The digital clock on the wall says 14:15, and I am watching a 28-year-old in a crisp white t-shirt explain to me why my 15 years of database architecture can be reduced to a Fibonacci number. He calls it ‘story pointing.’ I call it a slow-motion car crash of intellectual dishonesty. He is holding a green sticky note like it is a holy relic, and he is telling me that the refactoring of a legacy schema should only take ‘three points’ because, in his last sprint with a different team, they did something ‘similar’ in two days. He does not know what a schema is. He does not know why the latency spikes at 03:15 every Tuesday. But he has a certification, a stopwatch, and a very expensive haircut.
⚠
My left eyelid is twitching. It is the same twitch I had last month when I accidentally laughed at a funeral. It wasn’t that the death was funny; it was the absurdity of the eulogy, the way the speaker tried to turn a complex, messy life into a series of 5 bullet points. We are doing the same thing to our crafts. We are taking decades of intuition, the kind of knowledge that lives in your bones, and we are trying to process-manage it into oblivion.
I remember when expertise was a North Star. Now, it is treated like a bottleneck. In the world of the Agile Generalist, being a master of something is actually a liability because you are no longer ‘interchangeable.’ If you are the only one who can solve the 45 complex edge cases in the payment gateway, you are a risk to the velocity of the sprint. The goal is no longer to build the best thing; it is to build the most predictable thing. And if ‘predictable’ means ‘mediocre,’ the process managers will take that trade every single time. They would rather have a team of 15 generalists who are all equally average than one expert who might occasionally say, “No, that is a terrible idea and it will break in 25 ways you haven’t considered yet.”
The Map Replaces the Territory
Intuition Priced In
Interchangeability Valued
Adrian J., an algorithm auditor I know who spends 55 hours a week looking for bias in automated hiring systems, told me that we are essentially training humans to behave like the very machines we are trying to replace them with. He says the tragedy isn’t that the machines are getting smarter, but that the organizational structures are demanding that the humans get dumber, or at least narrower. Adrian J. once found a bug that would have cost a bank $575 million, but his manager was annoyed because the discovery happened outside of the ‘testing phase’ defined in the quarterly roadmap. We have reached a point where the map is not just more important than the territory; the map has replaced the territory entirely.
[the process is the product now]
Think about the last time you were truly impressed by someone’s work. It probably wasn’t because they followed a Jira workflow to the letter. It was likely because they had that rare, deep-seated competence that allowed them to see a problem before it manifested. It’s the difference between a person who can read a script and a person who understands the story. We see this erosion everywhere, from software engineering to city planning, where ‘data-driven’ becomes a shield for people who don’t want to make a difficult, expert-led decision. They want the numbers to tell them what to do so they don’t have to be responsible for the outcome.
The Cost of the Soul Leaving the Work
I’m not saying that structure is bad. I’m saying that structure without respect for craft is a prison. When we treat a senior engineer with 15 years of experience the same way we treat a junior who just finished a 15-week bootcamp, we aren’t being ‘agile.’ We are being delusional. We are ignoring the 10,000 mistakes the senior has already made and learned from. We are ignoring the fact that a ‘two-point’ task for an expert might be a ‘two-week’ disaster for a generalist, yet we price them the same in our imaginary points economy.
The Resource Mindset
There is a massive, unspoken cost to this. It’s the cost of the soul leaving the work. When you stop being an expert and start being a ‘resource,’ you stop caring about the long-term health of what you’re building. You start caring about clearing the board. You start gaming the metrics because the metrics are the only thing the 28-year-old scrum master understands. If they want 45 points of velocity, you’ll give them 45 points of velocity, even if 25 of those points are just fluff and technical debt that someone else will have to pay for in 5 years.
