- Futurism argues that AI is creating “doom loops,” in which the technology extracts value from human labor and content while simultaneously weakening the economic incentives to keep producing those inputs.
- Internal Microsoft documents made public during The New York Times’ lawsuit warn that AI content strategies could harm both model performance and the web as a whole.
- One document describes the paradox of AI products threatening the economic foundations of essential providers in the LLM “content supply chain.”
- Journalism is a prime example. Tools such as ChatGPT and Gemini can summarize articles without sending readers to the original source, potentially reducing traffic and weakening media organizations’ revenue models.
- A similar loop is emerging in music and film, as AI-generated content competes directly with creators whose work and data previously helped develop these models.
- The article cites figures suggesting that more than 200,000 creative-sector jobs disappeared in the US over roughly four years, although these figures do not by themselves prove that AI was the direct cause.
- The job market is also developing a similar loop: applicants use AI to generate large numbers of applications while employers use AI to screen them, increasing application volumes and making it harder to distinguish candidates.
- Greenhouse’s CEO describes a situation in which each side uses AI to solve its own problem, while that use makes the other side’s problem more severe, creating further incentives to keep using AI.
- Education is presented as another loop, with students using AI to complete assignments while teachers use AI to create coursework, reducing the role of direct thinking on both sides.
- The copyright debate remains unresolved. AI companies argue that model outputs are sufficiently different from source material to qualify as “fair use,” while content owners continue to challenge that position.
- The long-term risk raised by the article is that AI needs continuous access to new human-generated knowledge, culture, and content while potentially weakening the revenue and incentives that encourage people to keep producing those inputs.
📌 “Doom loop” is a useful concept for examining AI from an ecosystem perspective rather than simply measuring the productivity of individual companies. AI may help each participant reduce costs in the short term, but if many participants optimize in ways that weaken data sources, employment, revenue, or input quality, individual benefits can create collective costs. However, this is a strongly critical commentary on AI, so empirical evidence should be distinguished from the author’s causal claims.
