A phrase has been circulating through the AI world: everyone becomes a manager. Melissa and Johnathan Nightingale (Raw Signal Group) are among the sharpest voices to push back on it, and the pushback is right. The line sounds clever until you stop and ask what management actually is. It's worth taking the objection one step further than "that's not what management means" — because the mistake underneath the phrase tells you something about how organizations are already misreading the work AI is creating.

Much of the current AI conversation depends on inflated metaphors. We're told workers will command fleets of agents, supervise digital employees, and rise into higher-order roles where software handles the execution. It's an appealing story because it turns anxiety into status. You are not being displaced. You are being promoted.

But coordinating tools is not the same thing as managing people.

Real management has never been about assigning tasks alone. It's about earning trust from people who don't owe you instant confidence. Coaching someone whose talent exceeds their discipline. Giving hard feedback without humiliating anyone, recognizing effort without lowering standards, holding accountability when avoidance would be more comfortable. Understanding motivation, navigating conflict, building momentum, helping a team perform under pressure. No AI agent needs reassurance after a failed presentation. No model quietly resents a peer who was promoted first. No dashboard is burned out, underestimated, distracted by family stress, or wondering whether it belongs. Human management deals in emotion, identity, fear, ambition, and relationship. That's why it's hard.

What many workers are more likely to experience is something else entirely. They become reviewers of machine output. Editors of rough drafts. Approvers of recommendations. Coordinators of workflows that move faster than they used to. Instead of doing one task start to finish, they oversee five partially automated streams and decide what advances, what gets fixed, and what gets rejected.

That's a meaningful shift. It just isn't management.

In many cases, AI doesn't remove work so much as relocate it. The visible labor shrinks while the invisible labor grows. A person who once wrote one proposal now evaluates six generated versions. A marketer who once drafted a campaign now chooses among dozens of options and checks each for tone, accuracy, and strategic fit. An engineer who once wrote straightforward components now spends more time validating generated code, tracing errors, and deciding what should never have been suggested in the first place.

Output becomes abundant. Judgment becomes scarce.

That's the labor market shift many leaders are underestimating. When production accelerates, decision-making becomes the bottleneck. Someone still has to determine what matters, what aligns with priorities, what introduces risk, what can wait, and what should be ignored entirely. More options often create more drag, not less.

This is where decision fatigue enters quietly. A worker can end the day exhausted despite producing little directly. They reviewed twenty suggestions, approved twelve, rejected eight, rewrote three, escalated two, and carried a low-grade anxiety that something flawed slipped through. Busy all day. Also strangely detached from the work itself.

This will confuse organizations that still measure contribution through visible activity. They'll see dashboards full of movement and assume productivity is rising. They'll read speed as progress and volume as value. Some leaders will mistake oversight systems for leadership itself.

That mistake is already familiar. Dashboards can show missed deadlines, declining throughput, sentiment scores, ticket counts, response times. They cannot tell you that your strongest employee has stopped caring. They cannot reveal that two peers no longer trust each other. They cannot sense when fear has replaced candor in a meeting. They cannot inspire effort during a hard quarter, or make someone believe growth is still possible after failure.

Leadership starts where reporting stops.

Here is the part the tool conversation keeps missing, and it's where this gets more uncomfortable than "learn to use judgment." Judgment is not an individual trait you can hire for. It's a team capacity — and teams build it through exactly the practices AI is quietly removing. The code review where a junior engineer learns why a technically correct solution is still the wrong one. The hallway exchange that reveals how the company actually makes decisions. The messy meeting where people argue their way toward a shared sense of what matters. Those routines were never only about producing work. They were how discernment moved from the people who had it to the people who didn't. Automate the production and you can accidentally automate away the apprenticeship. Output climbs. The mechanism that builds judgment goes quiet. Organizations discover, usually too late, that judgment was infrastructure, not talent — load-bearing, inherited, running beneath the work whether anyone tended it or not.

So this is why human leadership may matter more in the AI era, not less. As routine execution gets easier, organizations need more people capable of discernment, context-setting, conflict resolution, prioritization, and real coaching — and they need the conditions that let those capacities develop in the first place. As communication becomes easier to automate, authentic communication becomes more valuable. As work fragments into systems and handoffs, someone still has to create coherence.

The premium skill won't be prompting tools. It will be judgment. Judgment knows when the efficient answer is the wrong answer. It knows when more analysis is avoidance. It knows when a team needs pressure and when it needs recovery. It knows when a generated recommendation is technically sound but culturally disastrous. It knows which problem matters now. That capability will separate high performers more than any technical trick.

So no, everyone is not becoming a manager. The phrase mistakes software supervision for people leadership, and complexity for status. Many workers will instead become editors, orchestrators, reviewers, and decision-makers operating inside faster systems with heavier cognitive demands.

That's still a profound change. It deserves to be described honestly. Because if leaders misunderstand the nature of the work, they'll design the wrong jobs, reward the wrong behaviors, and burn people out under the banner of innovation. And if workers misunderstand it, they may chase the language of management while missing the real opportunity in front of them: to become someone trusted for judgment, in a moment when judgment is quietly becoming the rarest thing in the building.