AI Models Collapse When Trained on Recursively Generated Data
Shumailov et al., 2024, AI models collapse when trained on recursively generated data (Nature, vol. 631)
Subscriber deck — free during early access
2 cards in “Discussion” · whole deck
Discussion
- Why does preserving low-probability events matter, according to Shumailov et al.?Rare events are often relevant to marginalized groups, so losing them harms fairness, and they are vital for understanding complex systems.
- Shumailov et al., 2024, AI models collapse when trained on recursively generated data (Nature, vol. 631), Discussion, pp. 758–759
- What practical consequences do Shumailov et al. draw for training future models?A “first mover advantage”: access to genuine human-generated data (pre-LLM web crawls, real human interactions) becomes increasingly valuable, and the provenance of crawled content must be tracked, perhaps through community-wide coordination among those building and deploying LLMs.
- Shumailov et al., 2024, AI models collapse when trained on recursively generated data (Nature, vol. 631), Discussion, p. 759
Source
- Ilia Shumailov, Zakhar Shumaylov, Yiren Zhao, Nicolas Papernot, Ross Anderson and Yarin Gal. 2024. AI models collapse when trained on recursively generated data. In Nature, vol. 631, pp. 755–759. License: CC BY 4.0. Flashcards adapted from this work. doi:10.1038/s41586-024-07566-y