That’s not an insult. It’s arithmetic. Do what the median person does, read what they read, take the jobs they take, hold the opinions they hold, and you get the median outcome. The market pays you what your inputs are worth.
Whether that’s a problem depends on the shape of the distribution you’re standing in.
The middle is emptying
Start with the belief almost everyone runs: work hard, get rewarded. Do X, get Y. It’s the closest thing we have to an instruction manual.
Look at what happened to the people who followed it.
In 1971, 61% of Americans lived in middle-class households. By 2023 it was 51%. Those people went somewhere, and the direction matters. The lower-income share rose from 27% to 30%. The upper-income share rose from 11% to 19%. The middle didn’t collapse downward. It emptied in both directions.
And where you start matters more than what you do. Researchers measured the fraction of children who out-earn their parents: 92% for those born in 1940, 50% for those born in 1984. Escaping your birth class became a coin flip in two generations. The decline was sharpest for children of middle-class parents, and the cause was the unequal distribution of growth, not a lack of it. The pie got bigger but the shape has changed.
So the hypothesis: hard work isn’t the mechanism anymore. At best it’s necessary and nowhere near sufficient. There’s no instruction manual, no recipe where you do X and get Y.
That leaves two honest responses.
One is to stop playing. I made that case in You can’t max a life: the sprinter and the person lying flat make the same mistake, because both believe there’s a race. You can’t be behind in a game you’re not playing. For some people, that’s the better answer.
The other is to play differently. Because some things you want for real. The career. The research bet. The business. The money. But they come with distributions attached. Most companies fail. Most bets don’t land. The distribution doesn’t care that your reasons are good.
Both answers reject the same move. Grinding up the leaderboard is the mistake in my last essay. Grinding harder on the standard playbook is the mistake in this one. And to be clear: effort was never the differentiated variable.
So here’s the fix, and it’s a slogan Apple coined in 1997: Think Different.
Why different? Because if you do what everyone else does, you get what everyone else gets. That isn’t a slogan, it’s a tautology. Same inputs, same model, same output. You read what your peers read, take the advice they take, benchmark against the person next to you, and your result lands in the middle of theirs. Grinding harder just runs the same model faster.
Differentiated returns need a differentiated generator. So how do you build one?
One woman, thirty years, one idea
Katalin Karikó arrived at the University of Pennsylvania in 1989 with one idea: inject synthetic mRNA into a cell and you could instruct it to build any protein you wanted. A drug for anything. A vaccine for anything.
Her field had already decided. Synthetic mRNA was inflammatory. Inject it into a mouse and the immune system attacked it. The mice got sick and some even died. mRNA was a dead end, and the verdict came from data.
So her grants got rejected. Then rejected again. In 1995 Penn demoted her off the faculty track and cut her pay. She stayed anyway, and kept working like nothing had happened.
What she did next comes down to three habits worth stealing.
Be open to new ideas
The insight didn’t come from working harder on mRNA. It came from a photocopier.
In the late 1990s Karikó was at a shared machine in the medical school when she started talking to Drew Weissman, an immunologist building an HIV vaccine. She was a biochemist trying to make mRNA behave. Neither was senior enough to have their own copier, so they started teaching each other their fields over it.
Half the answer came from Weissman’s world. Immunologists knew DNA can trip an immune receptor, and that a chemical modification called methylation switches the alarm off. If DNA has a chemistry that makes the immune system ignore it, why wouldn’t RNA?
The other half came from an anomaly in plain sight. Synthetic mRNA set off the alarm. But tRNA, the workhorse nobody was building a career on, didn’t. Why? Because tRNA is heavily modified. So they swapped the building blocks of synthetic mRNA, and when uridine became pseudouridine, the alarm went quiet and protein production went up.
No amount of mRNA expertise would have produced that. It took an immunologist’s rulebook applied to a biochemist’s problem. Read outside your lane. Take the meeting with the person whose work has nothing to do with yours. Every unrelated field you know is another place a connection can come from.
And the tRNA clue had been in the literature for years, filed under boring. The thing that puzzles you and won’t leave isn’t a distraction from the work. It’s the actual work.
Be willing to be wrong
Most people can’t do this one.
Karikó never ignored the evidence. Her mice really did get sick. What she refused was the interpretation. “Synthetic mRNA triggers inflammation” is a finding. “mRNA doesn’t work” is a conclusion. Everyone else collapsed the two and moved on to fundable work. She held them apart and asked why.
