From Maker to Manager

The Identity Shift AI Requires

The Identity Shift AI Requires

Most knowledge workers haven’t processed what’s actually happening yet. They’re incorporating AI into their existing workflows. Generate the outline, fix the grammar, expand the bullet point. They’re still makers. AI is just a faster tool. That’s Wave 2 thinking in a Wave 3 world. Joe Mariano’s framework from Gartner Digital Workplace Summit 2026 maps exactly where most organizations are standing right now, and why the position is more precarious than it appears. The frame is simple. The gap it reveals is not.

The Three Waves of AI Adoption

Wave 1 is AI as smart intern. The human does the work. AI assists where helpful. The knowledge worker is fully in charge, fully expressing their craft, and using AI the way they would use Google: useful when you need it, optional when you don’t. Most organizations started here. Many called it “AI adoption” when they arrived and congratulated themselves on the milestone. They weren’t wrong to celebrate. Wave 1 is a genuine shift. But it is not the destination. Wave 2 is AI as co-collaborator. The human and AI work together. You generate a draft, refine it, redirect it, and produce something you couldn’t have produced alone. At least not at that speed. The knowledge worker is still the creative authority, but they’ve accepted a collaborator who doesn’t need lunch breaks or vacation time. This is where most organizations live now. It’s also where most AI training programs leave people, which is a problem, because Wave 2 is not the edge. Wave 3 is what Mariano calls headless productivity. The user is the director. The AI is doing the making – writing, analyzing, summarizing, synthesizing, scheduling, drafting. The knowledge worker sets the agenda, provides the context, reviews the output, and makes the decisions. They’re not making. They’re managing AI agents. Managing AI agents is an entirely different job than making things with AI assistance. And making that transition requires something most organizations aren’t prepared to provide.

Why This Is Harder Than It Looks

Here’s what makes the Wave 3 transition different from every other technology adoption most leaders have managed: every previous transition in knowledge work changed what you did. This one is changing who you are in relation to your work. A maker has a craft identity. The quality of the output is the evidence of their skill. Whether the artifact is a design, a data model, an analysis, or a document, it all signals the same thing: I know my domain. I have taste. I’ve earned this role. When AI produces the artifact, that signal disappears. Managing AI agents means your value is no longer in the output itself. It’s in your judgment about what the agent should produce, your ability to evaluate it critically, and your skill at redirecting when it’s wrong. That’s manager thinking, not maker thinking. Most knowledge workers understand this intellectually. Very few have internalized it behaviorally. The distance between those two places is the real adoption gap, and it’s not closed by training programs or prompting workshops.

What the Gap Looks Like in Practice

The identity gap between Wave 2 and Wave 3 isn’t theoretical. It shows up in specific, observable patterns in teams that are stuck. The maker wants to improve the output. The manager asks whether the output is solving the right problem. The maker refines the artifact. The manager changes the brief. The maker is proud of the craft. The manager is proud of the decision. When a knowledge worker hasn’t made this identity shift, they don’t trust AI-generated work because it doesn’t feel like their work. They spend most of their time editing outputs that were directionally correct, adding their voice back in, making the artifact “theirs” in ways that produce diminishing returns. The effort is real. The marginal value of that effort has declined dramatically. The result is a team that has AI tools but isn’t operating at Wave 3 productivity. Gains show up in activity metrics but not in output quality or decision speed, which is where they actually matter. Leaders see the usage numbers and assume adoption is happening. What’s actually happening is Wave 2 with better prompts. There’s a compounding effect worth naming. Knowledge workers stuck in maker mode tend to be busier with AI, not less busy. They’re generating more output, then refining more output, then reviewing more output. The AI has increased the surface area of their craft work without reducing the identity investment required to finish it. That’s exhausting in a specific way. It accelerates burnout and frustration in ways that don’t surface in a usage report until it’s too late.

