Remember when the internet was a database? You could just query it. Now we are trapped in digital walled gardens, begging for API scraps while platforms throttle us to boost their IPO valuations.
We didn’t lose the open web because of a technical failure. We lost it because hoarding data is profitable.
You’ve probably noticed the rush to build “decentralized” protocols like ATProto (Bluesky). The pitch is seductive: we can query the web like a giant database again! Select * from Internet.blogposts. But here’s the twist: the open web already solved this. RSS, OPML, semantic HTML. We had the tools. The tech was never the problem.
Complex decentralized protocols aren’t the salvation of the open web; they are over-engineered distractions from the brutal reality of network effects.
Look at Reddit. For years, they begged developers to build third-party apps. They encouraged it. Then the IPO loomed, and suddenly, the API gates slammed shut. The developers who built on Reddit’s open API were priced out overnight. It wasn’t a glitch. It was the business model. Attention is all you need, and the platforms have it.
When you build on a platform’s API, you aren’t a partner. You’re a free R&D department waiting to be liquidated.
Stop pretending we can engineer our way out of this. ATProto is dazzling, sure. But is it truly decentralized, or are its proponents just blinded by their own cleverness? At the end of the day, network effects are a brute-force economic weapon. You cannot code your way around a monopoly that owns the audience’s attention.
The next time you think about building on a platform API, or investing your soul into a new decentralized protocol, remember the lesson of the open web. The architecture was perfect. The tech was flawless. But tech doesn’t win against economics.
The internet didn’t break. It was bought, fenced in, and monetized—and no amount of elegant code will buy it back.
FAQ
Q: Isn't ATProto actually fixing this by making the protocol open?
A: No, it's treating a socio-economic problem as a software bug. The open web already had RSS and semantic HTML. ATProto is over-engineering a solution to a problem that centralized network effects already won.
Q: Should developers just stop building third-party apps?
A: If you build on a centralized platform's API, assume you have an expiration date. You are free R&D. Build on independent protocols or prepare to be priced out when the platform needs to monetize.
Q: Is the open web actually worth saving?
A: The nostalgic dream of the open web is dead. The future isn't decentralized protocols; it's accepting that centralized attention is a monopoly that brute-forces its way to the top, regardless of the underlying tech.