We optimized for speed. We lost the room where ideas happen.
Someone on your team had an idea last week.
You probably missed it. It came out sideways, the way real ideas do. Not as a polished proposal or a Jira ticket or a Confluence page with acceptance criteria. It came as a half-sentence during a standup, or a Slack message that started with "this is probably dumb, but..." and trailed off when nobody reacted fast enough.
The idea died in the silence after the message was sent. The person who said it waited. Nobody responded. They moved on.
This keeps happening. Not as a single dramatic moment anyone would point to, but as a slow leak. A thing we used to have that we don't anymore, and most of us can't even name what it was.
I can. It was the moment someone looked you in the eye and said, "What if we just tried this?"
The Room That Used to Exist
There was a version of building software that felt different. Not better, exactly. Rawer.
You'd be stuck on something, really stuck, and instead of filing a ticket you'd say it out loud. And someone nearby who had no business caring about your problem would get curious anyway. They'd walk over, or ping you, or just start talking through it with you. The conversation would start with "I don't know if this makes sense" and end with both of you hunched over a screen, trying something neither of you was sure would work.
The key word is trying. Not planning. Not estimating. Not writing a ticket for someone else to prioritize. Trying. Right now. Because the idea was alive, and the only way to know if it was good was to see what happened when you let it loose.
There was room for mistakes in that version of building things. Room for experiments that went nowhere. Room for someone to say "fuck it, let's hack it" and for the room to say yes, not because the idea was sound, but because the energy behind it was worth honoring.
That room is gone now, and I'm not sure we noticed when we lost it.
What Replaced It
We replaced it with something faster.
Sprints. Standups. Velocity charts. Quarterly OKRs that cascade into monthly goals that cascade into weekly tickets that cascade into daily standups where everyone performs the same ritual: "Yesterday I did X, today I'm doing Y, no blockers." The machine runs. Features ship. The dashboards are green.
Everyone already knows this, but Microsoft quantified it anyway: the further a developer's actual workweek drifts from their ideal one, the worse everything gets. Satisfaction, productivity, all of it, sliding in the same direction.
Sixty-nine percent of developers lose eight hours or more per week to inefficiencies. A full working day, every week, gone to tech debt, bad documentation, and process overhead. Not building. Not exploring. Not wondering "what if."
Learning. The thing that keeps you curious. The thing that makes you look at a problem and think "I wonder what would happen if..." instead of "which ticket is next."
We know this. We talk about it at conferences and in blog posts with titles like "Reclaiming Developer Joy." Then we go back to the sprint planning meeting.
The Thing We Actually Lost
But here's what bothers me, and it's not the productivity numbers.
When I think about the best moments of my career, not one of them was about shipping on time. Not one was about hitting a velocity target or closing a sprint with zero spillover. The moments that mattered were the ones where someone got excited about an idea and that excitement was contagious.
You know that moment. Someone's explaining something and they lean forward. Their voice changes. They start talking faster, or slower, or they stop talking entirely and start drawing on a whiteboard. And you feel it. Not the idea itself, but the aliveness behind it. The willingness to not know, to be wrong, to try something just to see.
That's what we lost. Not innovation. Not creativity. Not even experimentation, exactly. We lost the willingness to be present for someone else's excitement.
We rarely look into people's eyes anymore when they share an idea. We're in a Zoom grid, cameras optional, mentally rehearsing our own update. Or we're reading the Slack thread and composing a reply before the other person has finished typing. We hear the words but not the thing behind the words. We hear "what if we tried this?" but not "I care about this enough to risk being wrong in front of you."
That second part is the part that matters. And it needs something we've optimized out of existence: slack. The unscheduled hour. The empty whiteboard. The conversation that starts about one problem and ends up somewhere nobody expected. You can't ticket that. You can't retrospective it back into existence. It emerges from a set of conditions that we dismantled in the name of efficiency, and now we act surprised that nobody proposes anything wild anymore.
Speed Filled the Silence
AI tools widened the gap, though not in the way most people think.
The conversation around AI and engineering is usually about code quality, or job displacement, or productivity gains. Those are real concerns. But the subtler thing AI did was fill the silence.
The silence was where ideas lived. The ten minutes you spent staring at a function, not sure how to approach it, letting your mind wander into adjacent possibilities. The twenty-minute debugging session that accidentally taught you how a system actually worked, not how the documentation said it worked. The friction that slowed you down enough to notice things.
When you can generate a working solution in thirty seconds, you don't wander. You don't sit with the discomfort of not knowing. You don't call someone over to think about it with you, because there's nothing to think about together anymore. The answer is already there. All that's left is to evaluate it.
Evaluation is solitary. Wondering is communal.
We used to wonder together. Two people at a desk, neither one sure what would work, building something just to see. That's the version of engineering that nobody writes job descriptions for, and nobody measures, and nobody puts in the quarterly review deck. But that's the version that produced the ideas worth having.
What Excitement Sounds Like When Nobody's Listening
You've seen this happen. Someone on your team mentions a side project they've been noodling on. Something involving a weird API edge case, an unexpected behavior they stumbled across. For a moment, they're animated. Talking faster. Leaning in.
Then they catch themselves. "Sorry. Nobody cares about this stuff anymore."
They're not being dramatic. They're describing a reality. The things that excite engineers — the weird corners of a system, the unexpected behavior that makes you want to dig deeper — those things have no home in a velocity-optimized workflow. They're not on the roadmap. They're not tied to a business metric. They're not even a story you can tell in standup without sounding like you're wasting time.
So the excitement goes underground. It becomes a private thing, something you indulge in after hours if you have the energy. And gradually, it fades. Not because people stop being curious, but because curiosity without an audience becomes hard to sustain.
There's a reason we used to build things together in rooms. Not just because collaboration is efficient, but because excitement is a social emotion. It needs a witness. Someone to say "wait, what?" Someone to lean in. Someone to say "I don't know if that works, but let's find out."
Evaluation is solitary. Wondering is communal.
When was the last time someone on your team wondered out loud?
The Lament Isn't the Answer
I don't have a fix for this. This isn't an article that ends with "here are five ways to bring experimentation back to your team." That kind of prescription is part of the problem. We've been solving human things with process for so long that we forgot some human things can't be solved. They can only be given room.
What I know is that the people I've worked with who were worth working with, the ones who made me better at this job, were the ones who said "let's just try it" and meant it. Not as a process improvement. Not as an innovation framework. As a human being who was excited about something and trusted the room enough to say so.
That trust is the casualty. Not the experimentation itself, but the environment that made experimentation feel safe. The room where you could be wrong without it going into a tracking system. The room where someone's half-baked idea got taken seriously not because it was good, but because the person behind it was worth listening to.
We built something faster. I'm not convinced we built something better.
And somewhere on your team, right now, someone has an idea they're not going to share. Because they've learned that the room isn't listening.