How Chrome Ate the Web

Yesterday Mozilla’s new CEO announced that Firefox will “evolve into a modern AI browser.” The internet melted down. Someone posted a perfect meme: Firefox spent years being “the only browser that hasn’t hit itself in the dick with a hammer,” and now they’ve brought out the hammer.

But the anger misses something. Chrome didn’t just win the browser wars. Chrome won the right to define what “the web” even is. Once that happened, Firefox stopped being a competitor and became a conceptual error. Mozilla’s announcement isn’t stupid—it’s the only move they have left in a game that’s already over.

Mozilla faces three options, all bad. Stay pure and die slowly of irrelevance. Compromise and betray their principles. Wait for regulation that will come too late. They chose compromise. Their userbase is furious. Their competitors know it won’t work. Because even if Firefox becomes a great “modern AI browser,” it’ll just be a worse version of what Chrome already defines as modern.

You can’t compete when your opponent gets to define what “winning” means.

Three Layers of Capture

Chrome’s dominance operates at three levels. Most people only see the first. The trap is in the deeper two.

Layer one is implementation capture. Chrome, Blink, and V8 control how 85-90% of web traffic renders. Developers test on Chrome because that’s where users are. Sites break in Firefox, users blame Firefox. The DOJ began to address this in 2024. But this misses the deeper layers.

Layer two is epistemological capture. Google doesn’t just control the technology—they control how we measure quality. Core Web Vitals: Google-defined thresholds for performance and interactivity. Google chose these thresholds. Google measures them through Chrome User Experience Report, which only collects Chrome data. Google enforces them through search rankings.

The circle is complete. Define the metrics. Measure with your browser. Enforce through your search engine. Claim it’s “objective performance.”

Layer three is ontological capture. Chrome redefined what “the web” fundamentally is.

The old paradigm: the web was a hypertext document network. Browsers interpreted W3C standards. Standards bodies had prescriptive authority.

The new paradigm: the web is an application execution platform. Chrome defines standards through what it ships. W3C documents Chrome’s behavior retroactively.

These paradigms are incommensurable. Firefox trying to compete as a “browser interpreting standards” is like Newtonian physics competing with quantum mechanics. Different games entirely.

How It Happened

PageRank made Google the gatekeeper. AdWords created the monetization layer. But browsers still interpreted independent standards.

Chrome launched in 2008 with V8, genuinely 10x faster. Android funneled mobile users. Google Maps, Fonts, Analytics meant developers built on Google’s stack without thinking. Sites began testing “in Chrome” rather than “against standards.”

In 2013 Google forked WebKit into Blink. Chrome moved to a six-week release cycle. W3C takes years. Chrome shipped features, W3C documented them retroactively. Standards became descriptive—documenting what Chrome did—rather than prescriptive—guiding what browsers should do. Chrome went from 25% to 60% market share. Firefox fell from 31% to 11%.

Core Web Vitals completed epistemological capture. Microsoft capitulated in 2018, switching Edge to Chromium. Even breaking up Google wouldn’t undo it—Chromium is open source. The paradigm persists independent of the company.

The Anti-Monopoly Paradox

To compete with Chrome you need AI-powered development, massive telemetry, billion-dollar funding, Chromium compatibility, and proprietary optimizations. Critics oppose all of this. Environmental damage. Privacy violations. Corporate capture. Loss of independence.

Mozilla can’t match Chrome’s resources without compromising their principles. Chrome ships faster, optimizes better, attracts developers. Sites test only on Chrome. The feedback loop is vicious.

Mozilla chose pragmatic compromise yesterday. It won’t work. When your opponent defines the rules, you’ve already lost.

Safari Isn’t an Exception

Safari survives because Apple controls iOS hardware and mandates WebKit. This isn’t resistance—it’s a different monopoly. Only monopoly power can resist monopoly power. Safari accepts Chrome’s paradigm—web as application platform—it just implements different values within it. Different flavor, same web.

The Discourse Is Backwards

When sites break in Firefox, users blame Firefox. Not the site for only testing Chrome. This reveals Chrome’s web has been naturalized as “the web” rather than recognized as one implementation among many.

Switching to Brave doesn’t help. Brave uses Chromium. You’re running Google’s engine, measuring by Google’s metrics, accepting Google’s web. You just blocked some ads.

The paradigm won.

It’s Happening Again With AI

This is the same movie starting again, compressed.

Implementation capture: Transformers are dominant. NVIDIA is the hardware monoculture. OpenAI’s API format is becoming standard.

Epistemological capture: Benchmarks like MMLU define what “intelligence” means. Just like earlier benchmarks favored V8, these favor transformer reasoning. Alternative approaches score “poorly” not because they’re worse, but because benchmarks weren’t designed for them. Define the metric, measure with your implementation, claim objectivity.

Ontological capture approaching: In five to ten years, will “AI” just mean “transformer-based LLMs”? Symbolic reasoning already gets dismissed as “not AI, just rules.” The question “can AI reason?” is becoming “can transformers reason?” without anyone noticing the substitution.

Chrome took 15 years. AI might take five. The next capture may be instantaneous.

Why There’s No Exit

The DOJ limited Google’s default placement deals. The EU fined them for bundling. These address market share. They don’t touch who defines metrics or what “the web” is.

Even if Google paid nothing for placement, developers would still test only in Chrome. Their code is written in “Chrome dialect.” Their performance is measured by Chrome metrics. Their users expect Chrome behavior.

Breaking up Google wouldn’t undo the paradigm. Chromium is open source. The code persists. The definitions persist. The architecture persists.

Traditional antitrust targets market power. Ontological lock-in is reality capture at infrastructure level.

What We Lost

Chrome solved real problems. V8 was genuinely faster. Blink was genuinely cleaner. That’s how they won.

We lost the ability to imagine a different web. A web where browsers protect readers, not publishers. Where data lives on your device, not in corporate clouds. Optimized for sustainability, not engagement.

These alternatives aren’t just harder to build. They’re harder to conceive. When Mozilla says “modern AI browser,” they mean “browser that fits Chrome’s definition of modern.” The words to describe something fundamentally different are slipping away.

That’s ontological capture. Not just controlling the market—controlling the vocabulary we use to imagine alternatives.

The game ended when Chrome redefined what “the web” means. Chrome became the web. When developers say “it works in Chrome,” they mean “it works on the web.”

The sovereignty didn’t just transfer. The very concept of sovereignty was redefined to make the transfer seem natural.

Ontological capture doesn’t announce itself. It feels like progress. By the time everyone agrees there’s a problem, the vocabulary to describe alternatives is already gone.