The Curation Trap.
Exploring the editor’s delusion and why the act of selecting is not the act of thinking.
We are currently living through the greatest migration of human effort in history. It isn’t a migration of location, but of function. We are moving, en masse, from the role of Originators—those who wrestle with the blank page, the raw data, and the structural logic of a problem—to the role of Curators. We have become the masters of the “Executive Glance.”
In this new workflow, the heavy lifting is done by a black box, and our job is simply to review. We sit at the head of a digital table, watching options slide past us like a sushi conveyor belt. We tweak a word here, adjust a colour there, and hit “send.” We feel like creative directors. We feel like we have reached a new level of strategic leverage where we are finally “working on the business, not in it.”
But there is a haunting silence in this transition. By skipping the struggle of the first draft, we are quietly outsourcing our sovereignty. We are confusing the act of selecting with the act of thinking. And as the gap between the “Prompt” and the “Result” continues to shrink, so does the depth of our own professional intuition. We are becoming masters of the menu, but we are slowly losing the ability to cook.
The illusion of recognition vs. the weight of cognition
The genius of the curation trap is that it feels exactly like work. When you prompt a model and it generates four potential strategies, and you choose “Option C” because it feels the most “on brand,” you experience a genuine hit of dopamine. You made a decision. You exercised judgement. You feel productive.
However, there is a fundamental cognitive difference between recognition and cognition. Recognition is a passive process: it’s a multiple-choice test. You are comparing what you see against a pre-existing pattern in your head. It is low-energy and high-speed. Cognition, on the other hand, is active. It is the architectural work of building a thought from nothing. It requires you to hold conflicting ideas in your mind, to resolve tensions, and to find the “middle way” that doesn’t yet exist.
When we spend 90% of our day in recognition mode, our cognitive capacity begins to narrow. We stop looking for the fifth option, the one the algorithm couldn’t see because it wasn’t in the training data. We become trapped within the boundaries of the median. If you are only ever choosing from what is presented to you, you are no longer the pilot of your career; you are the passenger who thinks they are driving because they get to pick the playlist. This is how “average” becomes the new ceiling for excellence.
The atrophy of first principles: why we are accruing “cognitive debt”
The “struggle phase” of any project (the messy research, the three failed attempts, the circular logic that keeps you up at 2:00am, just to name a few) is often framed as an inefficiency to be optimised away. This is a catastrophic misunderstanding of how expertise is built. That struggle is not waste; it is the exact moment the knowledge moves from the screen and into your bones. It is where first principles are forged.
When we outsource the “building” and only keep the “editing,” we accrue what I call cognitive debt. Just like technical debt in software, we are taking a shortcut today that we will have to pay for with interest tomorrow. The interest is the loss of our “B/S Detector.” If you haven’t done the math yourself, you won’t know when the AI’s logic is 5% off-centre. You might catch the typos, but you won’t catch the structural rot.
We are becoming shadow experts. We can talk fluently about the output, but we can no longer explain the plumbing. This creates a fragile leadership layer: people who can curate a brilliant slide deck but lack the deep, intuitive understanding required to pivot when the underlying assumptions of their industry change. In the age of AI, the ultimate competitive advantage isn’t being a faster editor; it’s being the person who actually knows how the machine was built in the first place.
Reclaiming the originator’s edge: paying the “originator’s tax”
To survive the curation trap, we must intentionally reintroduce friction into our lives. We have to treat our minds like a muscle that requires resistance training. This isn’t about being anti-AI; it’s about ensuring that the tool serves the master, not the other way around. We must pay the “Originator’s Tax”—the deliberate choice to do things the hard way first to ensure our judgement remains calibrated.
The “draft zero” rule (protection of the core): Never open an AI tool until you have produced “draft zero” manually. This isn’t a polished draft; it is a bulleted, ugly, raw mess of your own associations, biases, and structural ideas. By defining the soul of the idea before the algorithm offers you a better version, you anchor the project in your own unique perspective. If you don’t start with your own bias, you will inevitably end with the algorithm’s average. You must own the architecture before the AI handles the paint.
The “active deconstruction” protocol (auditing the logic): When you do use a generative tool to assist your work, forbid yourself from simply tweaking adjectives. Instead, perform a forensic audit of the output. Ask the tool, “Why did you suggest this specific hierarchy of information?” or “What data are you prioritising in this conclusion?” Then, try to disprove it. If you cannot defend the structural logic of the work as if you had built it yourself, you are not curating; you are just a relay station for an algorithm.
The “first principles” deep-dive (calibrating the detector): Dedicate 20% of your week to manual cognition. Pick a core aspect of your role—something you’ve been outsourcing to tools—and do it entirely by hand. Read the raw 50-page PDF instead of the summary. Sketch the wireframe on paper instead of using a template. This is calibration time. It ensures that when you do go back to being a curator, your eye is sharp enough to spot the uncanny valley of logic that others miss.
The last mile of responsibility
In a world of infinite, automated curation, the Human Premium will not be found in how well you use a tool, but in how much of yourself you refuse to outsource. We are approaching a point where the correct answer is free and instant. As we have seen before, when certainty becomes a commodity, the value shifts to the quality of the question, but also the weight of the responsibility.
The most valuable people in 2026 won’t be the ones who can “prompt” the most efficiently; they will be the ones who still know how to build a thought from the ground up when the lights go out. Curation is a skill, but cognition is sovereignty. The curation trap is comfortable because it removes the pain of thinking, but that pain is exactly where your value lives. Don’t trade your sovereignty for a faster workflow. Stay in the struggle; it’s the only place where original ideas are born.