When the World Stopped Reading

When Everyone Stopped Reading

The day everyone became an expert, and what it costs the people just starting out.

In late 2004, a fund manager in San Jose started pulling subprime mortgage-bond prospectuses: dense, technical documents that, as Michael Lewis tells it, not even the drafting lawyers had ever read cover to cover. He scanned hundreds. He read dozens. His name was Michael Burry.

Inside those documents, in plain sight, was the defect: loan pools stuffed with mortgages built to fail. The risk wasn't hidden. It was published. Every institution on Wall Street had access to the same pages, and nearly every one of them had decided the hard part was beneath them. They summarized. They trusted the ratings. They delegated the read.

Burry read. They laughed. Burry read. Then he bet against the entire market, and then made the entire fricken market pay him.

The Big Short immortalized the exchange. "I read them," says Burry. The investor across the table is incredulous: "You read them? No one reads them." "I read them," Burry says.

It was the entire trade. Not access. Not a better model. Not inside information. The documents were there. The risk was there. The edge went to the person willing to read.

The trade paid Burry roughly $100 million personally and returned more than $700 million to his investors. From 2000-2008, his fund returns ran 489%, while the S&P 500, returned just over 2%. And here is the detail that I absolutely love: while the bet was still on, Burry had to put up withdrawal gates, negating investors ability to withdraw their capital. They revolted and threatened to sue him. They didn’t read the documents, so they couldn’t see what he saw. He made them rich anyway.

Burry later joked that only someone with Asperger's would read a subprime-mortgage-bond prospectus. He was being modest about the wiring and quiet about the choice. Reading whole documents had become so rare that doing it seemed to require a diagnosis. It didn't. It required a decision, one available to every analyst on Wall Street that year, and one that almost all of them declined.

That was twenty years ago. So, what has changed since. Artificial intelligence showed up.

Nobody reads anymore.

Not you. Not your boss. Not the client you are trying to win. The article gets summarized. The book gets condensed. The report gets fed to a model that returns five bullet points, and the five bullet points get skimmed.

I'm not scolding. I'm guilty of the same offense. The tools are good... and getting better. Summarization is not a character flaw. It is a rational response to a world that generates more words than any person could ever process.

But when it happened... that day or week when the summary replaced the source... almost nobody stopped to name it.

It was the day that everyone became an expert. All at once.

Ask anyone anything. They know. Watch them produce and deliver a confident answer to any question in less than thirty seconds. Market structure. Leadership rationale. Regulatory changes. Deal terms. Organizational theory. The answer arrives polished, organized, and reasonably accurate. It sounds like expertise. It feels like expertise. It is formatted like expertise. In a document, it is becoming nearly indistinguishable from expertise.

And that is the problem. Because two very different things just collapsed into one appearance, and one of them will carry a career.

It's the shift underneath the noise. Artificial intelligence collapsed the cost of sounding like an expert... while removing the painstaking commitment of becoming one.

Sounding like an expert used to require years. You had to read the material, sit through the meetings, write the memos, get the memos torn apart, and slowly accumulate the knowledge that allowed you to speak top of mind with confidence. The sound of expertise took time and discipline, and it was expensive to produce. It was what allowed the sound to serve as a reliable signal of the substance.

That signal is now dead. A first-year associate and a twenty-year veteran can produce the same document this afternoon. The words no longer tell you who knows. The summary generation knows everything for sure. Thirty seconds at a time. Twain saw it coming a century ago:

"It ain't what you don't know that gets you into trouble. It's what you know for sure that just ain't so." -Mark Twain

But the process of being an expert never got cheaper. It wasn’t about possessing information. Expertise is what happens when the information runs out. It's when knowledge and hands-on experience collide, allowing the expert to look at a situation and recognize the pattern. It's the ability to have experienced which of five plausible answers the client is considering and to understand with conviction, which is the right one... and importantly, the wrong one. It’s hearing a client describe a problem and recognizing the problem they did not describe. That capability was never stored in the words. It was built in the person, through repetition, through struggle, through error, over time.

The science here is old and settled. Simon on why mastery takes years: experts do not think harder, they recognize faster, drawing on tens of thousands of stored patterns built through prolonged exposure. Ericsson on how those patterns form: not through experience alone, but through deliberate, effortful practice at the edge of your ability, the kind that feels bad while you are doing it. Kahneman and Klein on when intuition can be trusted: only when it was built in an environment with real feedback, where being wrong had consequences that were felt and remembered. Sparrow on what happens when we know a machine will remember for us: we stop encoding. The mind quietly declines to store what it believes it can retrieve.

