The Last Innings of Web 2.0
The cornerstones of Web 2.0 are under siege. Google’s search moat is eroding as users go to LLMs for knowledge-graph answers. Reddit’s users are sparring with the platform’s decisions and protesting. There are dozens of Twitter competitors — protocols, interest-based, and new networks. Meta has more enemies than ever — incumbents like Apple (ad-tracking and AR/VR) and more recent companies like TikTok.
So far, these companies have been resilient (nobody is moving off of Google or Twitter today). But narrative can be an even more powerful and self-fulfilling force. And the narrative has turned. Web 3.0 might not be crypto, but there’s a general feeling that we’re ready for something new.
Why now? Three main reasons, in my opinion:
Generative AI. While we’re in the last innings of Web 2.0, we’re in the very beginning of AI-generated content. None of these companies are in the right position to deal with that. Some respond by rejecting it outright (e.g., StackOverflow), and some are changing their core product to embrace it (e.g., Google).
Users will create vastly more content with the help of AI. More AI-generated images, AI-augmented short and long-form writing, AI-generated marketing copy, and AI-generated music.
The classic discovery methods (e.g., upvotes, thumbs-ups, likes) might not make sense for this vast amount of content. Is there enough signal to sort and filter AI content? And for many of these discovery problems, is user-generated content even the right fit, or was it simply a stop-gap?
Interoperability. The second tenant of Web 2.0 was interoperability. To support a wide variety of user-generated content, the companies became platforms. They did this through open “protocols” (both programmatic via API and not), from the video upload form on YouTube to Reddit’s 3rd party API.
But interoperability has evolved. There are a magnitude (or two) more programmers in the world who want to use APIs to interact with a platform. Platform changes and product sunsetting highlight the stress points of these Web 2.0 companies where users realize they can’t take their data with them.
There’s been a push towards decentralized networks — federated Twitter, Facebook, and Reddit clones. They all suffer the same problem of solving the hard first problem of scaling from 0-100s of users but aren’t the right solution past that. There are also more opportunities for platform extension — e.g., user-generated code as the new UGC.
ZIRP withdrawal. Zero interest rate policies and the pandemic shocked large technology companies more than they’d admit. Rapid hiring followed by widespread layoffs. A dramatic shift to work-from-home, and then rising tension as companies try to bring their employees back into the office.
Managing these like a CFO is difficult but straightforward. But there’s no doubt that the cultural ramifications of these policies will last for decades. Less motivated employees. Stock options that aren’t worth nearly what they thought.