Nobody Gets Rich From a Blue Link: A Ladder to the Table

I just remembered something sudden on the train: why are almost all the links on the internet that extremely saturated shade of blue?

Author: Koutian Wu; GitHub: ktwu01

Minds wandering, I realized the question really worth asking isn’t “why blue,” but: did the person who defined links as blue get rich?

No. We don’t even know his name. He defined a convention used by billions of people every day for thirty years, one of the most successful design defaults in human history, and made zero money from it.

This article is a ladder. Each rung looks like an answer, “go set the standard,” “go sell shovels,” “go build a private layer,” and then gets knocked down by the next rung stepping on it. We climb one rung at a time until we reach the top nobody wants to say out loud: the table doesn’t belong to the people who do the work, it belongs to the people who hold control. Even Elon Musk lost everything at this game.

Level 0: The phenomenon, the ones that lived and the ones that died

Let’s lay out the material first.

Surviving defaults: the blue link, the floppy disk = save, @ = email, Ctrl+C/V, the QWERTY keyboard (which wasn’t even designed for efficiency).

Dead defaults: <blink>, Flash, framesets, skeuomorphism, visitor counters, “Best viewed in IE6” badges.

The usual explanation is “good enough + first mover + network effects + the Lindy effect.” That explanation is tepid, post-hoc, and something anyone can say, it’s exactly the kind of “surface” I’m criticizing. It frames survival as a natural inertia, as if defaults descended from the sky on their own.

The truth is far colder: QWERTY got locked in not because it was good, but because Remington put typewriters into offices and training schools built their curricula around it, that’s distribution. Flash didn’t lose to a better technology; it died the moment Jobs decided in 2007 that the iPhone wouldn’t include it, because Apple controlled a new channel. The survival of a default is a function of distribution power, not of design quality. Remember that sentence; it will come back to haunt you at every rung of the ladder.

Level 1: “Sell shovels,” an idea that isn’t an idea

Everyone says: in a gold rush, you should go sell shovels. The Thiel-style first cut is: if an insight is common knowledge, it can’t be an advantage. You’ve heard that sentence in every startup tweet, which precisely proves it’s already fully priced in. The market for shovels is itself the next sea of blood: NVIDIA has already won the most obvious shovel, GPU. Vector databases, inference gateways, evaluation tools, every track has a hundred identical people standing in it.

A deeper cut is about mentality. “Selling shovels” is a kind of unclear optimism, its core is “I don’t know who’s going to win, so I sell to everyone,” it’s refusing to hold a judgment. And the world runs on the power law: hedging, diversifying, and making the average means your returns approach zero, because almost all returns concentrate on the single largest winner. California wasn’t built by blacksmiths selling shovels.

Climb.

Level 2: Then go set the standard, but standards can’t capture value

“Since shovels are a red ocean, let me define the protocol, occupy the standard layer, and make everyone default to me.”

That sounds profound. But it turns you into Tim Berners-Lee: he invented the World Wide Web, and deliberately, freely, without patenting, gave the world the HTTP/HTML/URL stack, the most valuable standard set in human history, and captured roughly zero personal value. The people who captured the value built on top of his standard.

This is a pattern, not an isolated case. SMTP defined global email, and nobody collected a cent from the protocol; the money was taken by Gmail. RSS is the “blue link” of content distribution; one of its key promoters, Aaron Swartz, not only never got rich but came to a tragic end, while the value was eaten by Twitter’s and Facebook’s private feeds. The value of the blue link itself was captured by Google’s PageRank, something that runs on top of “links” yet is private and uncopyable. Links were the free highway; Google built a tollbooth on the road.

Setting a default and capturing its value are two completely different games. Open standards are charity, not a business model. Climb.

Level 3: Then build a private layer on top of the standard, but that’s still labor

“Got it. I use the open parts to win distribution, and I build a private layer on top that nobody can route around, my PageRank: a data flywheel, network effects, and switching costs I personally manufacture and control.”

This is the first decent answer on this ladder. It’s also where most “deep analyses” stop.