Read that last one again, because it is the trap.

The struggle is not the inefficient part of learning. The struggle is the learning. When you take and review notes in a meeting, when you wrestle a difficult article to the ground, when you write the analysis yourself and discover halfway through that you do not understand it, when you sit with a problem that will not resolve, your brain is doing the slow, expensive work of building the pattern library that will one day let you see what others can't.

The Nobel laureate Albert Szent-Györgyi defined the payoff: "Discovery consists of seeing what everybody has seen and thinking what nobody has thought." Same documents. Same data. Same lens. The pattern library is what thinks the different thought. Burry is what it looks like with a Bloomberg terminal.

The summary skips the struggle: the discipline and the cogitation. And when those are bypassed, the outcome skips the construction. The expert now delivers the conclusion without building any of the machinery that produced it.

The Olympic lifter Jerzy Gregorek compressed the whole trade into eight words: "Hard choices, easy life. Easy choices, hard life." The summary is the easy choice. Every single time.

Consuming summaries feels like learning... and that is what makes it dangerous. It produces the sensation of understanding without the substance. You finish the five bullet points and feel informed. But ask yourself one hard question about the material and the feeling evaporates.

Cognitive offloading feels like leverage... it seems like an incredible advantage. But really, it compounds like debt. Buffett has a rule for hidden debt:

"Only when the tide goes out do you discover who's been swimming naked."

For ten years, AI is the tide that keeps everyone looking dressed. But when a client asks a hard question, and the tide quickly goes out.

Play it forward ten years. The ambitious graduate who spends the first decade of their career skipping the struggle arrives at year ten able to produce everything, yet lacking the ability to evaluate anything. The pattern library was never built. The judgment never formed. There is a title on the door and a deep, dark void behind it. And here is the scary part... that deficit is virtually invisible right up until the moment it isn’t.

Because knowledge is no longer detected in documents. The data and information that form it are free now. Expertise is detected behaviorally, live, in rooms.

It shows up in the moment of the response. The client asks the question you did not prepare for. The committee member challenges the number. The deal takes a turn the deck did not anticipate. The room goes quiet and looks at you. But the prompt window is gone. There is only whatever you built.

The Navy SEALs train on a brutal piece of honesty:

"Under pressure, you don't rise to the occasion.

You sink to the level of your training."

The room is the pressure. The reps you skipped were the training.

The person with the pattern library recognizes the situation, reframes it, and answers with calm specificity. The person without it reaches for generalities, and everyone in the room feels the difference, even if nobody names it. It's the Mike Tyson rule: "Everybody has a plan until they get hit in the mouth." Expertise is what happens when you get up off the floor and deliver a solution.

I spent twenty-plus years watching selection committees decide who to trust with serious money. I can tell you the decision was rarely made during the delivery of a polished presentation. Trust was earned during the moments when the presentation broke. The question from the other side of the table. The objection nobody expected. The pause before the answer. Trust was decided when certainty disappeared, because that is the only moment that reveals whether the expertise is real.

Artificial intelligence has made that moment more decisive, not less. When every document is impressive, the document still serves a purpose, but it stops carrying the ultimate value. The room is what’s left.

And getting invited to the room is becoming more and more challenging, as the invites aren’t coming from just humans… they’re coming from machines. The machines are reading too. Artificial intelligence summarizes markets, compares providers, and screens professionals before a human ever looks. They are learning, at scale, to distinguish the specific from the generic, the consistent from the assembled, the earned point of view from the synthesized one. Generic output trains the machines to see you as generic. The invisible deficit gets scored twice: once in the room, and once in the systems that decide who gets into the room.

So let's define this properly, because the word expert is doing too much work in this conversation.

Expertise is not knowledge or data possession. Information is abundant, free, and one prompt away for everyone, which means it is worth nothing as a differentiator. Expertise is the machinery built on top of information: pattern recognition, judgment under uncertainty, and the precise ability to help a colleague, a client, or a prospect move from not knowing to committing. It is knowing which question to ask, which answer to distrust, and what the situation in front of you resembles. None of that can be handed to you, because none of it is retrieved. All of it is built.

