Wikipedia celebrated its 25th birthday this week with a milestone that would make any Silicon Valley founder weep: 7 billion monthly visitors, zero ads, and a business model that somehow still works.

I say "somehow" with affection. As someone who spent years pitching VCs on growth metrics and monetization strategies, Wikipedia's continued existence feels like a glitch in the matrix. Every rule of modern internet business says it should have either sold out or shut down by now.

The technology is impressive. The question is whether anyone needs it. In Wikipedia's case, the answer has been a resounding yes for a quarter century. The site serves more traffic than Netflix, costs a fraction to run, and maintains editorial standards that put most newsrooms to shame. All while running on donations and operated by volunteers.

But 2026 looks different than 2001. Wikipedia now faces challenges that Jimmy Wales couldn't have imagined when he launched the project. AI companies are scraping its content to train language models, governments are increasingly hostile to open information, and a generation raised on TikTok has different expectations for how knowledge should be packaged.

The AI threat is particularly interesting. ChatGPT and its competitors have become the new first stop for quick facts, the exact use case Wikipedia dominated for two decades. Yet Wikipedia's traffic keeps growing. Why? Because LLMs hallucinate, and people have figured that out. Wikipedia might not always be perfect, but it shows its work. Every claim links to a source. Every controversy gets documented in the talk pages. It's transparent in a way that black-box AI models fundamentally cannot be.

The infrastructure powering this is surprisingly modest. Wikipedia runs on about 1,000 servers worldwide, a rounding error compared to the data centers AI companies are building. The Wikimedia Foundation's annual budget is around $180 million, less than what some startups burn through in a year before shutting down.

What makes Wikipedia work isn't the technology. It's the community. Over 300,000 active editors worldwide maintain the site, arguing over citations, reverting vandalism, and updating articles minutes after news breaks. They do it for free, motivated by something quaint: they think free access to knowledge matters.

The caveats are real. Wikipedia struggles with systemic bias, particularly around gender and geography. Articles about Western topics get far more attention than those covering the Global South. Edit wars can paralyze controversial topics. The editor base is aging, and recruiting younger volunteers has proven difficult.

Government pressure is escalating. Turkey blocked Wikipedia for three years. China maintains its ban. Russia has threatened blocks over content it dislikes. As authoritarian governments grow more sophisticated about controlling information, Wikipedia's commitment to neutrality and open access makes it a target.

What's next? Wikipedia isn't pivoting to video, launching an ICO, or getting acquired. The Wikimedia Foundation is focused on knowledge equity, working to improve coverage of underrepresented topics and languages. Boring, maybe. But it's worked for 25 years.

In an internet dominated by algorithmic feeds, paywalls, and AI slop, Wikipedia remains stubbornly committed to a simple idea: free, verifiable knowledge for everyone. The fact that this still works, at massive scale, might be the most impressive technical achievement of all. Not because the code is sophisticated, but because the model is sustainable.

Silicon Valley could learn something from that. But they probably won't.