The whiteboard marker is dying, a pathetic, dry rasp against the glass that sounds exactly like the soul leaving a project. I am staring at the gray streaks left by a man who hasn’t learned a new trick since 1998, yet he holds the keys to every server in the building. He’s currently outlining a 48-step deployment process that involves three manual hand-offs and a literal phone call to a guy named Gary. I suggested a CI/CD pipeline ten minutes ago, and the room went cold. The architect didn’t even look at me. He just sighed, a sound like old parchment rubbing together, and said, ‘We tried something like that in ’98. It didn’t work. Scaling requires discipline, not scripts.’

I’m still reeling from my own idiocy earlier this morning. I walked into the building and threw my entire weight against the glass entrance door. It said ‘PULL’ in large, sans-serif letters, yet there I was, shoulder-first, trying to force my way through a physical impossibility. It’s a humbling thing to be defeated by a door. It makes you realize that your internal map of how the world works-the ‘push’ map-often survives even when the external reality-the ‘PULL’ sign-is screaming at you. This is the fundamental tragedy of the Expert Beginner. They have spent 18 years pushing on pull doors, and they have become so good at the physical act of pushing that they’ve mistaken the resulting sweat for progress.

We mistake tenure for expertise because it’s easier to count years than it is to measure growth. If you do something for 10,008 hours, you are supposed to be a master. But what if you’ve just done the first hour 10,008 times? The Expert Beginner is the most dangerous person in a company because they possess the social capital of a veteran but the intellectual flexibility of a fossil. They have enough authority to stop any evolution that threatens their status. If we move to a cloud-native architecture, their deep knowledge of maintaining that specific 1998-era server rack becomes worthless. Therefore, the cloud is ‘insecure.’ If we adopt a flat management structure, their 58 layers of bureaucratic sign-offs disappear. Therefore, a flat structure is ‘unstable.’

Zephyr J.P. tells me about a room where players spent 88 percent of their time looking behind books. They assumed complexity meant complexity. The Expert Beginner thrives in this assumed complexity. They create 128-page manuals for processes that should take eight clicks. They build ‘immune systems’ for the organization-internal policies and legacy dependencies that attack any foreign body, even if that foreign body is a cure. It’s a survival mechanism. If the process is easy, the gatekeeper is irrelevant.

I remember a project where we tried to implement a simple automated testing suite. The lead dev, a classic EB who had been there since the company was basically three guys in a garage, spent 18 days explaining why our specific edge cases made automation impossible. He wasn’t lying; he truly believed it. He had spent his entire career manually checking those edge cases. His value was tied to his ability to remember them. To automate them was to delete his identity. It’s hard to convince a man of a truth when his salary depends on him not understanding it. When the conversation gets this heavy, I find myself looking for spaces where people are actually solving problems rather than defending territory, places where the community value outweighs the ego of the individual, much like the open discussions you might find at 꽁머니 where the collective experience serves to elevate the group rather than entrench the few.

The Gary Test: Single Points of Failure

Zephyr J.P. finally stops clicking his pen. He walks to the whiteboard. He doesn’t erase anything. He just draws a small, red ‘X’ on the 28th step of the architect’s 48-step process. ‘What happens if Gary doesn’t answer the phone?’ he asks. The architect blinks. ‘Gary always answers the phone. He’s been here 28 years.’ Zephyr smiles, and it’s a little terrifying. ‘In an escape room, if a puzzle relies on one person being in one place at one time, it’s a bad design. It’s a single point of failure. You’ve built an entire company around the reliability of Gary’s cell service.’

The room is silent. The architect looks at the red ‘X’ like it’s a puncture wound. This is the moment where the friction happens. The Expert Beginner doesn’t see a ‘single point of failure’; they see ‘tradition.’ They see ‘the way we’ve always done it.’ They have spent so long working around the limitations of their tools that they’ve started to view those limitations as features. They are like people who have learned to walk with a permanent limp and now believe that anyone who runs is being reckless.

I think back to my ‘push’ door incident. The reason I pushed was that I was in a hurry. I assumed I knew the interface. I didn’t look for the sign. The Expert Beginner is a person who has stopped looking for the sign. They are so confident in their ‘push’ that they will break the hinges of the company before they admit it might be a ‘pull’ door. We spend $888,888 on consultants to tell us why our throughput is low, and the consultants write a 58-page report that boils down to: ‘Your senior staff is blocking the exit.’

The Price of Entrenchment

Organizations die from the inside out, killed by the people who love their current state more than the company’s future. There is a specific kind of grief in watching a brilliant junior dev hit the wall of an Expert Beginner. They come in with ideas, with 108 fresh ways to optimize a database or streamline a client onboarding. They present them with the enthusiasm of someone who hasn’t been crushed yet. And the EB just smiles that indulgent, condescending smile and says, ‘That’s a great thought, but it doesn’t account for the legacy data from ’08.’ And just like that, the air leaves the room. The junior dev realizes that their job isn’t to innovate; it’s to perform the rituals of the past. Eventually, they either leave or they become an Expert Beginner themselves, guarding their own little 18-inch patch of turf.

Zephyr J.P. later told me that the hardest part of designing an escape room isn’t the puzzles; it’s making sure people don’t use ‘brute force’ to solve them. An Expert Beginner is the ultimate brute-force practitioner. They solve every problem with more meetings, more documentation, and more tenure. They don’t look for the elegant solution; they look for the solution that requires the most of their specific brand of history. They confuse ‘difficulty’ with ‘value.’ If it was hard for them to learn in ’98, it should be hard for you to do in 2028.

Rewriting the Metrics

We need to stop rewarding tenure as a proxy for talent. We need to start asking, ‘What have you learned in the last 18 weeks?’ rather than ‘How long have you been here?’ If the answer is just a list of things they used to do, then you don’t have an expert; you have a historian. And historians are great for museums, but they are lethal for tech companies. I’m still thinking about that door. My shoulder still hurts a little. The lesson wasn’t that I was strong; it was that I was blind. The Expert Beginner is strong in all the wrong directions.

Next time you see someone arguing for a process that feels like it was designed in a different era, look at their hands. Are they building something new, or are they just holding the door shut? If it’s the latter, stop pushing. Just look for the sign. It might just be a pull.