Here’s a question that’s been rattling around in my head: why do we celebrate two completely opposite pieces of advice as if they’re both obviously true?

“Never reveal your sources. Your edge disappears the moment everyone knows how you did it.”

“Share everything. Knowledge locked in one head is knowledge wasted.”

Both sound wise. Both are said with a completely straight face. So which one is actually right?

The case for the vault

Why does the Coca-Cola formula sit in a vault instead of on a plaque? Why do hedge funds pay lawyers more than they pay traders just to keep their models sealed? Why did magicians form actual guilds to punish members who explained the trick?

Is it because secrecy really is the moat? Because the moment your competitor knows your playbook, your advantage becomes their starting line?

Ask any founder who’s watched a bigger company clone their product six months after a conference talk. Ask any trader who’s seen a strategy stop working the day it went viral on a forum. Doesn’t that suggest something — that some kinds of value only exist in scarcity, and the instant you explain how you did it, you’ve handed away the one thing that made you valuable?

Or is that just fear, wearing a strategist’s coat?

The case for the open door

Now flip it. Why does open-source software now quietly run most of the internet — built by people who gave away the exact thing a “guard your secrets” mindset says never to give away? Why did Jonas Salk refuse to patent the polio vaccine, calling it the property of everyone? Why does a scientific paper live or die by whether a total stranger can reproduce it?

Is it possible that knowledge behaves differently from every other resource we know? That unlike gold or oil, it doesn’t shrink when you give it away — it multiplies? That the more people who understand something, the faster it improves, because a thousand minds catch what one mind misses?

And here’s the uncomfortable follow-up: if hoarding knowledge really worked, wouldn’t every hoarder eventually win? So why do so many of them get quietly overtaken by someone who simply taught more generously, or open-sourced the thing everyone else was sitting on?

The purist and the convert

Nobody embodies “share everything” more literally than Richard Stallman. In the late 1970s, a jammed printer at MIT started it all — he couldn’t fix the bug himself because the software was locked away from him. That one moment of powerlessness turned into the GNU Project in 1983, the Free Software Foundation in 1985, and a license — the GPL — deliberately engineered so that once code is set free, nobody downstream can quietly lock it back up again.

Ask yourself: how far would you go for a principle like that? Stallman didn’t just release free software. He resigned from MIT to build an entire operating system from scratch rather than build on top of anything proprietary. Decades later he’s still doing it, still refusing convenience for the sake of the idea. Is that integrity — or is that the kind of certainty that only holds up if you never once let yourself doubt it?

Now put the opposite character next to him. Steve Ballmer, running Microsoft through the 1990s and 2000s, once dismissed the entire free software movement as a “cancer.” Under his watch, Microsoft treated open code the way a castle treats a battering ram — something to fortify against, not something to open the gates for.

So here’s the puzzle: two leaders, two philosophies, both utterly convinced they were protecting something valuable. One was protecting freedom. The other was protecting a business. Which one turned out to be right?

Then the story flips

In 2014, Microsoft got a new CEO. Within months, Satya Nadella stood on stage in front of a slide that would have gotten him laughed out of the building a decade earlier: “Microsoft loves Linux.”

What followed wasn’t just a slogan — it was a decade of decisions. .NET, Microsoft’s flagship development framework, went open source. Visual Studio Code, now arguably the most used code editor on the planet, was built and released as open source. PowerShell followed it out the door. Microsoft joined the Linux Foundation — the very organisation built around the thing Ballmer once called a cancer — as a top-tier member. In 2018, Microsoft paid $7.5 billion for GitHub, the literal home of open source software.

Now here’s the question that should stop you: was that Nadella becoming a convert to Stallman’s philosophy? Or was it something else entirely — a company realising that in a cloud-and-developer economy, staying closed had quietly become the actual liability, and giving code away was simply the smartest move on the board?

Does it matter which one it was, if the outcome — millions more developers building on Microsoft’s tools, faster innovation, a stronger ecosystem — looks the same either way? Or does motive change everything, because Stallman gave knowledge away to set people free, while a platform company gives code away to make people dependent on everything built around it — the cloud subscriptions, the services, the ecosystem you can never quite fully leave?

Compare that to the companies that never blinked. Oracle still guards its database internals like a vault. Apple still builds walled gardens where the hardware, the software, and the store all answer to one closed system — and both companies remain hugely profitable doing it. So are they wrong? Or is Microsoft’s “open” era actually just a more sophisticated version of the same instinct to control — dressed up, this time, in community goodwill?

So which is it?

Maybe the real question isn’t “secrecy or sharing.” Maybe it’s “sharing — for what?”

Does it matter whether you give knowledge away to free people, the way Stallman did — or give it away to build a bigger, stickier moat around them, the way a platform strategy dressed up as generosity might? Is there even a meaningful difference between the two, if the knowledge itself is equally useful either way?

Or does it come down to something simpler — whether you’re protecting a technique or protecting an outcome? A recipe, or a philosophy? Is a chef’s secret sauce the same kind of thing as a chef’s technique for building flavour? You can hide the sauce recipe forever — but you can teach the technique to a thousand cooks and still be the best one in the room, because technique lives in practice, not in possession.

I honestly don’t know if there’s a clean rule here. I’m not convinced Stallman, Ballmer, and Nadella were even playing the same game.

But I do know where I land.

Where I land

I believe knowledge is meant to be shared. Not because secrecy never works — it clearly does, sometimes, for some things — but because I’ve never once regretted teaching someone what I knew. I have regretted staying quiet. I have never regretted explaining the why, not just the what.

And here’s what strikes me most, looking at Stallman and Nadella side by side: it almost doesn’t matter that one shared out of principle and the other shared out of strategy. Both ended up building something no amount of hoarding ever could have — a movement in one case, an entire developer ecosystem in the other. Secrecy protects what you already have. Sharing is the only thing that has ever been shown to multiply it.

Hoarding knowledge protects a position. Sharing it builds a reputation — or a movement, or a platform — and those compound in ways positions never do.