Tailwind CSS is more popular than ever. Downloads are at record highs. The framework powers millions of projects.

And this week, the company behind it laid off 75% of its engineering team. That’s three people. Tailwind Labs had four engineers. Every headline screams “75% layoffs” because it’s technically accurate, but the framing obscures the reality: this is a tiny team, not a mass redundancy.

This isn’t the story you’ve been hearing. AI didn’t replace the developers - developers are using Tailwind more than ever. AI replaced the web as the interface layer. And that broke the business.

When a GitHub PR Reveals a Business Crisis

In November 2025, a developer named quantizor submitted PR #2388 to the Tailwind docs repo. The proposal: add an /llms.txt endpoint that would concatenate all 185 documentation files into a single AI-optimized text file.

On January 6, 2026, Tailwind Labs laid off three of their four engineers.

On January 7, founder Adam Wathan rejected the PR and explained why:

75% of the people on our engineering team lost their jobs here yesterday because of the brutal impact AI has had on our business. Traffic to our docs is down about 40% from early 2023 despite Tailwind being more popular than ever.

— Adam Wathan, CEO of Tailwind Labs

The math doesn’t lie. Tailwind sees 75 million npm downloads per month. Revenue is down nearly 80%. Making documentation more accessible to AI would only accelerate the collapse.

The paradox in one sentence

Tailwind is growing faster than it ever has, is bigger than it ever has been, and the business can no longer afford to maintain it.

This Isn’t About Fake Content

I’ve written before about how the web is bifurcating and how ad budgets are feeding bots. Those posts covered the malicious side: bot farms, click fraud, $238B wasted on fake impressions.

Tailwind’s collapse is different. This is legitimate AI usage killing a legitimate business.

A developer asks Claude Code: “How do I add responsive padding in Tailwind?”

The AI answers directly, correctly, instantly. The developer ships their feature. They never visit tailwindcss.com. They never see the link to Tailwind Plus. The sale never happens.

I never took the dead internet theory that seriously but it seems like there are really a lot of LLM-run twitter accounts now.

— Sam Altman, OpenAI CEO

Altman’s talking about fake accounts. But the Tailwind situation reveals something more structural: the “dead” in dead internet isn’t fabricated content. It’s the economic death of traffic as a business metric.

Humans haven’t disappeared. They’re just not visiting websites anymore. AI is their new browser.

How AI Breaks the Traffic Economy

The funnel used to be simple:

  1. Developer searches Google
  2. Lands on documentation
  3. Discovers commercial products
  4. Converts to paying customer

AI short-circuits at step one. The question gets answered without a website visit. Steps 2-4 never happen.

The crawl data tells the story. Google’s crawl-to-referral ratio sits around 14:1. For some AI crawlers, it’s 73,000:1. They consume everything and send back nothing.

This affects every business model built on traffic:

  • Documentation sites that monetize through premium products
  • Tutorial platforms that convert readers to paid courses
  • News publishers that sell ads against pageviews
  • Open source projects funded by awareness funnels

When AI mediates the relationship between humans and information, the attention economy breaks.

What Comes Next

The infrastructure for a post-traffic web is already being built:

  • Agentic Commerce Protocol: OpenAI and Stripe developed ACP to enable transactions inside AI interfaces. PayPal, Shopify, and Etsy are already onboard. Commerce moves from websites to prompts.

  • Pay-per-crawl: Cloudflare’s HTTP 402 experiment lets publishers charge AI crawlers for access. TIME, AP, and Conde Nast signed on. But it’s unclear if AI companies will pay.

  • Sponsorship: Within 48 hours of the layoffs, Google’s AI team, Vercel, Gumroad, Lovable, and Macroscope all pledged financial support. Guillermo Rauch called Tailwind “foundational web infrastructure.” It’s a band-aid, not a solution. Foundation funding doesn’t scale to every affected project.

Update: Jan 9

Wathan posted that he’s been “overwhelmed by the support” and clarified: “we’ve still got a fine business (even if things are trending down), just not a great one anymore.” Not existential - but a forced contraction that didn’t have to happen.

Experiments, not solutions

These are early attempts to rebuild monetization for an AI-first world. The open source funding model hasn’t found its replacement yet.

The Interface Has Moved

For 25 years, web traffic meant usage. You could measure engagement in pageviews, build businesses on conversions, fund open source through awareness funnels.

That equation broke. Usage now bypasses traffic entirely.

The web isn’t dying. The traffic economy is.

Businesses that built on discovery through search and documentation are most exposed. The ones that survive will be the ones that follow the interface: embedding into AI workflows, participating in agentic commerce, finding ways to monetize where users actually are.

That’s not a prediction about the future. That’s a description of what Tailwind proved this week.

Related Posts