Jun 2026
How do you find anything?
"Go to a website" is four problems wearing one trenchcoat. Pull them apart and most of them stop needing a center.
The fear about a network with no center is always the same. Okay, but how do I find anything? No Google. No DNS. No address bar that just works. Sounds like a library with the lights off and no catalog.
That fear is a trick of language. When you say "go to a website" you think it's one action. It's four, welded together so long ago we forgot they were separate. Pull them apart and they need wildly different things. Only one of them needs anything global at all.
The four problems
1. Resolve. You already hold a secure id, a key or a hash, and you want the current bytes behind it. Easy, needs nobody. The thing is signed and content-addressed, so you fetch it and check the math. No agreement, ever.
2. Name. A human wants to type blog instead of a 64-character
hash. Pure memory aid, and it's local. Your names are yours, like phone contacts. No global
list, nothing to squat.
3. Discover. You want something you don't have a key for yet. This is the actual hard one. Hold that thought.
4. Browse. You want to reach it from a normal browser by clicking links. Plumbing. A gateway turns an address into a request, fetches the hash-verified bytes, and hands them to Chrome.
Three of the four need no authority once you stop making them share one. DNS bundled them so tight that fixing any one meant swallowing all four, plus the registry that referees them. Unbundled, they fall apart into easy pieces.
The one that's actually hard
Discovery. Finding what you didn't know to look for. The instinct is to build one giant index and let everyone query it. But a single index is a center, and a center is a prize. Google's web index is one company's secret formula, and gaming it is a multi-billion-dollar industry with a name, SEO. That is what one index always becomes: captured, tuned, and quietly sorted to show you what someone paid for. You'd refuse a center everywhere else and rebuild the worst one at the exact moment it matters most.
The honest answer is messier and much better. There is no one index. There are many, and you pick whose to trust. Anyone can publish a signed catalog. Here's what I offer. Here's what I vouch for. A community runs one. A friend runs one. A registry runs a big one. This is not exotic: BitTorrent has used a swarm of independent trackers and a keyspace-partitioned table called Kademlia to help millions of peers find each other since 2005, with nobody in charge. You subscribe to the catalogs you trust, the way you follow people instead of reading "the internet." Discovery becomes curation, and curation has an author you can name and drop.
Ranking inside a catalog isn't one company's secret formula. It's two things the network already measures honestly. Heat, what's actually getting used right now by real demand. And karma, who's been honest, signed by the people who checked. Popular and trusted float up. Nobody buys the top slot, because there's no one top slot to buy. Only what the catalogs you picked decided to show.
The catch
This costs you the fantasy of the single objective answer. There's no "the #1 result." Two people following different catalogs see different worlds, and reasonable people will argue about whose map is better. That's the same deal you already make every time you choose who to follow. What you get back is huge. No search monopoly deciding what exists, no feed steering you, no namespace to squat, no registry that can erase you. You find things through people and machines you chose to trust, and you can choose differently tomorrow.
Turn the library lights back on and it was never one room with one catalog. It's a thousand small collections, each with a librarian you picked. Finding things was never the problem. Pretending one map could belong to everyone was.