There is a story we tell about difficulty.
That it is the price of an outcome. The unpleasant tax on the result you actually want. The thing to be minimized, optimized away, reduced as much as possible so that the output can arrive as efficiently as the tools allow. Difficulty is the problem. The solution is whatever reduces it.
This story is so culturally pervasive that it feels like common sense. It is built into every productivity system, every efficiency tool, every piece of software designed to make something that was hard easier. Including AI. Especially AI. The entire value proposition of the most powerful thinking tools ever built is precisely this, the difficult thing, done for you. The struggle, eliminated.
And for many tasks this is exactly right. The struggle of manually calculating a spreadsheet is not a curriculum. It is friction between you and the actual work. Eliminating it is genuine progress. The tool serves the outcome without costing anything that matters.
But there is another category of difficulty that the story gets completely wrong. A category where the struggle is not the price of the outcome but the substance of it. Where what you’re trying to produce is not an output but a thinker. Where the difficulty is not friction between you and the result but the process by which the result, the genuine result, the one that matters, gets built.
In this category eliminating the struggle doesn’t deliver the outcome more efficiently. It eliminates the outcome entirely while producing something that looks like it.
The output arrives. The thinker doesn’t develop. The answer exists. The understanding doesn’t. The essay is written. The thinking it was supposed to build isn’t.
This category is larger than most people realize. It includes most of what matters about education, about professional development, about the building of judgment, about the formation of the examined mind. It includes most of what the previous twelve videos of this series have been pointing toward.
And in this category the story we tell about difficulty is not just wrong. It is producing the specific harm it presents itself as preventing.
Why does struggle produce what smoothness cannot?
The answer is in the biology of how background gets built. And it connects everything this series has been building to the Flores and Maturana foundation that the Computers and Cognition series established.
The nervous system, Maturana’s operationally closed biological system, doesn’t develop through reception. It develops through perturbation and recovery. Through encountering something that doesn’t fit the existing structure and being required to reorganize in order to maintain coherence. The reorganization is the development. The perturbation that required it is what produced the development.
This is what struggle is, biologically. It is the perturbation that requires structural reorganization. The encounter with something that doesn’t yield to existing patterns and therefore forces the development of new ones. The productive resistance that deposits new background in the system that successfully navigates it.
When the struggle is eliminated, when the AI resolves the perturbation before it can require structural reorganization, the development doesn’t happen. Not because you chose not to develop. Because the condition that produces development was removed before it could do its work.
The developer who debugs code they wrote develops something the developer who generates and accepts AI code does not. Not because debugging is more virtuous. Because debugging is a perturbation, a structural encounter with a system that isn’t doing what was intended, that requires the developer to develop a new background in order to resolve. The AI-generated code that works doesn’t perturb. It resolves. And resolution without perturbation produces output without development.
This applies across every domain where genuine knowing matters.
The writer who struggles with an argument, who drafts, discards, approaches differently, finds the actual structure of the thinking through the resistance of trying to express it, develops something the writer who prompts for an essay and edits the result does not. Not the essay. The thinking. The background that makes the next piece of writing more genuinely the writer’s than the last one.
The student who wrestles with a difficult concept, who holds not-understanding long enough for understanding to arrive through genuine cognitive work, develops something the student who searches for the explanation and receives it fluently does not. Not the knowledge. The structure of knowing. The background that makes the next difficult concept more navigable because the previous one was genuinely worked through.
The professional who thinks through a complex decision without immediately outsourcing the analysis, who sits with the competing considerations, follows the implications, examines the assumptions, develops something the professional who prompts for a decision framework and applies it does not. Not the decision. The judgment. The developed capacity to navigate complexity that only develops through genuine navigation of complexity.
In each case the struggle was the curriculum. Not the price of the outcome. The substance of the outcome that actually mattered.
And here is what makes this the most urgent argument in the series.
The outcomes that can be outsourced, the outputs, the information, the analysis, the frameworks, are also the outcomes most subject to automation. The things AI can produce for you are the things that will increasingly not require you to produce them. Their value as outcomes is declining.
The outcomes that cannot be outsourced, the judgment, the background, the knowing that develops through genuine struggle, are the outcomes whose value is increasing precisely because they cannot be automated. The thinker who has developed genuine knowing through accumulated productive struggle is becoming more valuable, not less, in a world where fluent output is infinitely available.
The struggle that everyone is trying to eliminate is the thing that is becoming most worth having.
This is not an argument for doing everything the hard way. It is a distinction between two categories of difficulty that require completely different responses.
The first category is friction, difficulty that sits between you and the actual outcome without contributing to its production. Manual calculation when the math is understood. Formatting when the thinking is done. Information retrieval when the information is needed and the understanding to use it is already present. This difficulty should be eliminated. AI is extraordinary at eliminating it. Use it freely, use it well, use it without guilt.
The second category is curriculum, difficulty that is the production process for the outcome that actually matters. The struggle of genuine thinking. The perturbation of encountering something that doesn’t yield to existing patterns. The productive not-knowing that reorganizes the structure of your knowing when you stay in it long enough. This difficulty should be protected. Not fetishized. Not sought for its own sake. But recognized as the substance of the development you’re trying to produce and protected accordingly.
The examined thinker knows which category they’re in. And that knowledge comes from the diagnostic developed in the previous video, the felt difference between genuine thinking and its substitutes. Resistance, moving uncertainty, surprise, changed relationship. When those qualities are present the difficulty is the curriculum. When they’re absent and ease has replaced them, the difficulty that was curriculum, has been replaced by consumption.
Protect the curriculum. Eliminate the friction. The distinction between them is the most practically important thinking skill available in the current moment.
Here is the practice that makes the distinction concrete.
Before outsourcing any cognitive task to AI ask one question, is the struggle here friction or curriculum? Is the difficulty sitting between me and an outcome I understand how to produce, or is the difficulty the process by which I would develop the capacity to produce it?
If friction, use the tool freely. Let it eliminate what doesn’t need to be hard.
If curriculum, stay in the struggle. Let the perturbation do its work. Let the not-knowing remain unresolved long enough for genuine knowing to develop. Use AI afterward, to extend, to examine, to stress-test what the struggle produced, but not instead of the struggle that produces it.
This is how the examined thinker inhabits the age of AI. Not by rejecting the tool. Not by fetishizing difficulty as virtue. By knowing the difference between the difficulty that serves and the difficulty that is the service.
The struggle is the point. Not always. Not everywhere.
But in the territories where genuine knowing matters, where judgment is the outcome, where background is what’s being built, where the thinker is what’s being formed, the struggle is not the obstacle between you and what you’re trying to produce.
It is what you’re trying to produce.
Every time you stay in it, every time you resist the oracle, follow the question before the question, find the assumption underneath, use the tool as Socratic partner rather than answer machine, you are building something that compounds.
The background that is yours. The judgment that cannot be prompted. The knowing that lives in the structure of a thinker who was present for every perturbation that produced it.
That is not a productivity metric. It is not measurable in outputs per hour. It does not show up on a dashboard.
It shows up in the quality of your thinking when it matters most. In the judgment that navigates the edge case the pattern didn’t cover. In the understanding that survives the breakdown the fluent output didn’t anticipate. In the thinker you are becoming through every struggle you chose not to eliminate.
The tool is extraordinary. The thinker who uses it well is more extraordinary still.
Be that thinker.
The struggle is how.
@ Lee Greene