When everyone can generate a thousand images in an afternoon, the image itself becomes worthless. When language models produce passable copy at scale, the copy is no longer the asset. The scarcity has shifted. What matters now is not the ability to make things—it’s the ability to know which things matter, why they matter, and to whom.
For decades, production was the bottleneck. Photographers, writers, designers, strategists controlled supply through specialized skill and costly tools. The market rewarded scarcity of output. The bottleneck has now moved upstream. Production is abundant and cheap. What becomes scarce is judgment.
This inversion is already visible in how sophisticated operators deploy AI. They are not using it to replace taste-makers. They are using it to amplify them. A creative director at a major fashion house doesn’t use generative AI to eliminate vision—she uses it to iterate faster on a vision only she can articulate. A publisher doesn’t use language models to write editorial thesis; he uses them to execute against a thesis only he can author. A founder doesn’t ask AI to decide product strategy; he uses it to pressure-test strategy, surface patterns, and accelerate decision cycles around choices he alone can make.
The Commoditization Layer
What AI actually does is commoditize execution. It turns labor-intensive, skill-dependent work into a utility. This is economically healthy—it reduces friction and democratizes access. But it also means any work that can be fully automated or fully specified is no longer a source of competitive advantage.
A mid-market agency used to compete partly on its ability to produce high-quality work quickly. Now that ability is table stakes. The survivor is the agency with taste—the ability to understand what the client’s audience actually needs, what signal they’re trying to send, and what moves the needle. That judgment cannot be outsourced to a model.
Consider content strategy. Five years ago, media companies competed on volume and speed. Now volume is trivial. A single person with the right prompts generates more words than a newsroom. The publications with genuine economic moat are those with editorial perspective: a point of view about what matters and to whom. Bloomberg survives because it has judgment about capital. The Atlantic survives because it has judgment about ideas and culture. Generic content mills have no moat because volume is worthless.
This extends across knowledge work. The consultant who can just analyze data and hand off findings now competes against AI that does the same thing faster. The consultant who survives is the one who says: “Here is what the data means for your business. Here is the decision I recommend, and why.” That synthesis, that judgment call, that willingness to be wrong—cannot be delegated to a language model.
Taste as Operational Foundation
In a world of AI abundance, taste becomes foundational infrastructure. It is the layer all other work builds on.
The operator with taste compounds advantage quickly. She specifies what she wants more precisely because she knows what good looks like. She evaluates output faster because she has standards. She identifies what’s missing in ways algorithmic thinking cannot, because taste is fundamentally about understanding context, history, audience psychology, and cultural signal.
This explains why luxury brands invest heavily in creative leadership even as production becomes cheaper. It explains why venture capital consolidates around partners with taste for markets and founders, not just data analysis. It explains why the most successful founders hold strong points of view about culture, design, and user experience—not just capital or technical skill.
Taste is the ability to make bets on what matters before the market confirms it. It requires pattern recognition, cultural fluency, historical knowledge, and confidence. It cannot be learned from a dataset. It comes from exposure, discernment, and often from failure—living through the consequences of your judgments.
The Return of Institutional Curation
We will likely see a return to institutional curation—but for different reasons than the gatekeeping of the past. In the age of scarcity, gatekeeping controlled access to expensive production. In the age of abundance, curation filters signal from noise. The human editor, the tastemaker who says “this matters and this doesn’t”—becomes more valuable, not less.
This has real implications. It concentrates platform and influence around those positioned as arbiters of taste. It creates new hierarchies and rewards those who have already established credibility and audiences. It makes it harder for genuinely new voices to break through—they lack the platform to assert their taste as authoritative. This is the opposite of the decentralization narrative that typically surrounds new technology.
But it also creates opportunity. If taste is valuable, developing taste becomes a sound investment. Young operators should spend less time optimizing production skills—AI will handle that—and more time developing judgment. This means reading history, studying culture, consuming across mediums, traveling, building taste through exposure and reflection. Cultivate relationships with people who have sharp judgment and learn how they think.
The Strategic Shift
For established firms: Is your competitive advantage actually taste, or was it execution? If execution, you face structural headwinds. If taste—genuine editorial vision, product intuition, market judgment—then AI is a tool that accelerates what you were already good at.
For individuals, the shift is starker. The person whose value comes from being able to execute a technical skill—write code, design graphics, produce copy—faces commoditization. The person whose value comes from knowing which code matters, which design direction to pursue, which story deserves telling has options. That person becomes more valuable with better tools to execute their judgment.
Production skill does not become irrelevant. Understanding how language models work, how image generation functions, what the constraints are—this is important context for making good judgments about output. But the person who can only make things, not judge things, faces structural headwinds.
The economic signal is already visible. Salaries and equity flowing to creative directors, editors, product strategists, and operators with vision are rising. Salaries for pure execution roles are flattening or declining. Capital flows toward people and firms with taste, toward those who can make confident bets on what matters next.
The AI revolution is not eliminating the need for human judgment. It is making human judgment the only thing that cannot be automated. Taste, discernment, and the ability to make good calls under uncertainty are no longer luxury skills. They are foundational infrastructure for anyone who wants to create value in an AI-augmented world.
