It Is Time to Boil the Ocean
AI is not competition. It is rocket fuel. The question is what we finally aim at.
For twenty years, the advice was always the same: don’t boil the ocean. Scope it down. Ship the minimum. Stay realistic. And for twenty years, it was mostly right. Execution was expensive. Engineering hours were scarce. Ambition without resources was just daydreaming with a backlog.
The cost of execution just dropped by an order of magnitude. Which means the advice just inverted.
The dangerous move is no longer thinking too big. It is thinking too small. And yet here we are, most of us, applying the scarcity reflex to a moment of abundance: trimming budgets, “optimizing headcount,” scoping down to fit a format that no longer exists. It is, if I may say so plainly, exactly backwards.
AI is not competition. It is rocket fuel. The question is not what it takes from us. It is what we can finally attempt.
In “The Kitchen Is Open,” I argued that AI democratized capability and taste became the scarce resource. This is the sequel: now that we are in the kitchen, what do we cook? Something safe that fits the old menu? Or the thing we have always wanted to make, the one we shelved because we could never justify the ingredients?
The Scarcity Playbook
Here is how the scarcity reflex works in an organization. A new capability appears. Leadership convenes a task force. The task force identifies ways to reduce costs by ten percent. The board approves the reduction. Everyone celebrates the “efficiency gains.” Meanwhile, somewhere else, someone uses the exact same capability to build something that was previously impossible.
The first organization saved money. The second created a category.
This pattern is so old it has a name. In 1865, William Stanley Jevons observed that as England’s steam engines became more efficient, coal consumption did not decrease. It exploded. Cheaper energy per unit meant people invented entirely new uses for energy. New factories. New railways. New industries that had not existed when coal was expensive.
Economists call this the Jevons Paradox. The plain version is simpler: when something gets cheaper, we do not use less of it. We dream bigger. Steel got cheap and we built skyscrapers. Computing got cheap and we invented the internet. Bandwidth got cheap and we got video calls, streaming, and an entire generation working from their kitchens. Every time humanity reduced the cost of execution, the response was not restraint. It was ambition.
The cost of building software just collapsed. And we are sitting in meetings talking about headcount.
The Reflex We Trained Into Ourselves
Nobody chose the scarcity reflex. It was trained into us, one sprint at a time.
We learned to scope down because resources were genuinely scarce. Engineering hours were expensive. Shipping was slow. The rational thing was to cut features, reduce scope, and focus on what was achievable within the quarter. Over years, that discipline became instinct. We stopped asking “what should we build?” and started asking “what can we get away with building?” The ceiling became the goal.
There is something quietly cruel about this. The people who are best at scoping down, the disciplined ones, the ones who reliably ship, are often the ones with the largest ambitions locked away. They did not lose the ideas. They learned to keep them quiet, because the economics never justified saying them out loud. You do not pitch the cathedral when the budget covers a shed.
The economics just changed. The budget covers more than a shed now. Possibly more than a cathedral. And the instinct trained by twenty years of scarcity is still whispering: “Stay realistic. Do not overcommit. Ship the minimum.”
That whisper used to be wisdom. Today it is a cage.
What the Evidence Actually Shows
The organizations paying attention are not cutting. They are expanding.
Shopify grew revenue thirty percent in 2025, driven in part by AI tools that opened entirely new commerce channels for their merchants. They did not use AI to cut costs. They used it to expand into markets they could not previously reach. Closer to the metal, small teams are shipping features in days that used to take departments months, because the cost of turning ideas into working software dropped by an order of magnitude. The pattern is consistent: organizations that treat AI as abundance find more to do, not less.
This is Jevons, playing out in real time.
When execution gets cheaper, the bottleneck shifts. It moves from “can we build it?” to “should we build it?” and then quickly to “what else becomes possible now that this is easy?” The organizations that follow that thread discover demand they did not know existed. Problems they had written off as unsolvable become tractable. Markets that were too expensive to serve become reachable. The scope does not shrink. It multiplies.
The scarcity reflex says: same work, fewer resources. The Jevons response says: new work, same resources, ten times the output. One of these is a defensive posture. The other is a growth strategy. They are not equally rational.
Montreal built a world-class AI ecosystem from a mid-sized, French-speaking city in a cold northern country. That was not scarcity thinking. That was Bengio and a handful of researchers deciding the ocean was theirs to boil, long before anyone gave them permission. The lesson was never “be realistic.” The lesson was: pick the right ocean.
Go Look at the Shelf
Here is what I want to say plainly, because I think most of us need to hear it.
There is a shelf. Every builder has one. It holds the ideas that were too expensive, too slow, too ambitious for one person or one small team. The product that would require a team of twenty. The internal tool that nobody would fund. The side project that would take two years of evenings. The thing that makes you think “that would be incredible, but I could not possibly…”
The “but I could not possibly” is the part that just changed.
AI repriced half of what is on that shelf. Not all of it. Some ideas are hard for reasons that have nothing to do with execution cost. But the ones that were shelved because of engineering hours, because of sheer building effort, because one person could not do what required ten? Those are live again.
The fear that runs deeper than job security surfaces at 2am, after watching a demo that made our stomachs drop. It is not about org charts. It is this: what if I do not have anything on the shelf? What if I have spent so long shipping minimums that I have forgotten what my own ambitions sound like?
I know that fear. I suspect most of us do. And I want to say something about it directly: it is the fear itself that is the scarcity. Not the absence of ideas. The discipline of keeping them quiet. The person who worries they have nothing large inside them is almost always the person who has been disciplined enough to keep something large shelved.
The shelf is not empty. It is heavy. And the economics just shifted under it.
The Ocean Is Already Boiling
Here is the deepest paradox, and it is a kind one: the fear of AI is largest precisely where the ambition is smallest.
If all we bring to the table is efficiency, then a more efficient tool will always feel like a threat. But if what we bring is vision, taste, and the hard-won judgment of what is worth building and why, then the tool is simply a better instrument. Scarcity sees a power tool and worries about the carpenter. Abundance sees a foundation being poured and starts drawing blueprints.
The great lie of “be realistic” was always that it sounded like wisdom. “Don’t boil the ocean” was its strategic cousin. Both were, more often than not, the world asking us to be smaller than we are. And we obliged, because we are responsible people, and responsible people do not pitch the thing that does not fit the budget.
The budget just changed. The cost of execution collapsed, and the collapse is not slowing down. Somewhere right now, someone with taste and ambition and a clear sense of what matters is picking up an idea that used to require a department and building it over a weekend. That person is not reckless. That person simply noticed that the economics changed before the reflex did.
Go look at the shelf. Pick the idea that scares you. Say it out loud. “I am building something that does not fit the old budget.” That is vulnerable. It is also the only move that matches the moment.
The ocean is boiling whether we light the fire or not. The only question is whether we are cooking something worthy of the heat.
It is time to boil the ocean.
What if instead of seeing AI as a nightmare, we saw it as a means to unlock our dreams?
That is the paradigm shift. Not “how many of us will they still need?” but “what can we finally build?”
We all shelved ideas that were too ambitious, too expensive, too unrealistic. Not because they were bad ideas. Because the economics did not justify them. The economics just changed.
This is the first in a series. The paradigm shift is the easy part. The hard part is the skills: knowing what to ask for, judging whether the output is good, decomposing problems into pieces agents can carry, and managing a team that works at 3am and has no sense of direction. One article per skill, starting next week. The hitchhiker’s guide to the new job market.