"But I bet you can't tell me what it smells like in the Sistine Chapel." -Sean Maguire, Good Will Hunting

That is the difference. Knowledge can describe the ceiling. Expertise knows what you see and feel when you are standing under it.

Which brings me to the sentence this entire paper exists to deliver.

AI rewards the expertise it appears to replace.

The people extracting the most from these tools are the deepest experts, and it is not close. Expertise is what lets you ask the machine the right question. Expertise is what lets you recognize that the confident answer is wrong. And yes, the machine can be confidently wrong. Expertise is what lets you take five plausible outputs and know, in seconds, which one the culture of the client will embrace, which one leadership will stand behind, and which contains the error that would have embarrassed you in front of the client.

The expert uses AI as a lever and moves more than ever before. The novice uses AI as a substitute and becomes the intern of their own tools: producing work they cannot evaluate, signing their name to judgment they don't understand or possess.

AI makes experts faster. It makes novices louder. The gap between them is not closing. It is compounding, quietly, one skipped struggle at a time.

If you're early in your career, this part is for you. And I'm going to be direct, because comfortable advice is usually where bad patterns go to hide.

You are being offered a trade, every day, and it's dressed up as productivity... as operating efficiency. Get the machine to read it for you, write it for you, analyze it for you, and prepare you for the meeting.

Take the wrong side of the trade during the first decade of your career, the part when you build library of experience, and you arrive at the destination without building any form of foundation. You won't be warned, because your output will look fine. The debt only comes due in the rooms that matter. And your Venmo account won't be able to pay the piper.

Meanwhile, almost everyone around you is on the wrong side of the trade. And understand what that means: every person who stops reading makes the one who reads worth more. The market for judgment works like every other market. Supply is collapsing. You are being handed Burry's edge, every single day, and almost nobody is picking it up. You do not need to outrun everyone. You need to be one of the few still doing the reps when the reps stopped being required.

So here is what it looks like in practice. Not abstinence. Discipline. Will Durant, summarizing Aristotle, said it best:

"We are what we repeatedly do.

Excellence, then, is not an act, but a habit." -Will Durant

Expertise is not a decision you make once. It is a behavior you repeat until it becomes who you are. So here are the behaviors.

Read. Think. Re-read. Read whole things. Not everything, but some things, regularly, chosen deliberately: the dense article, the actual filing, the full transcript, the book. Reading whole things is where you learn how arguments are constructed, where the weaknesses hide, and what quality feels like from the inside. While the AI summary can tell you what a document says, it is the details of the document that will teach you how to think.

Be willing to do the little things. Take meeting notes and read them after the meeting. Assess the sequence of the discussion. Begin to recognize the patterns. When the work is in your domain, create the first draft yourself. Then let the machine tear it apart. The process, the struggle of producing the draft... well, it's where the learning lives. Then use the machine: its critique is where the polish comes from. Run the sequence in that order and you get both. Reverse it and you get neither.

AI is a sparring partner, not a ghostwriter. Use it to push against your thinking. Ask it for the strongest objections to your position. Ask it what you missed. Ask it where your argument is thin.

That is where the value is. Interrogating the machine builds judgment. Copying from it borrows the appearance of judgment, and the interest comes due the first time leadership expects you to think clearly in the room.

Put yourself in rooms where being wrong comes with a cost. Judgment only forms in environments that deliver real feedback. Volunteer for the client call. Take the question live. Make YOUR recommendation and own the outcome. The discomfort is not a hazard of the process. It is the process. It is what forges the steel of differentiated expertise.

Expose yourself to behavioral judgment wherever you find it. Watch how the experienced handle the moment the meeting breaks. What they say, what they refuse to say, how they buy time, how they reframe. The behavior you will experience is the curriculum the summaries will never contain.

None of this is nostalgia. I am not asking you to work like it is 1995. I am telling you the tools changed and the target moved, and most of your generation is aimed at the old target. The old target was knowing things. That game is over; the machine won.

The new target is building relationships and forming judgment. The ability to engage. The capacity to evaluate. The lens to recognize. The strength and confidence to decide. And, importantly, the intellect to move others from uncertainty to commitment. That game is wide open, because the machine cannot play it, and most humans have stopped training for it.

Everyone stopped reading. Everyone became an expert. And in doing so, everyone made the genuine article rare.

Rare is a career strategy. It might be the only one left.

You're in the sequence.