Out of Ideas

For twenty years the engineer was the bottleneck, and everyone above him arranged their entire working life around the wait. The product manager wrote the ticket and waited. The founder pitched the roadmap and waited. The delay was load-bearing — it was where all the important people got to look busy and blameless, because the thing simply took time, and the taking of time excused everyone standing near it. Then the wait dropped to zero, the music stopped, and it turned out there were not enough chairs. More precisely: not enough ideas.

The feature that used to eat a sprint now takes ten seconds. This isn't hyperbole; it's a scheduling problem. A developer can ship the thing faster than a product manager can open Jira, find the "New Issue" button, pick a priority he doesn't really understand, and paste an acceptance criterion cribbed from the last one. The queue drains faster than it fills. And a queue that drains faster than it fills has one catastrophic property, from the standpoint of the people whose whole job was to fill it: it hits empty, and then everyone can see the bottom of it.

Because the backlog was never merely a list of work. It was a hiding place. As long as there were forty tickets and two quarters of runway, nobody had to ask whether the forty tickets were any good — whether a single human being on the planet wanted the thing they described. The scarcity of engineering hid the poverty of the ideas. Now engineering is nearly free, the board clears by Thursday afternoon, and the PM is standing in front of an empty column being asked the exact question the entire apparatus was built to postpone: what should we build — and, quieter, underneath it, does any of this matter to anyone at all.

And here is the sentence no one will say into a microphone: a great many product managers cannot answer that. They were never the visionaries the org chart implied. They were a translation layer — a cushion between a leadership that didn't know what it wanted and an engineering team that couldn't be addressed in plain speech. Remove the translation, let the machine take plain English directly, and the layer has no remaining function. What's left is a midwit with a Figma license and a calendar full of "syncs," suddenly asked to produce, on demand, the one thing he was never actually hired for: a true idea about what a person might want.

Look up the chain and it is the same emptiness in a better suit. The founder, the CEO, the VP of Product — the tide went out on all of them at once, and the water was technical difficulty, and it had been covering everyone. For years "it's hard to build" was the universal alibi: it explained the flat revenue, the launches that slipped, the roadmap that never quite arrived. Delete "hard to build" and the naked question just stands there in the daylight — if building is instant and the company still isn't winning, then the problem was never the engineering. The problem was the idea, which means the problem was them.

And the diagnosis, once you're finally allowed to see it, is unglamorous and everywhere: most of these teams are sprinting to build things nobody wants, and always were. It merely used to take eighteen months and a Series B to discover it, and now it takes a weekend — which is far too fast, because the eighteen months was the product. It was the runway of plausible deniability during which everyone drew a salary and got to feel like a builder. Speed didn't make them better. It just moved the verdict up, from "someday" to "this afternoon," and the verdict is not kind.

Beneath all of this sits the rest of the org, watching with perfect clarity and saying precisely nothing. The engineer who now ships in ten seconds knows the backlog is thin and the ideas are worse. He can see that the emperor and the emperor's product manager are both standing there undressed. And he keeps his mouth shut and presses 1, presses 2, approves, ships, nods — because the person who names the emptiness out loud is the person who volunteers to be the first line item cut. Silence is the rational move when you can see the layoffs forming on the horizon and would much prefer to make the second wave than the first. So he looks away, professionally, and lets the pretense run its course.

Because the whole performance has exactly one audience that counts, and it was never the customer. It is the investor. The hush at the top is not confusion — it is choreography. Leadership worked out, faster than they will ever admit, that the day the market learns the product is aimless, the pipeline is thin, and the moat evaporated along with the cost of code, the money reprices overnight. So the quarterly deck stays bullish, the all-hands stays "energized," the phrase "AI-native" gets swung over the whole thing like incense over a body, and everyone in the building — the clueless PM, the exposed CEO, the silent engineer with his finger on approve — is quietly conscripted into a single shared project: keep the investors from finding out what every person inside the company already knows.

So the org turns itself inside out. The people who spent careers being waited on now have nothing to hand down, and the people who spent careers waiting can suddenly produce anything — except the one thing that was never theirs to supply. The bottleneck didn't vanish. It moved up, into the precise seats that were always insulated from ever being measured, and it sat down in the corner office. For the first time the slowest thing in the company is the idea, the idea was always their job, and there is nowhere left to park the delay. The tide has gone all the way out. They're still smiling at the shareholders and pointing bravely at the horizon — but look at the board behind them. It's empty. It clears by lunch now. And everyone on the floor has already read what it doesn't say.