Answer engine
Search engines scraped and indexed copyrighted content from the web in exchange for driving traffic back to those sites. It was a deal that benefited everyone: users, content creators, search engines, and advertisers. That flywheel fueled the growth of the internet for over two decades.
Then came the answer engines (Perplexity, Bing Chat, ChatGPT). They don’t just take search traffic (which is fine); they also take away website traffic from content creators. They deliver great value to users — who are even willing to pay for it — but in doing so, they broke the balance that kept content creation going.
What’s needed now is a new balance. A pricing model that redistributes subscription income from users. We already have two models to emulate: Netflix and Spotify.
Like streaming services, chatbots could sign up large content providers — from the New York Times to Hacker News, from the Financial Times to Substack. These players have enough content to form lucrative partnerships.
For everyone else, especially blogs, the model could be YouTube Premium or Spotify: per-view redistribution from a shared subscription pool.
The quality of answer engines will increasingly depend on the breadth of available content (and the network company can block the engines1), as the underlying tech becomes less of a moat.
We need that flywheel of content creation to continue to maintain the internet worth having, rather than a cesspool of enshittified posts and AI generated images.
Aside from YouTube, Meta, Amazon and Netflix - which covers most of Internet - the rest goes through Cloudflare, CloudFront and Akamai that could team up to enforce the pay-per-view redistribution model.