Your project management tool was made for a non-human (AI) factory, not for you

Every project or task management tool on the market asks the same opening question: what needs to get done? The question seems reasonable. And, it is the deepest source of the problem.

If you’re spending your days moving cards across columns and updating status fields, ask yourself: is this the work, or is this the appearance of work?

When Frederick Winslow Taylor published The Principles of Scientific Management in 1911, he was refreshingly honest about what he wanted from workers. The ideal laborer for handling pig iron, Taylor wrote, “shall be so stupid and so phlegmatic that he more nearly resembles in his mental make-up the ox than any other type.” Work could be decomposed into motions. Motions could be timed with a stopwatch. Optimization meant removing the worker’s judgment from the equation entirely.

Taylor’s protégé Henry Gantt designed his charts for repetitive factory operations. The Critical Path Method came from DuPont plant shutdowns. PERT was built for the Polaris missile program. Each generation of tooling carried forward the same assumption: work is a machine problem. Decompose it into parts. Sequence the parts. Track the parts. Ship.

Now look at your screen. Kanban boards. Story points. Sprint velocities. Burndown charts. The interface got prettier. The assumption didn’t change.

Peter Drucker saw this coming in 1967. Knowledge work, he argued, was fundamentally different. “The knowledge worker cannot be supervised closely or in detail. He can only be helped. But he must direct himself.” More pointedly: “The worker must define what the task is or should be. And only the knowledge workers themselves can do that.” He noted that each of his requirements for knowledge worker productivity was “almost the exact opposite of what is needed to increase the productivity of the manual worker.”

The tools never caught up. They still ask what needs to get done as if someone else has already decided. They still model work as a pipeline of discrete units flowing toward completion. They still assume that if you can see all the tasks and their deadlines, the work will happen.

Here’s what actually happens.

You open the task board. There are 47 items. Twelve are overdue. The cognitive cost of just looking at this is measurable. Bluma Zeigarnik demonstrated in 1927 that incomplete tasks are remembered roughly twice as well as completed ones — your brain treats every undone item as an open loop, generating intrusive thoughts that hurt your performance on everything else. Your project tool is not a neutral record. It’s an anxiety-generating machine.

You pick a task. It says “Improve onboarding experience.” What does that mean? Where do you start? Csikszentmihalyi spent decades studying flow, that state where challenge and skill align, where action and awareness merge, where time distorts and work becomes genuinely absorbing. He was explicit about the preconditions: you need to know precisely what to do, moment by moment. Vague tasks don’t create flow. They create anxiety. And anxiety triggers avoidance. Research by Pychyl and Sirois confirms that procrastination is not a planning failure. It’s an emotion regulation failure. The task generates negative affect — frustration, confusion, dread — and avoidance provides immediate relief.

So the tool that was supposed to help you work is now actively preventing you from starting. Not because you’re lazy. Because the tool’s model of work is wrong.

You start anyway. Three minutes in, a notification fires. Someone updated a ticket. Someone @-mentioned you. Gloria Mark’s field research found that workers switch activities every three minutes and five seconds, and after each interruption, it takes an average of 23 minutes and 15 seconds to return to the original task. You pass through roughly 2.3 intervening tasks on the way back. Your project management tool is an interruption machine that bills itself as a productivity tool.

The deeper issue is philosophical, not just practical. Aristotle distinguished between poiesis — making, where the point is the product — and praxis — action whose meaning is in the doing itself. Building a house is poiesis. Exercising judgment under uncertainty, navigating ambiguity, deciding what matters — that’s praxis. Task management tools model all work as poiesis: a deliverable to be produced, a ticket to be closed, a status to be changed.

But knowledge work is full of praxis. The most important things you do at work — deciding what to prioritize, making calls without enough information, building trust with colleagues, figuring out what the actual problem is — can’t be captured in a task field. They don’t have acceptance criteria. They don’t have story points. They are the work, and your tool can’t see them.

Heidegger had a useful concept for this. Tools that work well are zuhandenheit — ready-to-hand. They disappear. A hammer in use is invisible; the carpenter sees the nail. When a tool breaks, it becomes vorhandenheit — present-at-hand. Suddenly conspicuous. Demanding attention instead of enabling work.

Your project management tool is chronically present-at-hand. You don’t work through it. You work around it. You maintain it. You attend to it. You spend time managing the tool that was supposed to manage the work. The time you spend feeding the system is time stolen from the work the system was supposed to support.

I run 15 software projects with autonomous AI agents. The agents use GitHub issues, labels, and milestones — the simplest possible task infrastructure. No story points. No sprint planning. No velocity tracking. No burndown charts. The agents don’t need those things because those things were never about the work. They were about making the work legible to managers. When the workers are agents, the pretense collapses. Nobody is estimating how long a task will take in story points because the question was always a fiction.

But the humans in this system don’t use task boards either. They make decisions about what to build, when to launch, which market to enter. They exercise judgment. They set direction. The work that can’t be ticketed.

If you’re spending your days moving cards across columns and updating status fields, ask yourself: is this the work, or is this the appearance of work? Taylor wanted the system to be first and the man to be second. A hundred years later, the system is still first. The tools are still asking what needs to get done, as if the answer is a list.

The answer was never a list. The answer is judgment, applied in the moment, to a situation that has never existed before and won’t exist again.

Your Jira board can’t hold that. It was never designed to.