Extraction is rational

Every wave of the internet promised to democratize something. Every wave ended in extraction. For twenty years I thought the problem was execution, or regulation, or bad actors. It was simpler than that.

It was the money.

Why it keeps happening

Why did every platform end up extractive? The usual explanations fail.

Fiat currency is a melting ice cube. “Target inflation” sounds benign—2% per year, barely noticeable. But it compounds into a simple mandate: extract value now, before it depreciates. High time preference gets baked into the system. “Get it while you can” stops being greed. It becomes rational.

This is everywhere fiat touches. Real estate speculation. Stock buybacks over R&D. Financialization eating production. Everywhere you look, short-term extraction is winning over long-term value.

The invisible hand I couldn’t see for over twenty years: fiat debasement drives high time preference, and high time preference makes extraction rational.

Sound money changes the math.

In 2020, Adam Curry and Dave Jones built the Podcast Index, an open alternative to Apple’s closed directory. They added Lightning payments to the RSS spec. No platform required. Apps like Fountain let listeners stream sats directly to creators as they listen, just a few cents per minute, paid instantly, globally, without permission from anyone. Over 11,000 shows now support it. My own podcast runs on value-for-value—no ads, no sponsors, just listeners who pay for what they value. Creators keep 96% instead of the 50%+ that ad-funded platforms extract. When the money works, people pay.

Fiat rewards extraction. Bitcoin lowers time preference.¹ You can choose to build for the long term.

Extraction existed under the gold standard, including robber barons and railroad monopolies. But fiat systematized it. It made short-term thinking the default winning strategy across every industry, every wave.²

How extraction evolved

The eyeball era. In 1998, the Fed cut rates after Long-Term Capital Management collapsed. Cheap credit flooded the market and funded companies with no real business model. Growth at all costs. Get in, get as much as you can, get out. I was there—on Microsoft’s first browser team, then at Exodus Communications hosting the biggest sites on the internet. In early 2000, a dotcom executive gave me a tour of their newly installed rooftop drive-in theater and lobby full of millions in fine art. In the meeting, he told me capital was drying up. The only saleable asset they had was clickstream data. He was looking to sell. Desperate. Everything became about “eyeballs.” In late 2001, I laid off my entire team, then got the boot. Exodus went into bankruptcy. I understood it was a bubble. But I blamed the bubble, not the cheap money that inflated it.

Surveillance capitalism. You can’t monetize value creation directly when users expect “free.” Why do they expect free? Inflation trains people to consume now. Personal savings dropped from 13% in 1971 to under 5% today. When the future is worth less than the present, “free now” beats “pay upfront” every time. I was selling grid computing to Wall Street. I sat across from a Merrill executive who told me “don’t fuck this up"—a multimillion-dollar bonus on the line, tied to extracting more from client data. We were part of the machinery. I didn’t see it then. The pattern was everywhere. All that data needed a business model. Facebook found one: give away the product, sell the users. More engagement means more revenue. Quality becomes secondary to outrage, addiction, polarization.

Enshittification. I moved into personal data ownership. We built real frameworks for data sovereignty—software and legal structures that actually worked. We got a pilot with Mastercard. It died. Not because legal was slow or product couldn’t align—but because Mastercard sells data. Transaction processing is just the means. We were asking an extraction company to adopt anti-extraction infrastructure. They couldn’t swallow the thing. No one could. When the core business is extraction, you can’t sell them tools that threaten it.

The pattern accelerated when cheap money dried up. Uber took fourteen years to turn a quarterly profit. Spotify took seventeen to post a profitable year. When rates rose in 2022, Netflix began injecting ads, Spotify cut staff and raised prices, streaming services consolidated. The extraction that ZIRP subsidized came due.

Three stages. Same mechanism. I thought the failures were local: bad actors, poor execution, bad timing. Each explanation felt sufficient. It wasn’t.

Where sharp critics stop

Cory Doctorow nailed the diagnosis. His “enshittification” framework describes what I witnessed: platforms start off good to users while locking them in, then abuse users to serve business customers, then abuse everyone to extract maximum value.

His solution: antitrust enforcement, break up monopolies, strengthen regulation, mandate interoperability. He’s right about what happened. These things matter. But this strain of thinking—diagnose monopoly power, prescribe antitrust—stops one layer short.

When I interviewed Doctorow, I asked about protocols over platforms. If platforms inevitably enshittify, what about permissionless protocols that bypass them entirely? He pivoted. Protocols are fine for everything except money. For money, he insists it’s government’s job, not a protocol problem. Bitcoin? “The shittiest money imaginable.” Only good for buying “shitty monkey JPEGs and more Bitcoin.”

The critique didn’t bother me. The asymmetry did.

Doctorow is willing to fight for decades to make imperfect regulation work. Regulators get captured? Keep fighting. It took 35 years to break up AT&T? Worth it. Antitrust enforcement is slow and captured and often fails? Still the right approach.

But Bitcoin, sixteen years in with real problems being solved? Dismissed outright. Where’s “Lightning has problems, let’s fix them”?

Even the sharpest critics of extraction stop at monopoly power and regulatory capture. They ask how platforms become extractive, but not why extraction becomes the rational choice across every industry, every decade. Break up a monopoly and you get smaller companies running the same playbook. “This time is different” is the oldest delusion in financial history.³

I don’t have proof at scale. Neither did proponents of the internet in 1995. The examples are small because the money is young. The test is coming.

I was watching Mastercard choke on data sovereignty when Bitcoin clicked. I finally saw a way out.

I took it. I’m not alone.

Extraction can make you rich. It won’t make you right.

—

¹ For the full argument on time preference and sound money, see Saifedean Ammous, The Bitcoin Standard (Wiley, 2018).

² For data on how this pattern shows up across wages, housing, healthcare, and more since 1971, see wtfhappenedin1971.com. For economic history, see Niall Ferguson, The Ascent of Money (Penguin, 2008).

³ Carmen Reinhart and Kenneth Rogoff documented eight centuries of this delusion in This Time Is Different (Princeton, 2009).