The fluorescent light in Conference Room 3 is flickering at a frequency that suggests it’s about to have a nervous breakdown, or maybe it’s just trying to communicate in Morse code. Owen F. is clicking his pen-a cheap, blue plastic thing that sounds like a miniature hammer against my skull. We are staring at a screen that displays a Gantt chart so beautiful it belongs in a museum of impossible dreams. The bars are long, straight, and perfectly diagonal, marching toward a launch date that exists in a future where no one ever catches a cold or forgets a password. Owen, our lead seed analyst, looks at me with the eyes of a man who believes the numbers more than he believes his own mother. He sees 103 milestones; I see 103 opportunities for the universe to laugh at us.

Owen F. taps the screen, specifically at a task labeled ‘API Integration-Vendor B.’ He’s allocated 13 days for it. ‘They’re reliable,’ he says, as if reliability is a static property of a corporation and not a fragile state of grace maintained by overworked humans. I want to tell him about the 233 unread emails in Vendor B’s support queue right now. I want to tell him that the lead engineer at Vendor B just bought a puppy that doesn’t sleep, and therefore, for the next 13 days, that engineer’s cognitive capacity will be reduced to that of a lukewarm potato. But Owen is an analyst. He analyzes seeds. He thinks if you put the right data in the soil, the growth is a mathematical certainty. He ignores the fact that a stray dog might decide to dig up that soil just because it likes the way the dirt feels between its toes.

Planning is a psychological defense mechanism against the terrifying randomness of our colleagues. We create these charts not to map the future, but to calm our current anxiety. We pretend that ‘Coordination’ is a zero-cost activity. We assume that if we have 3 stakeholders, the communication lines are simple, ignoring that 3 people actually create 3 distinct relationships, and 13 people create 78 potential points of misunderstanding. It’s a geometric explosion of ‘Wait, I thought you were doing that’ and ‘I didn’t see that attachment.’

Take the legal department, for instance. Our plan has a ‘Legal Review’ window of 3 days. It’s a tidy little box at the end of the month. But Legal isn’t a box. Legal is a sentient entity composed of 13 different lawyers, each of whom has a unique interpretation of risk and a shared love for the word ‘notwithstanding.’ One of them will go on vacation; another will find a comma that offends their very soul. Suddenly, your 3-day window has expanded into a 43-day odyssey through the underworld of redlined PDFs. Owen F. doesn’t account for the soul of the lawyer. He accounts for the ‘function’ of the lawyer. This is the fundamental error of the modern manager: treating humans as interchangeable units of production rather than as volatile chemical compounds that react unpredictably when mixed.

I’ve spent the last 3 years watching projects bleed out because of this. We build these systems that are ‘lean’ and ‘optimized,’ which is just another way of saying they have zero tolerance for human error. If Brenda in dev-ops has a bad day, the whole architecture wobbles. If the CEO watches a 3-minute video on a new technology over the weekend, the entire scope of the project shifts by 33 percent on Monday morning. We are building glass houses in a world where everyone is holding a bag of marbles.

There’s a strange comfort in the chaos, though. If everything went exactly according to the plan, we wouldn’t be needed. We’d just be the fleshy interfaces for a sequence of automated events. The friction is where the humanity is. It’s in the frantic 13-person Zoom call where someone finally admits they don’t know how the database works. It’s in the 3:03 AM Slack message where a junior dev finds the bug that four seniors missed. This is why certain industries, despite their complexity, manage to maintain a sense of order. In the high-stakes world of digital entertainment and operational reliability, platforms like Gclub have to build their entire infrastructure around the assumption that humans are going to be chaotic, that traffic will spike at the worst times, and that the system must survive the people using it, not just the code running it.

Owen F. is still talking. He’s moved on to ‘Phase 3: Stabilization.’ He’s allocated $833 for unforeseen contingencies. I almost laugh. $833 wouldn’t cover the coffee budget for the amount of overtime we’re going to need when the vendor’s API turns out to be documented in a language that hasn’t been spoken since the 13th century. But I don’t say anything. I just watch the pen click.

I’m thinking about the elevator again. When the doors finally opened, the technician didn’t apologize for the ‘system failure.’ He just said, ‘Yeah, the heat makes the sensors twitchy.’ The heat. Not a logic error, not a mechanical breakdown. Just the environment being itself. We forget that our projects live in an environment. They live in an office that’s too cold, or in a remote-work world where the neighbor’s lawnmower is a constant acoustic threat. They live in the hearts of people who are worried about their mortgages or their 3-year-old’s cough.

We need to stop calling them ‘delays’ and start calling them ‘the cost of being alive.’ If a project takes 103 days instead of 53, it’s not necessarily because someone failed. It’s because 103 days is how long it took for the human elements to align. Owen F. would argue that better data leads to better timelines. I’d argue that better empathy leads to more realistic ones. We should build plans that have ‘pockets’ for the inevitable. A pocket for the week the lead designer gets dumped. A pocket for the 3 days the entire team spends arguing about the hex code for a shade of blue that no one will ever notice.

We pretend that if we measure enough things, we can control them. We track 43 different KPIs. We use 3 different project management tools that all sync with each other in a recursive loop of productivity theater. Yet, the most important metric-the ‘Human Friction Index’-is never on the dashboard. It’s the measure of how much energy is lost just trying to get two people to agree on what ‘done’ looks like.

Owen finally stops clicking. He looks at me, waiting for approval. He wants me to say that the diagonal line is achievable. He wants me to validate the fantasy. I look at the chart, and then I look at the 13 tabs open on my own laptop, each representing a different person who hasn’t replied to my messages yet. I think about the elevator and the feeling of being suspended in a void, waiting for a human hand to reset the system.

As we walk out, the light finally gives up and dies. The room goes dark, save for the glow of Owen’s laptop. He doesn’t even notice. He’s already thinking about the next 3 weeks. I’m just hoping the elevator works this time. I’m hoping that whatever sensors are twitching in the dark decide to cooperate for just long enough to get me to the ground floor. Because in the end, that’s all any project really is: a series of small, miraculous cooperations between flawed parts, moving toward a destination we only half-understand, while we try not to think too much about the cables holding us up.