The Kitchen Is Open

The kitchen just opened. The question is not who gets cut. It is what we finally get to cook.

Are we asking the wrong question?

In quarterly reviews and team standups alike, the arithmetic of elimination is running. “If AI can write code, how many of us do we actually need?” It is, like most reasonable questions, a question that has carefully reasoned itself into a ditch.

I understand the fear. It is not irrational. When you watch a four-person team at Anthropic ship a major product feature in ten days, something tightens in your chest. When you see demos of agents writing, testing, and deploying code while your team is still in sprint planning, the math feels obvious. Optimize headcount. Rationalize the org. Improve operational efficiency.

But the math is wrong. Not because AI is overhyped. Because we are solving the wrong equation entirely.

I wish this transition had not come so fast. So does every senior engineer I know. But wishing does not slow it, and the only question that matters is what you do with the time you have been given.

The Kitchen That Was

For twenty years, the technology industry operated like Gusteau’s kitchen before the rat arrived. Execution was expensive. Writing software required years of training, specialized knowledge, and a tolerance for complexity that most humans do not possess. This scarcity created a priesthood. If you could write production code, you were in. The kitchen was gatekept, and the gatekeeping felt like merit, because it mostly was.

You built your career in that kitchen. You learned the frameworks. You suffered through the debugging. You earned the quiet, unglamorous wisdom that separates someone who has shipped software from someone who has merely studied it. Every production incident. Every late night untangling someone else’s optimistic architecture. Every decision you made correctly only because you had once made it wrong. That was the long apprenticeship, and you paid it willingly, because the work was good, and the kitchen rewarded those who stayed.

Chef Skinner ran the kitchen on this logic. Only credentialed chefs cook. Only trained hands touch the stove. The hierarchy exists to protect quality. And for a long time, it did.

But the world has a way of changing around the people who built the walls. And the walls, which felt permanent, were never the point. The cooking was the point.

Gusteau Was Right

Then the constraint moved. There is a concept in economics called Jevons’ Paradox. William Stanley Jevons observed in 1865 that when steam engines became more efficient, coal consumption did not drop. It exploded. Efficiency did not reduce demand. It unleashed it.

This pattern repeats with a persistence that should, by now, be familiar. Steel got cheap: we did not build the same number of buildings for less money. We built skyscrapers. Computing got cheap: we did not do the same tasks faster. We invented the internet, then mobile, then cloud. Every time humanity made execution cheaper, the response was not restraint. It was ambition.

Execution just got cheap again. Anthropic shipped a major product in ten days with four people. Whoop is hiring 600 people, nearly doubling their workforce, while going all-in on AI. The cost of turning an idea into working software dropped by an order of magnitude.

And every organization is staring at this and asking Skinner’s question. Guarding the kitchen door, defending it with great courage and complete irrelevance, while the real crisis is that nobody in the dining room knows what dish to order.

What Is Actually Scarce

Skinner’s fear is the one we all share: if you open the kitchen, quality collapses. If anyone can cook, nobody cooks well.

This fear is understandable. It is also precisely backwards.

Here is what is actually scarce now: clarity (the ability to look at a problem and know what to build), ambition (the willingness to attempt what was previously too expensive to try), and distribution (knowing how to get a solution to the people who need it).

Taste. That is the word. Remy did not succeed because he could operate a stove. He succeeded because he had taste. He knew which flavours belonged together before he combined them. He could sense what the dish needed before the dish existed. AI has capability without judgment, speed without direction, a kitchen full of ingredients and no idea what the customer actually needs.

You have taste. Fifteen years of production incidents, architectural mistakes, and hard-won intuition gave it to you. And here is the thing about taste: it does not depreciate when the tools change. A master carpenter does not lose his eye for joinery when someone hands him a better saw. The tools moved. The knowing stayed.

Small advantages, held faithfully over time, are what change the course of things. You have been accumulating yours for years.

You Are Linguini (This Is a Compliment)

Linguini is not a great chef. His value is that he is human, standing in a human kitchen, navigating human politics, translating between a genius who cannot speak and a world that is not ready to listen.

Remy is brilliant. But Remy cannot talk to the health inspector. Remy cannot manage the kitchen staff. Remy cannot read the customer’s face and decide that tonight, this restaurant needs to serve something it has never served before.

You are the human in the loop. Not as a checkpoint. Not as a bottleneck. As the one who understands what the work is for. AI can build to specification. You are the one who knows whether the specification is worth building. That is not a small thing. In a world suddenly flooded with capability, it may be the largest thing of all.

What To Do Monday Morning

Stop treating your code as sacred. Code produced at near-zero cost should be treated as near-zero cost. Let it be rewritten. Let it be thrown away. The craft is in the choosing, not the typing.

Kill one meeting. Monday. Replace it with a working demo. The permission loop is costing you more than the thing you are asking permission for. Begin the thing, then show it. You will be forgiven for building too fast far sooner than for deliberating too long.

Build something you do not have credentials for. The translation layer that kept you in your lane is dissolving. The borders between disciplines are thinning. Cross them. The credential mattered when execution was expensive. Now execution is cheap, and the only credential that counts is the work itself.

Ship before you are ready. Feedback at 30% complete is worth ten times more than feedback at 90% complete. The instinct to polish before showing is a habit from an era when showing was expensive. That era is over.

Plan in weeks, not quarters. Over-planning is the new underperforming. The road ahead is not fully mapped, and no amount of planning will map it. Walk it. Adjust as you go.

Montreal Knows This Story

This city should not be a world capital of AI. By every rational metric, the kitchen should belong to someone else. Somewhere larger, richer, more credentialed. And yet: Bengio, Mila. A mid-sized, French-speaking city in a cold northern country, producing foundational research that reshaped an entire industry.

Montreal is proof of something the spreadsheets will never capture. Great things come from overlooked places. They always have. The center of the map is not where the important work happens. It is where the important work gets noticed. The work itself starts in the margins, among people who simply began.

The Critic’s Speech

At the end of the film, Anton Ego writes: “Not everyone can become a great artist, but a great artist can come from anywhere.”

Your career is not ending. It is being rewritten, and you are holding the pen. The skills that made you excellent (judgment, pattern recognition, the instinct for what works and the harder instinct for what does not) did not depreciate. They became the scarcest resource in the building. You were forged by the long apprenticeship. That forging holds.

There is a kind of courage that does not look like courage. It looks like a person sitting at a desk on a Monday morning, opening a blank file, and building something they are not yet sure they can build. It looks like starting before you are ready. It looks like trusting that the years were not wasted, that the hard-won things endure, that the kitchen rewards those who cook, not those who guard the door.

The kitchen is open. Everyone can cook. Not everyone will cook well. You know what the dish should taste like. You have always known.

Put down the knife. Pick up the menu. Start cooking.