The Idiot Index of Higher Education

The image at the top of this essay is about rockets, but it is really about institutions.
The question is simple. If you break a rocket down into aluminum, titanium, copper, carbon fiber, steel, and other raw materials, what should the thing cost? Not what rockets have historically cost. Not what competitors charge. Not what the industry has learned to tolerate. What should it cost if the materials were already sitting on the floor and the only remaining problem was arranging the atoms correctly?
That gap is the Idiot Index. It is the distance between the raw ingredients and the finished product. If the finished product costs fifty or one hundred times more than the materials that compose it, you may have discovered sophistication. You may also have discovered waste.
Now apply the same logic to the university.
What are the raw materials of higher education? Motivated students. Expert faculty. A body of knowledge. Peer networks. Time. Institutional legitimacy. Access to tools. Access to employers. A campus or digital environment where people can practice, argue, revise, fail, and try again.
Those are powerful materials. In many cases, they are extraordinary. A good student walking into a good university is already a bundle of latent potential. Put that student near serious faculty, ambitious peers, real problems, and modern tools, and the ceiling is very high.
So what is the finished product supposed to be?
Not credit hours. Not a transcript. Not a sequence of completed assignments. Not four years of seat time converted into a credential.
The finished product is judgment. Capability. Professional confidence. Intellectual seriousness. A stronger sense of agency. The ability to walk into ambiguity, make sense of it, work with others, and produce something valuable.
That is what universities claim to manufacture.
The uncomfortable question is whether our manufacturing process is any good.
For a long time, universities could avoid that question because comparison was difficult. Education was expensive, but so was everything around it. Expertise was scarce. Books were scarce. Access to research was scarce. Feedback was scarce. Tutoring was scarce. Professional networks were scarce. Computing power was scarce. If students wanted structured access to knowledge, expert interpretation, and a recognized credential, the university was the obvious place to go.
That world is gone.
AI has not made education irrelevant. It has made the old cost structure visible.
It can explain concepts at midnight. It can generate examples. It can critique drafts. It can translate technical language. It can simulate a customer interview, a negotiation, a financial model, a sales call, a legal issue, a coding problem, or a boardroom argument. It can help students practice before they perform. It can give feedback in seconds instead of weeks. It can reduce the cost of first drafts, first analyses, first prototypes, and first attempts close to zero.
That does not mean the machine is wise. It means many things universities treated as expensive are no longer expensive.
This is where the Idiot Index gets painful.
If the raw materials are students, faculty, knowledge, tools, time, and networks, and the finished product is a capable human being ready to create value in an AI-shaped world, how much of the current university system is actually arranging the atoms into the right shape?
How much is still lecture as information delivery?
How much is still standardized homework designed for ease of grading rather than depth of learning?
How much is still slow feedback that arrives after the student has emotionally and cognitively moved on?
How much is still curriculum built around departmental convenience?
How much is still a compliance machine wrapped around an educational mission?
How much is still the assumption that if students sit in the right rooms for the right number of hours, transformation has occurred?
This is not an argument against universities. It is an argument against confusing the inherited process with the product.
The university still has materials AI does not have. It has legitimacy. It has human density. It has faculty who know when an answer is technically correct but intellectually thin. It has peers who create pressure, comparison, friendship, ambition, and accountability. It has rooms where students have to speak out loud in front of other people. It has internships, projects, labs, studios, clinics, competitions, alumni networks, and the social proof that still matters in labor markets.
Those things are not trivial. They are the parts of the rocket that actually need engineering.
But if universities spend their energy defending the old arrangement of atoms, they will lose the plot. The grind is ending. The old bargain was that students would prove themselves by doing the lower-level work. Read the chapter. Summarize the case. Write the memo. Build the spreadsheet. Search the database. Format the deck. Clean the data. Draft the email. Complete the assignment. Accumulate enough evidence of compliance and eventually earn a shot at more interesting work.
AI is eating that bargain from the bottom.
Entry-level work is being automated, compressed, or turned into supervision of machines. That means the educational ladder cannot keep using low-level task completion as its first rung. If AI can produce a passable version of the assignment in twenty seconds, then the assignment is no longer a reliable proxy for learning. The assignment may still be useful, but only if it is redesigned around judgment.
This is the move universities have to make.
Stop asking whether students used AI. Start asking whether the work would survive contact with reality.
