In the span of ten seconds, a fleeting moment, lives can be transformed forever. 100 Meters (Hyakuemu), produced by Rock 'n' Roll Mountain, is a film based upon this very fact. While this post isn't a film review or critique of the film, I want to "delve" deeper into topics of pursuit. Specifically, what does pursuit mean for those of us who build software to solve problems, where the finish line keeps moving, and the tools have started running alongside us?
In the film, we meet prodigies burdened by expectation. We meet runners who have lived their whole lives in second place. We meet those who run simply to outpace their fear. Ultimately, all of them are bound by the fact that they run, but what do they run for?
It's a question I find myself asking in a different arena: the daily work of writing code, and what it means now that AI has entered the race.
People reflect upon the goal for their lives often and use it as a compass to guide them into action and work on their priorities. While some may achieve what they set out for, the feeling is ephemeral. Humans long for order, resolution, and meaning. All these together create a system, one that brings calmness and comfort. But does this system enable you to progress, or does it make you stagnate?
Camus explored this in The Myth of Sisyphus. We strive for order in a world that offers none. Sisyphus, condemned to roll a boulder uphill only to watch it tumble back down for eternity, represents the struggles of human life: the desire to achieve and reap reward. But even when we reach the summit, satisfaction slips away. It feels like a butterfly that glides through flowers and leaves, leaving only a trace of its presence before time carries it off.
Entropy is omnipresent. All work is like this, meant to be completed, then fade into oblivion as new things rise and old trends fall. As an engineer, I build things that grow and evolve over time. But there's a cost: the constant cycle of redesigning, refactoring, rebuilding.
A system ships, stabilizes, and then requirements shift. New frameworks emerge. The codebase that felt elegant six months ago now feels like technical debt. You tear it down. You build again.
Camus called this the absurd: the gap between our longing for permanence and a world that offers none. But he didn't see tragedy in Sisyphus. He saw defiance. The push is the meaning.
There's a particular feeling when you're deep in a codebase, and the whole system lives inside you. You can trace how a change in one module ripples through three others. It feels like you're walking through it like a house you've lived in for years. That's when the real work happens. Not when you're consulting documentation, but when the problem and the solution exist together in your head, rotating slowly, waiting for the piece that fits.
AI can extend this. When you're already holding the context, a model becomes useful: suggesting patterns you half-remembered, catching edge cases at the periphery, holding adjacent concerns while you focus on the core. You're still the one thinking. It's a sharper tool, not a replacement for your hands.
But I keep watching something else happen.
Code appears. It looks right. It runs. The diff gets approved. No one held it in their head. No one walked through the house; they glanced at a photograph and called it home.
This feels productive. It feels like progress. But it's the same sensation as scrolling through short-form content, video after video, sometimes informative, always moving. You finish an hour later, thinking you've learned something. But have you? Could you reconstruct any of it? Or did it just wash over you?
Vibe coding has the same texture. You prompt, you receive, you ship. The output exists, but the understanding doesn't. The code is in the repo, not in your head.
I see this quite often now. A paper or a new model drops, and everyone scrambles to implement/utilize it. A few weeks later(well, at this point, never know if it is hardly days or hours :/) it's a baseline, superseded by something else. The model you fine-tuned last quarter is already legacy and can be improved. You read the next paper. You rebuild.
The rhythm is relentless, and if you're not careful, you start skimming: skimming code, skimming papers, skimming your own understanding. While one can say that it has always been like this, the pace at which this happened was not something we expected, and some still don't, and are coming to terms with it.
And this matters because code being cheap doesn't mean software quality improves. The rate at which we build will increase, and expectations will skyrocket. But who digs in? Who takes accountability for what the agent wrote? Or do we just approve and move to the next ticket, one merge after another, feeling busy, feeling fast, never stopping to ask if we actually know what we just built?
So if the understanding isn't there, and the work is never truly done, what's left?
Here's what I keep returning to: traditional crafts have endings. A novel is finished. A painting is signed. A sculpture is cast. You get to say done.
I might be wrong, but software doesn't work like this. There is no final form. You ship, and then requirements shift. You refactor. You redesign. You rebuild. The elegant system becomes legacy. You tear it down. You start again. There are no finish lines, only checkpoints. Constant sprints, punctuated by walks and jogs as energy waxes and wanes.
This is why the athletes resonate. They don't run to stop running. The race ends, and they line up for the next one. They run to run.
Software is the same. You don't build to be done. You build to keep building. The goal isn't completion; it's continuation. Flourishing not through reaching a destination, but through the excellence of the activity itself. And if you can't find meaning in that, in the process itself, the refactoring, the debugging, the slow daily shaping of something that will never be finished, then the work becomes unbearable.
Camus would recognise this. Sisyphus doesn't push the boulder to reach the summit. He pushes because pushing is what he does. The struggle itself, fully owned, is enough.
Getting to say "done" is not the goal. The goal is to keep going.
So now, what are you running for?
Notes & further reading:
- Paul Graham, Holding a Program in One's Head.
- Mihaly Csikszentmihalyi, Flow: The Psychology of Optimal Experience.
- Albert Camus, The Myth of Sisyphus.
- Aristotle Nicomachean Ethics.
- 100 Meters (Hyakuemu), Rock 'n' Roll Mountain.