Sometimes the most dangerous lessons are the ones we think we've learned too well.
I've written a few times about how Blockbuster blew up—but there's a key piece of the story that I wasn't aware of until recently.
The commonly told tale is that Blockbuster was stupid. They had the opportunity to buy Netflix for pennies on the dollar, turned it down, and now you'll only ever see their logo in nostalgia-bait romcoms based on the early 2000s.
But Blockbuster wasn't stupid. They were actually pretty smart and forward-thinking - they just had the cosmic misfortune of learning a good lesson at the worst possible time.
The part of the story most people overlook is that Blockbuster's first big bet on an Internet company wasn't Netflix. It was Enron.
Yes, that Enron. The one that turned out to be running one of the biggest corporate frauds in history. Blockbuster had partnered with them to stream movies over broadband networks—an idea that was absolutely ahead of its time. But when Enron imploded spectacularly, it left Blockbuster with deep scars and a newfound scepticism of flashy tech companies promising to revolutionise their business.
So when Netflix came knocking, Blockbuster's executives weren't just being blind or stupid. They were being careful. They were applying the lessons they'd learned from getting burned. They were doing what smart people do - learning from their mistakes.
They just happened to learn exactly the wrong lesson.
This is the cruel irony of experience: the lessons that seem most hard-won can often be the ones that lead us astray. We pattern-match based on past pain, not realising that the pattern itself might be wrong.
Sometimes the lessons that protect us from past dangers can blind us to future opportunities. We become so focused on not making the same mistake twice that we make new ones instead.
There's a concept in mathematics called ergodicity - it deals with how systems evolve over time, and how some events have such asymmetric impacts that they fundamentally change the trajectory of everything that follows. Some mistakes you only need to make once for the impact to be near-permanent.
A trader who loses everything on a bad bet doesn't get to average out their returns over time. A climber who fails to double-check their harness doesn't get a second chance to be more careful. A company that misses a fundamental shift in their industry doesn't always get the opportunity to pivot later.
These are what mathematicians call non-ergodic systems - where single events can have permanent, irreversible impacts. Where averages don't matter because you don't get enough tries. Where learning the wrong lesson once can be worse than not learning at all.
The tragedy of Blockbuster isn't that they failed to see the future. It's that they saw it too early, got burned, and then flinched when they saw it again. They weren't blind, they were scared. They weren't foolish, they were fearful. And fear of the past made them miss the opportunity of the present.
There's a subtle but fundamental difference between prudence and paranoia. It's not always obvious, but you've probably seen this play out many times.
A bad relationship makes us shut down emotionally rather than learn to choose better partners. A failed business venture makes us avoid all risk rather than learn to evaluate risk better. A betrayal makes us trust no one rather than learn to trust more wisely.
The challenge isn't just learning from experience - it's learning the right things from experience. It's being able to say "this went wrong because of X, not because of Y" and then watching out for X while remaining open to Y.
The next time you find yourself absolutely certain about a lesson you've learned, ask yourself: Am I learning from experience, or am I just afraid of getting burned in the same way twice? Am I being prudent, or am I running scared?
Some opportunities only come once, and learning the wrong lessons can be just as costly as not learning at all.
Learning the wrong lessons: Why being right about the present blinds you to the future
Read next
In Praise of Bad Writers
In the summer of 2026 a professor began to suspect that most of his students were using AI to cheat.
Skin in the game
This is just a draft of a post I'll share in much better shape next week. I just
Go Get a Name
If I started calling you 'Wainwright', you might wrinkle your nose. Not because it's a weird