July 20, 2026 · 7 min readCopy link

writing investing business

The Mistake Is in the Magnitude

In 1995, nobody doubted the internet would matter. But holding the exponent in your head was impossible. Nobody doubts AI is a big deal either. But who is accurately imagining how big?

Scale estimation remains elusive in a world where the crowd can decipher direction. The mistake is in the magnitude.

Take Jeremy's viscerally titled "Billion Dollar PDFs". Which he expands on in Patrick's podcast:

every once in a while someone basically crystallizes a notion right at the right time in the right way that becomes the foundational viewpoint or opinion in a certain era...the Billion-dollar PDF thing is this idea that you can form billions of dollars of capital one way or another around simply setting a new idea

His framing is about the insight to set the story first, to take charge of the narrative. Part of what makes these stories¹ resonate is their common arc: how everybody is directionally correct in a prevailing mega-trend, but is off on its magnitude.

Take Situational Awareness. Leopold's seminal paper described a future where AI is the dominant knowledge force...by 2027! Written in June of 2024, the framework presented was contrarian. Which feels weird. The AI narrative was well under way already. By publication date:

• ChatGPT had 200MM+ users

• Apple Intelligence was in full swing

• The EU had already created an AI Act

Even the layperson was familiar with AI. But Aschenbrenner's insight was you all are drastically underestimating how important AI will be, how quickly it will propagate, and how gigantic the resources dedicated must be. And not by 10 or 20%. The mob was off by 1000x or 2000x.

Bill Gates's 1995 Internet Tidal Wave memo established the same notion. The number of websites was growing 5x monthly. Netscape had IPO'd at a $2.3B valuation (and doubled). Nearly 10 million adults in the US alone were online. The internet was far from a phenomenon banished to obscurity.

Gates's insight was in his fourth line of the memo:

In the next 20 years the improvement in computer power will be outpaced by the exponential improvements in communication networks.

I dare even say Gates didn't realize how large the web would become over the next 30 years.

The growth rates are astounding, and being able to visualize a compounded rate so high over decades is impossible. Arguably, we still have a long way to go before we stop gaining speed. Maybe we are still underestimating it.

Peter Thiel² endured a blind spot as well:

Whenever a tech startup has a strong up round led by a top tier investor (Accel counts), it is generally still undervalued. The steeper the up round, the greater the undervaluation.

He passed on Facebook's Series B. $500MM valuation, a $1B offer six months later, then a $15B valuation eighteen months after. Now to $2 trillion. Even the greats underestimated how good being right can be.

In 1984, Buffett explained his investment school of thought via the coin-flipping analogy in The Superinvestors of Graham-and-Doddsville.

By then, Paul Volcker's fame had peaked and greenmail was a well-established practice. Outside of academics, rarely was there disagreement about why Buffett's school of thought worked. But nobody took these schemes as far as Buffett had for so long. Duration was the trade. The longer you followed his philosophy, the bigger the underestimate turned out to be.

Screen nearly 9,000 13F filings for concentrated funds that held a single position for over a decade. The list doesn't crack a hundred names. The positions that survived are no secret. Big tech and Berkshire. Names we all owned at one point or another. It was easy to agree investing was riding a great wave. Staying on it for a decade: not so much.

Or as Soros put it to his mentee who pitched an investment:

That is the most ridiculous use of money management I've ever heard. What you describe is an incredible one-way bet. We should have 200% of our net worth in this trade, not 100%.

Druckenmiller's idea was precise. He just didn't realize how right he was. He later explained how he channels his mentor, saying:

I do like it when I have extreme conviction and no one else believes it. It gives me even more conviction.

This instinct plays out in operations: not who spots the trend, but who dares to push it further than anyone else believes rational. Extraordinary companies tend to max out 1-3 variables and are OK with being mediocre at the rest.

The Replaced also knew the big factors but got it wrong. Maybe they strayed too much; other times they were just not as excessive.

Take Microsoft replacing IBM. Software was once Big Blue's game. Yet Microsoft figured out just how extreme building and distributing software could be. 100x more radical than IBM, which enabled them to surf for a very, very long time.

Or Sears, which was replaced by Walmart. Walton was simply 100x better at scale-merchandising. Kodak, which was the technological leader in film, got creamed by Canon, which embraced digital photography. Canon just went further out on the adoption curve. Something Kodak had historically dominated.

All of these examples are fine with being passable in other areas of their business. Microsoft is not innovative. Walmart isn't the best at ecommerce, and Canon doesn't make the most portable cameras. But they all figured out the 1-3 variables to be immodest on, and made the rest good enough. And these variables were not new to the industry! They were just not well believed.

Riding the edge of the bubble is dangerous. Not every factor should be taken to the extreme. Not all waves will churn for decades. Billion Dollar Ideas don't come around every quarter. The multitudes can be wise!

You also shouldn't over-index on one specific trend. A failure mode is the world changing and your notions stagnating. Sometimes the crowd over-corrects and you should adjust enthusiasm accordingly. Don't get stuck! Always Be Correcting assumptions.

And there may be no story at all. Maybe it was false³, overstated⁴, or plain wrong⁵ from the start.

But sometimes it holds. If your story is true, the dangerous question isn't "what if this goes wrong?" It's legitimate, but it crowds out the underrated one: what if this goes right? If everyone is drastically underestimating how good the upside can get, that's not a reason to doubt yourself. It could be the whole opportunity.

Thank you to Duncan, Justin, Andrew, Chris, and Ted for reading drafts and contributing

← All posts