It’s brutally hard to actually do. Most of us do the reverse: we hold our beliefs tightly and our evidence loosely. When something doesn’t fit, we explain it away, because the alternative is admitting a thing we’ve built an identity on is incomplete. The more fundamental the belief, the harder we defend it, which is exactly when updating pays most.
Having the insight was the easy part but holding it hurt. Nature rejected her paper. Science rejected it. Immunity ran it in August 2005, and nothing happened. Nobody invited them to present and the paper collected citations at the rate of a small specialist result for fifteen years.
If your differentiated view is comfortable to hold, it probably isn’t differentiated.
Be willing to go deeper
Take a differentiated position and you’re almost guaranteed to be wrong at some point. In public, maybe for years.
Karikó was demoted for a decade before that photocopier conversation, and never had proof she wasn’t the one making the mistake. From the outside, “brilliant scientist ignored by a myopic field” and “stubborn person wasting her career” look identical. No test tells you which one you are until it’s over. Being atypical puts you in the left tail as often as the right, and most people who look like Karikó at year ten are simply wrong.
So how do you stay in for a decade with no proof? Not by grinding.
Grinding is suffering now for a payoff later, and that trade usually isn’t worth it. Karikó never made it. She told Eric Topol that if her goal had been to advance, the demotion would have ended her, because advancement was exactly what she’d lost. Her goal was to understand the thing. That’s the difference.
Ask her about those thirty years and she doesn’t describe a sacrifice. Her line is that she felt successful while others considered her unsuccessful, because in the lab she was in full control. Asked how to be successful, she says success isn’t the thing; being happy is — then talks about running six kilometers every morning before work. That detail may have endeared her to me more than anything else.
She was never in doubt, either. Pushed out of Penn in 2013, she told her chairman the building would be a museum one day. She thought she was right, and she was.
That’s the habit. The demotion couldn’t touch her because she wasn’t climbing, and you can only be demoted if you’re keeping score. Not keeping score is what let her go deep. Thirty years on one problem turns “this fails” into “this fails because of this mechanism, which I can attack.” There was no eureka: a decade of failure, a chance conversation, a borrowed analogy, then years more work.
In 2013 she left for BioNTech, a German company with no approved product. In 2020 a pandemic arrived that I don’t need to recount. The part that matters: two companies designed vaccines on a computer in days, built on her pseudouridine trick, at 95% efficacy, and shipped billions of doses. In 2023 she won the Nobel.
For thirty years it looked, from the outside, exactly like losing. She was never in the race those people were scoring.
Your organization is doing this to someone right now
One key uncomfortable part isn’t the field’s skepticism. It’s Penn. It’s the system working “against” you.
The insight that would end a pandemic sat in their building the whole time, and they demoted the person holding it. However, nobody there was a villain. There was no one inside out to get her. Penn almost certainly ran a reasonable process: funding signals merit, unfunded researchers are a bad allocation, the org chart should reflect that. The process worked exactly as designed and threw away a Nobel.
Your company does this through three doors, all well-intentioned. You optimize for predictability, so plans stop absorbing new information. You optimize for zero errors, so people stop making the jumps that look like risks on a spreadsheet. You build hierarchy, so insights get filtered out three levels below the person who could act. Karikó needed a decade of funded failure to produce one result, and no quarterly review would have approved that.
If you lead anything, replace “that’s not how we do things” with “tell me more about that.” Reward the person who surfaces a contradiction as much as the person who ships on time. Ask what would make the crazy idea work before you ask what’s wrong with it. And notice who in your building is standing at the photocopier with something you’ve already discounted to zero.
The bet
Thinking differently isn’t a personality. It’s a few habits you can start now.
Separate the finding from the conclusion, especially when everyone agrees on the conclusion. Go find the person whose work has nothing to do with yours. Chase the anomaly everyone has already decided is boring. When you lose a bet, spend an hour on the evidence before a second defending yourself. And stay on the problem past the point where staying looks like a mistake.
None of it guarantees a good draw. That’s what a skewed distribution means. But the alternative, following the crowd, guarantees the median, and the median is a number you don’t want.
Be average, expect average returns. Or don’t be.
Kariko was authentic. She just wanted to geek it out on the thing she can't stop thinking about, and do her daily 6k runs. She was happy, and that was how she defined success. She wasn't trying to think different. She was just thinking for herself.