The Perception Gap That Compounds Everything

One of the most consistent findings in AI transformation research is the gap between how executives and frontline knowledge workers experience AI adoption. Executives see early wins: improved speed, more output, better first drafts. They extrapolate forward and conclude that Wave 3 is essentially here or imminent. They calibrate their expectations and resource decisions accordingly. Frontline workers experience the friction more directly. The tool works. The output isn’t wrong. But something about it feels off, and the effort required to make it feel “right” often exceeds what they’re willing to admit to their manager. They’re not struggling with capability. They’re struggling with identity, quietly, in ways that don’t surface in a productivity metric. This is the gap that breaks AI transformation programs. Leaders who think they’re at Wave 3 make decisions suited for Wave 3: reduce headcount, accelerate timelines, increase output expectations. The team, still operating somewhere between Wave 1 and Wave 2, absorbs the gap as stress rather than progress. Naming the transition explicitly is the first step out of it. Most organizations haven’t done that yet.

What Leaders Need to Do

The shift from maker to manager isn’t trained into people in a single workshop. It’s cultivated over time through the right conditions. There are four that matter most.

Name the transition explicitly. Most organizations haven’t said out loud that they’re asking knowledge workers to change their relationship to their own craft. Giving language to the Wave 3 shift, framing it as a genuine professional development challenge rather than a tool rollout, reduces the defensiveness around it. People can engage with a named transition. They mostly defend against an unexplained tool mandate.

Create space for practice without judgment. The identity shift happens through repeated exposure to AI-generated work, structured experimentation, and honest conversation about what’s still “yours” when the AI produced the artifact. This requires psychological safety: the ability to say “the AI got my first draft” without that signaling incompetence. Organizations that normalize Wave 3 working patterns and build safety around the directing role move faster than those where the unspoken expectation is that real expertise means doing it yourself.

Redefine what quality means out loud. When output speed is no longer a differentiator because AI can generate more than anyone needs, what counts as excellent work changes. Leaders need to make this explicit. Better problem framing. Tighter judgment about what’s good enough. Faster iteration through smarter directing. These are the skills that separate strong Wave 3 performers from everyone else. They need to be named and measured before they can be developed.

Model the transition publicly. Leaders who narrate their own experience of moving from maker to manager give their teams permission to do the same. This is the multiplayer dimension of the shift: the team cannot make the transition if leadership is still performing Wave 2 expertise in meetings and public communication. If the VP refines every AI-generated summary before it goes out and never mentions it, the signal is clear. Wave 3 work is somehow less legitimate. That signal moves through an organization quickly, and in ways that are difficult to reverse.

Why This Is a Multiplayer Problem

AI fluency is contagious in both directions. Teams where most people are at Wave 2 pull Wave 3 thinkers back toward manual work norms. Teams where leadership has made the identity shift and talks openly about managing AI agents create conditions where the transition accelerates for everyone. This is why the maker-to-manager transition is not just an individual development challenge. It’s a team design challenge. The question isn’t whether any given person can make the shift. It’s whether the organizational environment makes the shift possible, or subtly punishes it. Human collaboration remains the highest-leverage variable in AI transformation. Not the quality of the prompt. Not the model selection. The degree to which a team has collectively moved from a maker identity to a manager identity, and built the trust, norms, and communication patterns that make that shared identity stable. That’s a facilitation problem before it’s a technology problem. It requires the same conditions that any identity-level organizational change requires: safety to experiment, visible modeling from leadership, honest naming of what’s actually shifting, and a recognition that the bottleneck is not capability. The bottleneck is psychology. Specifically, the psychology of a workforce that hasn’t yet made peace with directing work they didn’t make themselves.

The Stakes

The organizations that get this right will have a fundamental productivity advantage. Not because they have better tools. Access to AI is roughly equal across most industries right now. Because their people have actually made the identity shift and can operate at Wave 3 consistently. The organizations that don’t will have talented people doing the least valuable work available to them: polishing outputs that were directionally correct, refining work that needed direction not correction, and maintaining a relationship with craft that the technology has already changed whether or not they’ve accepted it. The maker’s job was to get the artifact right. The manager’s job is to get the decision right. That is a significant shift in what work means and where value lives. It doesn’t happen through a training program. It happens through sustained leadership attention, facilitation support, and a willingness to name the transition as exactly what it is: a change in professional identity, not a tool upgrade. The organizations that build those conditions will move faster, produce better work, and retain people who are genuinely growing. The rest will find that their AI investment shows up in usage dashboards but not in business results. The difference is not the technology. It’s whether the humans directing it have made the shift from making to managing.