How AI Is Breaking the Internet (and What to Do About It) | Matthew Prince | TED

The internet, a miracle of the last 30 years, has been driven by a consistent business model: generating traffic to ideas, then monetizing that traffic through sales, subscriptions, or ads. This model has fueled web growth across platform shifts, from desktops to social media and mobile devices, with the humble banner ad largely financing it. Google, for 27 years, has been the superhero, building the search engine and the monetization system that paid content creators for traffic. However, the world is changing with AI, marking another significant platform shift. Increasingly, people consume information directly from AI chatbots, transforming the internet from a "treasure map" search engine to an "answer engine." While beneficial for consumers, this shift is devastating for publishers because traffic, the internet's currency, is no longer flowing to their sites. Data shows it's 5.2 times harder to get traffic with Google's current model, and exponentially harder with OpenAI (750 times) and Anthropic (32,000 times). This decline has worsened, with some media outlets reporting a 30-90% drop in traffic, threatening their survival. The good news is that AI companies, and society at large, value journalism and research. The speaker proposes a new business model centered on "filling the holes in the cheese"—creating new knowledge rather than rehashing existing stories with inflammatory headlines. AI companies are willing to pay for content that generates new ideas and advances human knowledge. This contrasts sharply with the past decade's media model, which often prioritized clickbait over genuine information. The music industry’s transformation offers a hopeful precedent. Twenty-three years ago, rampant piracy through services like Napster threatened the industry. Steve Jobs’ introduction of 99-cent songs, though not the final model, paved the way for services like Spotify, which now pay billions to music creators. Spotify even identifies unfulfilled queries, leading to new music creation, like a Danish artist earning €40 million annually writing songs for specific, niche requests. Another example is the speaker's own experience with the Park Record, a local newspaper in Park City, Utah. Three years ago, he and his wife bought the dying paper, not for profit, but to support local news and voter turnout, which they significantly increased. Surprisingly, they anticipate making more from licensing their unique local content to AI companies by 2026 than from digital advertising. This demonstrates the potential for AI to reward original, local content that fulfills specific information gaps. The main obstacle to this new model is Google. The speaker argues that Google believes it has a "God-given right" to all online content without paying for it, leveraging its massive data advantage. Google is reportedly threatening content creators: "You can't stop us from training on your content or we will take you out of search." This behavior, from the former "Superman" of the internet, is holding back progress. However, a solution may lie in AI companies' shared priority: beating Google. The speaker suggests that content creators should charge AI companies for the ability to "catch Google." This strategy, while not the end game, could be the "iTunes moment" for content, demonstrating its value and paving the way for a better internet business model where original content creators are compensated.