The blue light from the dual monitors is currently searing a very specific, rectangular-shaped headache into the bridge of my nose. I can feel the pulse behind my left eye syncing up perfectly with the blinking cursor on the screen-share. Across the digital void, Mark-my manager, though ‘facilitator of software’ feels more accurate-is hovering his mouse over a crimson-colored bar on a burndown chart. The fan in my laptop is doing that high-pitched, desperate whine again, the one that sounds like a mosquito trapped in a tin can, struggling to keep up with the weight of forty-eight open browser tabs and the heavy existential dread of another Monday morning.

‘Looks like your ticket completion rate was down 8% last week, Avery,’ he says. He doesn’t ask how I am. He doesn’t ask about the AI training sets I’ve been curating, which have become increasingly surreal lately. He doesn’t ask about the fact that I spent eighteen hours of my personal time last week trying to figure out why the model kept identifying images of clouds as ‘industrial smoke.’ He just sees the 8% dip. To Mark, I am not a person who curates the logic of a machine; I am a variable within a different machine, one called Jira, and currently, my variable is underperforming. It’s the quantification of trust, or rather, the total outsourcing of it. We’ve reached a point where management doesn’t require human intuition or observation anymore. It just requires a login and a basic understanding of X and Y axes.

The Illusion of Velocity

I’m Avery T., and my job title is ‘AI Training Data Curator,’ which is a fancy way of saying I spend my days teaching silicon brains how to recognize the difference between a stop sign and a very persistent pigeon. It’s tedious, soul-sucking work that requires a strange blend of high-level linguistic understanding and the patience of a saint. But Mark doesn’t see the pigeons. He sees the throughput. He sees the ‘velocity.’ He sees the data points that end in 8 and 18 and 28, and he assumes those numbers tell the whole story of my life.

This is the great abdication of our era: the belief that if the dashboard looks healthy, the people must be thriving. It’s a convenient lie because it allows managers to avoid the messy, uncomfortable, and deeply inefficient work of actually talking to their employees about their lives or their careers.

I recently had to explain the internet to my grandmother, which was an exercise in profound humility. She’s eighty-eight years old and still thinks of a ‘platform’ as something you stand on to board a train. I told her the internet is basically a giant, invisible library where all the books are on fire and everyone is screaming at the top of their lungs, but also there are pictures of cats.

She asked if the ‘Cloud’ was where the files went when it rained. I laughed, but then I realized that my manager’s view of my work isn’t much more sophisticated than Nana’s view of the Cloud. To him, my productivity is a meteorological event. It happens, it’s tracked by instruments, and if there’s a drought of tickets, he puts on his metaphorical raincoat and asks for a status update. There is no acknowledgment of the terrain or the soil quality. There is only the rain gauge.

Managing the Tool, Not the Person

This obsession with software-led management has created a weird, hollowed-out version of the workplace. In my previous role, I worked for a guy who actually walked around. He’d see the look on your face-that specific ‘I’ve been staring at Regex for six hours’ thousand-yard stare-and he’d tell you to go take a walk or grab a coffee. He managed the person. Mark manages the tool. He spends his entire day optimizing the Jira workflow, adding custom fields, and tweaking the automation rules, seemingly unaware that the humans on the other end of those rules are slowly evaporating into a cloud of resentment.

There’s an irony in my work with AI. I spend 1508 hours a year trying to make machines more human, while my boss spends his time trying to make me more like a machine. I’m teaching a model to recognize ‘sarcasm’ and ‘nuance’ in text, while Mark is stripping all the nuance out of our 1-on-1s. He wants binary. He wants ‘Done’ or ‘In Progress.’ He doesn’t want ‘I’m struggling with the ethical implications of this dataset and it’s making me move slower because I’m double-checking the bias.’ That doesn’t fit in a 238-pixel wide column.

The Beauty of Physical Presence

We’ve lost the sense of the physical environment in all this digital noise. Everything is flat. Everything is a screen. Sometimes I crave something with actual texture, something that doesn’t refresh every eight seconds. I was looking at the new office designs for a friend’s firm, and they were focusing on the tactile-real wood, stone, and architectural elements that you can actually touch. It’s a funny transition to think about, because even when we try to fix our physical spaces, we often look for shortcuts.

But there’s a beauty in things that are built to last, that have a physical presence. It’s like when you see a well-executed exterior; there’s a structural honesty there. It reminded me of some of the materials from Slat Solution, where the goal is to create a facade that actually adds value and protection, rather than just hiding a mess behind a digital curtain. In software management, the dashboard is the facade, but unlike a solid wall, it provides no protection for the people behind it. It just makes the house look finished from a distance while the foundation is rotting.

Gaps Between the Data Points

I’m not saying we should go back to paper ledgers and telegrams. I love my 1008-pixel wide monitor. I love that I can work from home in my pajamas while teaching an AI to understand the concept of ‘joy.’ But we have to stop pretending that the map is the territory. A Jira ticket is a representation of work, not the work itself. When a manager forgets that, they stop being a leader and start being an administrator of a very expensive, very boring video game. They lose the ability to see the eighteen different ways an employee might be contributing that aren’t captured by a ‘points’ system.

They lose the trust of their team, because trust isn’t built on a dashboard. Trust is built in the gaps between the data points. It’s built during the five minutes of a call when you aren’t talking about the project, but about the fact that your cat is sick or that you’re worried about the state of the world.

Hanging Up the Line

Mark finally stops sharing his screen. The relief is physical. The glare on my glasses subsides, and I can see his face again. He looks tired, too. He’s spent his morning looking at fifty-eight different charts for fifty-eight different employees. He’s probably just as overwhelmed as I am, but his role has been so narrowly defined by the tools we use that he doesn’t know how to reach out across the digital divide. He asks if there’s anything else he should know. I want to tell him about the 158 images of ‘dogs’ that were actually ‘mops’ and how it made me laugh for ten minutes. I want to tell him that my grandmother thinks he lives in the Cloud. I want to tell him that I’m more than an 8% dip.

Instead, I just look at the battery icon on my laptop-currently at 88%-and say, ‘No, Mark. The tickets are all updated. Everything is in the system.’ He nods, satisfied that the software is happy. We hang up. I sit in the silence of my room for a moment, the high-pitched hum of the laptop fan finally dying down as the Zoom process ends. I realize that I’ve become quite good at managing the software, too. I know exactly what to click and when to click it to make the charts look the way he wants them to look. We are two people, separated by a million miles of fiber-optic cable, both performing for a piece of project management software that neither of us particularly likes. It’s a strange way to live, but as long as the velocity stays high, I suppose we’ll keep playing our parts in the machine. I just hope that one day, we’ll remember how to manage the people again, not just the tickets they leave behind.