This trend of devaluing the specialist in favor of the process-compliant generalist is creating a world of surface-level experiences. It’s the ‘IKEA-fication’ of professional life. Everything is flat-packed, easy to assemble, and ultimately disposable. But some things shouldn’t be disposable. Some things require the nuance of someone who has spent 35 years walking the same streets or 15 years looking at the same codebase. This is why I find myself gravitating toward businesses and experiences that still celebrate the individual guide, the person whose knowledge isn’t just a manual they read last night.
For instance, if you are looking for a segway tour koeln, the difference between a guide who just knows the route and one who actually knows the history of every 15th-century stone in the city is the difference between a chore and a memory. The expert guide isn’t following a rigid, point-based sprint; they are reacting to the environment, the group, and the hidden stories that only reveal themselves to those who have looked long enough. They aren’t interchangeable ‘tour-providing units.’ They are the value.
We’ve lost the ability to value that. We think we can automate it, or document it into a wiki, or manage it with a stand-up meeting every morning at 09:15. We are wrong. Documentation is just the shadow of expertise; it isn’t the light itself. When the expert leaves the room, the documentation becomes a tombstone for the knowledge that used to live there. I have seen 25-page SOPs that couldn’t explain what a master technician knows by simply listening to the sound of a motor for 5 seconds.
The Limits of Speed
Adrian J. often jokes that he’s going to retire and become a woodworker. He says wood doesn’t care about your agile methodology. Wood has grain, it has history, and if you don’t respect the material, it will split. You can’t ‘story point’ a piece of oak into becoming a chair faster than the wood allows. There is a physical limit to how fast excellence can happen. In the digital world, we’ve forgotten those physical limits. We think that by adding 5 more people to a project, we can make it go 5 times faster, despite 45 years of evidence to the contrary. Brooks’s Law is as true today as it was in 1975, but we ignore it because it doesn’t fit the ‘velocity’ narrative.
Expertise is a conversation with reality
The Language Barrier
I’m sitting in this meeting, and the scrum master is asking if we can ‘parallelize’ the database migration. I try to explain that it’s like trying to have 9 women give birth to a baby in one month. He looks at me with a blank expression and asks if that’s a ‘blocker’ or a ‘risk.’ I realize then that we aren’t even speaking the same language. I am speaking the language of reality, and he is speaking the language of the spreadsheet. He isn’t interested in the 5 reasons why his plan will fail; he is interested in whether he can mark the task as ‘In Progress’ by 17:45.
I recognize my own hypocrisy here. I’m still sitting in the chair. I’m still taking the paycheck. I’m still pointing the stories, even though I know the numbers are made up. I’m part of the problem. That’s the most painful part of the realization. We’ve built a system that is so efficient at self-preservation that even the people who hate it find themselves feeding it. I told Adrian J. this, and he just laughed. It was that same dry, inappropriate laugh from the funeral. “We’re all just auditors now,” he said. “We’re just auditing the decline.”
The Path Back: Cultivating Nuance
But maybe there’s a way back. Maybe it starts with admitting that not everything can be measured, and not everyone is replaceable. Maybe we need to stop asking ‘How fast can we do this?’ and start asking ‘Who is the best person to do this?’ We need to create spaces where expertise is allowed to be messy, where ‘three points’ can take three weeks if that’s what the quality requires. We need to stop treating the 28-year-old with the certification like he’s the captain of the ship and start treating him like what he is: the person who holds the clipboard while the actual sailors navigate the storm.
Predictable, Manageable, Average
Messy, Deep, Lasting
In the end, the ‘agile’ movement was supposed to be about people over processes. It was right there in the manifesto. But like every other good idea, it got swallowed by the corporate machine and turned into its exact opposite. It became a way to prioritize process over people, to turn the person into a variable that can be solved for. But you can’t solve for expertise. You can only cultivate it, respect it, and-if you’re lucky-get out of its way long enough for it to build something that actually lasts more than 5 minutes after the next deployment. If we don’t start doing that soon, we’re going to find ourselves in a world where everything works perfectly on paper, and nothing works at all in the real world. And that’s not a ‘risk.’ That’s a guarantee.