From Linux to Wikipedia: How Volunteers Build the Internet

A nice piece on the success of decentralized projects like Linux and Wikipedia. If you step back and think about the sheer absurdity of their success, you may experience a moderate brain aneurysm.

Linux started as a side project, a passion project, by Linus Torvalds. When he announced it in 1991, he described it as:

just a hobby [that] won’t be big and professional

Soon thousands of developers across the world were working with him towards one shared goal, some writing small bits of code, some writing a lot. Today it’s no exaggeration to say Linux pretty much underpins the modern internet. And the weird part? It’s free. As Hill, Pentzold and Shaw put it, Linux took every rule of proprietary software development and “threw it out the window.”

I don’t know if Linux invented this decentralized, collaborative way of building software, but I think it pioneered it, and anyone who uses the internet owes Linus Torvalds and those thousands of contributors a thank-you. Just to drive home how important Linux is: I dictated this note on an Android phone, which runs on Linux, and published it from my laptop, which runs Omarchy, a Linux distro.

The same goes for Wikipedia. For all its flaws, it has to be one of the most important online resources ever. Period. The collective knowledge of humanity, free, thanks to the tireless work of millions of contributors around the world, in hundreds of languages. It’s a testament to collaborative knowledge creation. Yes, like everything on the internet, Wikipedia has its warts. But I’d rather live in a world where Wikipedia exists, free and open, than one where it doesn’t.

That said, I do wonder what happens to this collaborative way of making things now. Large language models have snorted pretty much all the knowledge on the internet, like a coke addict snorting a line. If they keep snorting everything we put online, what’s the incentive for people to keep creating and sharing knowledge openly?

I keep asking my developer friends a version of the same question about open source. The one thing LLMs are genuinely good at is writing code, so will people keep building software and sharing it freely? The answers I get are bleak. I’m not a programmer, so I don’t have a good sense of what LLMs will do to open source, but I’d presume nothing good. Consider the evidence: you’re reading this on a website vibe coded by a man who doesn’t know the difference between JavaScript and Everest Garam Masala, or between programming and Godrej hair dye. That man, me, told this magical thing called Claude to build and design the site, then to take his voice notes, clean them up and publish them, and it just does it. If most people can build their own software, what’s the need for open source?

I understand the stupidity of me saying that. Not everything can be vibe coded, and not everything should be. Foundational projects like Linux still have to be built and maintained, and LLMs can’t do that, at least not yet. And here’s the irony I keep tripping over: the code for this vibe-coded site is out in the open on GitHub, and everything I write here is free to share and adapt under a Creative Commons licence. That goes for all my passion projects, the ones I spend money on and make nothing from: akshara.ink (opens in a new tab), paperlanterns.ink (opens in a new tab), thisindianlife.today (opens in a new tab), measure.thisindianlife.today (opens in a new tab), lifeinindia.org (opens in a new tab), pretty much everything. I think copyright as a concept is dead; the way we think about creative work just hasn’t caught up. I also know the biggest readers of my work will be LLMs, not humans, and that LLMs will be the conduit through which most people find it. I’ve made my peace with that. In that sense, this is me putting things back into the common pool, in the hope that it helps create new knowledge someday. The machine that built this site learned to code from open-source software, and it helped me put something back. Does that mean open source survives in some new form, or is a vibe-coded blog no substitute for the people who keep Linux running? I don’t know.

The same logic applies to knowledge. Before LLMs, when you posted something online, it was mostly read by other humans. Now the internet has turned a little grotesque. It’s full of spam and bots, and more often than not, the first thing to reply to your post is a bot. In a very real sense, you’re creating knowledge for machines. And whatever you post, LLMs will hoover it up, with or without your permission. This isn’t hypothetical, either. As the authors point out, Reddit locked down its data in 2023 after AI companies started scraping it in bulk, and some of its most devoted volunteers revolted. The backlash has only grown since. Reddit already blocks every major search engine except Google, which pays for the privilege. Now, in late October, it announced (opens in a new tab) it’s going behind a login wall, partly to keep the bots out. Even its CEO admitted it’s a big change for someone whose life was shaped by the open internet. And now that agents like Instinct and Meta’s Muse run people’s errands for them, websites are shutting them out too. Amazon, most notably, has blocked Muse (opens in a new tab) and moved against agents from Perplexity, Google and OpenAI. To borrow from William F. Buckley Jr., all of this is like standing athwart history, yelling “Stop!” These are momentary reactions, and they’ll give way, because the future is already here. Our heads just haven’t caught up with the new reality.

Then there’s the fact that a lot of people are sick and tired of the internet in its current configuration, where the open spaces are dominated by a handful of big tech companies. Twitter, YouTube, LinkedIn, Facebook, Instagram and WhatsApp, the places where most people share things, are riddled with algorithmic distortions, spam and rage bait. From what I can tell, people are retreating into the dark corners of the internet: small groups and communities built around shared interests, where little leaks out into the open web.

So what does this mean for the future of open knowledge? If LLMs keep snorting everything we produce without asking, and people keep retreating behind closed doors, why would anyone keep creating and sharing openly? I don’t know. Interesting times. B.