Can the student frame the problem?
Can the student decide what evidence matters?
Can the student tell when the model is bluffing?
Can the student improve the output?
Can the student defend a recommendation in front of people who disagree?
Can the student notice the thing the prompt did not ask for?
Can the student understand the human context well enough to know what should not be automated?
That is education after the grind.
There is a line I keep coming back to: “You’re asking me how a watch works. For now, let’s just keep an eye on the time.”
It is a great line because it is not really about watches. On the surface, someone is asking for a detailed explanation of a complex mechanism. The answer is not anti-intellectual. It is not a dismissal of curiosity. It is a statement about sequencing.
You do not need to understand gears, escapements, balance wheels, quartz crystals, and power reserves to know what time it is. Most people can use a watch perfectly well without understanding how a watch works. That does not mean watchmaking is fake expertise. It means expertise comes in layers.
This is one of the hardest lessons for young people, apprentices, junior engineers, new managers, and students entering any serious field. When you are new and driven, you want the whole model immediately. You want to know why the system works this way, what every part does, where every dependency lives, and what happens under every possible condition.
Those are good questions. They are the right questions. But sometimes the honest answer is that the explanation requires six months of background knowledge before it will make sense.
An experienced person may be saying something very specific: you are asking a valid question, but you are asking it several chapters before the book has introduced the necessary concepts.
This matters for education because AI makes the illusion of instant expertise much easier to produce. A student can ask the machine for the whole explanation and get one. It may even sound good. But sounding good is not the same thing as understanding. A fluent answer can hide the fact that the student has not yet earned the conceptual scaffolding needed to judge it.
So the goal is not to suppress curiosity. The goal is to teach timing. Sometimes students need the mechanism. Sometimes they need the time. Sometimes they need to act before they understand every gear inside the system. Good education helps them know the difference.
In business schools, the implications are especially direct. We cannot prepare students for organizations by training them to complete isolated tasks in artificial conditions. Organizations are not short on task completion anymore. They are short on people who can transform messy systems.
A graduate who can ask better questions, configure tools, interpret weak signals, build trust, evaluate outputs, and move a team through uncertainty is valuable. A graduate who can merely produce the same artifacts AI now produces faster and cheaper is in trouble.
The university has to become less like an information delivery system and more like a capability factory. That does not mean everything becomes vocational. It means the liberal, professional, technical, and social parts of education have to be recombined around actual human development.
Students should use AI constantly, but not passively. They should use it to generate alternatives, test assumptions, practice arguments, build prototypes, expose gaps, and accelerate iteration. Faculty should stop treating AI as a cheating device and start treating it as the new baseline condition of knowledge work.
The question is not whether the student can produce a polished answer. The machine can do that. The question is whether the student can become the kind of person who knows what answer is worth producing.
That requires harder teaching, not easier teaching. It requires more live performance, more oral defense, more project work, more messy data, more ambiguous cases, more interaction with real organizations, more reflection, more critique, and more accountability. It requires faculty to move up the value chain too. If AI can explain the textbook, the professor has to do what AI cannot do as well: set standards, notice growth, model judgment, create pressure, tell the truth, and design experiences that change students.
The worst possible response is to preserve the old system by policing the tools. That is reasoning by analogy. That is asking what schools have always done, what other schools are doing, and what the policy language should say.
First-principles thinking asks a better question.
What are we trying to make?
What materials do we actually have?
What does AI make cheap?
What remains expensive because it is genuinely human?
What arrangement of people, tools, practice, feedback, and trust would produce the graduate we claim to value?
If we started there, much of the university would look different.
Courses would become less about content coverage and more about capability progression. Assessment would become less about artifact submission and more about defended judgment. Faculty time would shift from repeating explanations to coaching performance. Students would spend less time proving they can complete tasks and more time proving they can think with tools under real constraints.
The university would still matter. Maybe more than before. But it would matter for a different reason.
Not because knowledge is hard to find.
Because transformation is hard to fake.
The Idiot Index of higher education is not a spreadsheet calculation. It is a moral and strategic question. How much of what we ask students to pay for is actually turning raw potential into capability? How much is just the inherited cost of arranging atoms badly?
AI did not create that question.
It made it impossible to avoid.
Get essays like this in your inbox
New essays on AI, education, and the future of work. No spam, unsubscribe anytime.