, “but most of my day is run through Claude at this point.”
Her path remains uncommon. Roughly half of U.S. employees now use AI at work at least occasionally, but only 15% do so daily, according to Gallup. A recent *Harvard Business Review* analysis of 1.4 million AI interactions among more than 2,500 KPMG employees found that only about 5% qualified as sophisticated users—those doing iterative, higher-impact work beyond casual prompting. For most people, AI still means assistance with the immediate task: drafting an email, summarizing a document. Useful, but disposable.
So where are the builders?
Some barriers are structural—governance, access, time, and incentives. Once those are addressed, identity often becomes the hidden bottleneck. Enterprise work has long trained people into a division of labor: a small group of specialists builds systems; everyone else operates inside them. That made sense when building required deep engineering skill. For many everyday problems, it no longer does. What hasn’t changed is how people see themselves. Most employees still identify as consumers of technology, not creators. And identity is stubborn. As Herminia Ibarra’s research on reinvention shows, people rarely think their way into a new self-image. They act their way into it, and identity follows.
Specialists still own complex systems, security-sensitive applications, and anything customer-facing that requires tight oversight. But a large share of day-to-day work problems sit in safer territory—places where the person who builds the tool is also the one best positioned to judge whether it works. What stops many people is the quiet belief that building is “not my lane.”
AI has made building far more accessible. It has not yet convinced most people that it is for them. Every organization has employees sitting where Davis was two years ago. The question is whether leaders leave them there.
Three practices help close the gap and activate a builder identity across the workforce.
1. Make the first build unavoidable.
Training alone rarely shifts identity. A required build can. In my sessions, every participant must create a tool or agent that solves a real problem and then demo it to the room. Once it works, posture changes. People who arrived as AI users begin describing what they built. SharkNinja ran a similar approach at scale, pausing normal work for a four-day, company-wide AI hackathon involving roughly 4,000 employees. Leaders assigned about 20 major initiatives; employees added roughly 400 projects of their own. The mindset shifted from waiting on IT to “I have a problem. I can fix it.” Force the first successful build, and a builder identity starts to take root.
2. Make builders visible.
When the only people seen building are engineers, most employees default to seeing themselves as users. When peers and leaders create practical solutions that streamline work, the reference point expands. Airtable’s CEO Howie Liu builds in the open, deliberately sharing what he creates so the company sees its leader shipping. Rather than circulating a document about a new capability, he has built landing pages in Replit and shared the links—and even the prompts—so others can follow the method. He pushes teams toward “prototypes over decks”: working demos that can be tried, not words in a product requirements document. When leaders build publicly and nontechnical colleagues demo useful tools, the definition of who gets to build begins to change. Visibility creates more builders.
3. Measure the builds.
Most companies track proxies: active users, tokens consumed, employees trained. Those numbers show who is using AI. They reveal little about who is creating deeper value. Better questions focus on what was actually built, whether others adopted it, and whether it materially changed a workflow. At BBVA, employees have created more than 20,000 custom GPTs; roughly 4,000 are now in frequent use. Tracking those figures surfaces the builders already present and signals that building is expected, counted, and valued beyond the IT department. Job titles stop deciding who is allowed to build.
Organizations looking for greater returns from AI will not find them in adoption metrics alone. They need to narrow the builder activation gap—turning more employees who see themselves as users into people who build. So return to the opening question. Ask your people what they have actually built with AI, and notice whether the room goes quiet. The real work is making sure it doesn’t.