But it’s still not the table. Because this whole thing, the protocol you wrote, the private layer you built, the clever flywheel you made, is all labor. It’s the most advanced, most valuable labor in the world, but its nature is the same as the person who colored links blue: you’re creating value, and who the value belongs to is decided by a completely different thing.

This is the sentence users keep pressing me on: even if you’re the customizer of the OpenAI protocol, so what? The engineers who wrote the GPT spec don’t own OpenAI. The more of a genius you are, the more astonishing your contribution, the more you’re creating wealth for the people who hold this company, if you’re not one of them. Climb.

Level 4: So hold equity, but equity gets diluted and can become worthless paper

At this level, the topic finally switches from “what you do” to “what you hold.” This is the real dividing line.

Thiel’s own life is the evidence: he didn’t get rich by writing PayPal’s code, he got rich by owning a big chunk of PayPal; his real wealth came from the roughly 10% of Facebook he bought for $500,000 in 2004. He didn’t build Facebook, he held a piece of it. Ownership > contribution, even heroic contribution.

Labor exhausts itself for a salary, it’s sweat; the people at the table get ownership. Everyone from levels 0 to 3, the blue-link formatter, the protocol writer, the private-layer builder, is selling labor; the only difference is the unit price. To get a seat, you have to fully switch from “what did I contribute” to “what do I hold.”

But equity isn’t the end state. Equity can be diluted into insignificance, can come without voting rights, and can be swept into worthless paper by cleanup clauses in the next round. Holding equity without control, you’re still just a higher-tier bystander. Climb, this is the last rung, and the cruelest.

Level 5: The real table is control, even Musk got toppled

Look at OpenAI.

Elon Musk is a co-founder of OpenAI. He supplied the vision, the early money, the name and the gravitas, almost everything a founder can have. Then in 2018 he left and gave up his seat. The lawsuit went all the way to a 2026 hearing, he sued OpenAI for betraying its non-profit founding mission and tried to pull it back. The court didn’t give control back to him. Musk, as strong as he is, with all his money, his prestige, and his “I created it” moral claim, in the OpenAI matter, the kind of power he held isn’t the kind of power that can overturn the table.

Now look at the other side: Sam Altman is widely reported to hold almost no equity in OpenAI. Yet he sits at the dead center of the table. Why? Because what he holds isn’t equity, it’s control: control over the board, over the company structure. The November 2023 coup lit this up: the board could fire him overnight, and he could get himself reinstated within days by mobilizing Microsoft and the whole workforce. That wasn’t an equity fight, it was a control fight. Control > equity.

That’s the blunt truth at the top of the ladder: the table isn’t the “who created value” table, and it’s not even the “who owns shares” table. It’s the “who, at the most critical moment, controls that entity” table. Contribution, standards, founding vision, even capital, are all just water flowing toward the controller.

Level 6: There’s an even bigger power above control

To be honest, we have to look one step higher, even if it’s uncomfortable.

Altman’s control isn’t absolute either; it hangs over Microsoft’s compute and capital leverage, and over regulators, states, and geopolitics. Power is a stack: labor < standards < products < equity < control < and, above control, the raw forces that can directly rewrite the rules, the distribution of the giants, sovereign states, capital itself. Musk lost to Altman because in that moment, in that entity, the level he stood on was lower than the other’s. And above Altman, there’s Microsoft.

Nobody stands at the top of the stack. Here, most “deep analyses” wrap up: recognize which level you’re on, and make peace with it. But this is exactly where a few chisels are still missing, because the whole ladder is built above ground, and the real action is below it. Climbing up, you’re just climbing a little faster than everyone else; the truth is in the foundation.

Underground level −1: the ladder itself is a trap

Why is every level, from shovels to control, a red ocean? The standard economic explanation (supply and demand, competition) is shallow. The real base of Thiel’s thought isn’t economics, it’s René Girard’s mimetic desire: people want a thing not usually because of its intrinsic value, but because other people want it.

