For 30 years, creators published freely to gain attention, which they converted into reputation, jobs, or customers. AI search intercepts this attention by synthesizing information, removing the rational self-interest for creators to share knowledge openly and pushing them to create paywalls.
AI models are architecturally designed to summarize the past. As new, creative, and forward-looking knowledge gets paywalled, the majority of users relying on free AI tools are fed a constant stream of the 'recombined past,' which may systemically stifle future innovation and critical thinking.
The retreat of legitimate experts from the free, open web creates a content vacuum. This space is not left empty; it's actively filled by malicious actors—scammers, content farms, and influence operations—for whom polluting the public knowledge pool with low-value content is their business model.
As valuable human knowledge moves behind paywalls, only well-funded AI labs can afford to license it for premium models. Free, mass-market AIs will be trained on an aging, increasingly synthetic public web, creating a significant information gap between paying users and the majority.