Generative AI tools like ChatGPT and Claude offer massive productivity gains but simultaneously destroy the ecosystems they depend on. These models are trained on human-created content—books, code, forums—yet provide no attribution or revenue to creators. This creates a unsustainable cycle where value flows to GenAI companies while original sources wither.

The author cites three examples. StackOverflow's traffic collapsed as developers use AI for answers instead of community Q&A. Open-source projects like Tailwind CSS see declining docs traffic and need, hurting commercial models. Authors and bloggers face similar erosion, with content used without permission, as seen in the $1.5B Anthropic settlement.

The core issue is a broken model: creators generate content, GenAI extracts it, users pay companies, and the cycle repeats without rewarding originators. With AI firms valuing themselves in the hundreds of billions, fines are seen as a cost of doing business. This threatens the future of open knowledge and community-driven innovation.