This makes me re-read the whole article. Gold, shovels, standards, equity, control, they’re not five different things, they’re one and the same desire object, stared at simultaneously by everyone. Its scarcity isn’t natural; it’s manufactured by the convergence of the crowd. “Selling shovels” is a red ocean precisely because it became consensus, and consensus is the signal that the crowd has already converged. Every level you push to climb is a level where you’re hand-to-hand with a whole crowd of people who want the same thing you do.

So the conclusion is counterintuitive: you can’t win a mimetic war; you can only leave it. The word Thiel worships, “monopoly,” is really just a structural name for “successfully exiting the crowd.” The question was never “how do I climb higher than others,” but “how do I not stand on the ladder everyone else is climbing.”

Underground level −2: the table is built when the prize doesn’t yet exist

So when exactly does control, the real table at level 5, get obtained?

Not in the scramble. By the time the thing is valuable and the crowd has arrived, grabbing control is already too late; that’s another mimetic war. Control is quietly welded into the structure at the moment when the prize doesn’t yet exist, the entity is worth nothing, and there is no crowd.

Re-examine Musk: his real loss of OpenAI wasn’t the 2026 hearing; it was 2015 to 2018, in the years when the company was nothing and nobody was fighting over it, when he didn’t write his position into the structure, and then walked away. Altman really won OpenAI not on the night of the 2023 coup, but earlier, in the unremarked details of how the entity was architected. The table was built in advance by someone willing to bow down and design the table while everyone was staring at the gold. That’s the only time control is cheap, because it’s before the crowd.

That’s what the antidote to mimetic desire looks like brought down to earth: don’t fight for control of a valuable entity (the crowd has arrived, you’re waging an inevitably mediocre battle), but when it’s worthless and nobody cares, structurally make the control yours, and then wait.

Underground level −3: the only thing no one can take from you is a position that regenerates

But there’s one last chisel, the most honest one: even structural control can be confiscated. Dilution, cleanup clauses, courts, regulators, bigger fish, state violence, any single “holding” can in principle be taken away by the power above it.

So is there anything that can’t be taken? There is, but it’s not a thing, it’s a position: the ability to see the next secret ahead of the crowd, and the “allocator’s seat,” holding the position that can repeatedly convert that ability into new positions. This is exactly why Thiel didn’t end up being CEO of some company but went to Founders Fund: what he holds isn’t a company, it’s the meta-position that decides where capital goes. A company can be disrupted, confiscated; but a seat that, in every round, ahead of the crowd, recasts a new control, regenerates.

This isn’t the cheap chicken-soup of “your real wealth is your mind.” It’s a cold structural judgment: don’t hold a position that can be taken away; hold the position that gives birth to new positions.

Coda: this article is destroying itself

Now the darkest layer surfaces.

Everything above, “find the secret before the crowd / weld control into the structure at the founding moment / hold a seat that regenerates,” once it’s stated as advice, begins to fail. Because advice is the seed of imitation. When these sentences become yet another viral tweet, they collapse into a new, crowded level 1, exactly like “go sell shovels.” The deepest secret cannot survive being spoken aloud. The article you’re reading is, with its own hands, destroying the thing it describes.

This also explains the two ghosts at the start: the blue-link formatter and Berners-Lee saw the secret before the crowd, yet gave up the throne, vision without will. Musk had the will, but struck after the crowd had already gathered, will, but wrong timing. All three fell at different positions on the same ladder. The rare few who truly win hold, at the same time, three things that can barely be held together: the vision to see before the crowd, the will to weld control shut while it was worthless, and a seat that regenerates, and the reason this is so rare as to be nearly un-teachable is precisely that once it’s taught, it no longer holds.

So stop asking whether you should be the gold-digger or the shovel-seller. That’s a level-1 question, spoken to the people still standing on the ground. The real question is underground, and it’s not friendly:

While everyone was staring at the gold, with no crowd yet, and the thing still worthless, did you see it, and did you have the nerve to bow down and weld the table into yours?

If so, you won’t say it on Twitter. If you’re reading an article telling you how to do it, including this one, you’re probably still standing on the ground.