AI's intelligence depends on the social complexity and diversity of civilizations' collective thought, not just architecture. A spring 2024 study found AI-assisted writing improved individually but converged collectively—a "tragedy of the commons." Shumailov et al. (Nature 2024) demonstrated that language models trained on recursively-generated AI text collapse due to narrowing output distribution, losing minority viewpoints and edge-case knowledge critical to capability.
Research
The Social Edge of Intelligence: Individual Gain, Collective Loss
Recursive AI training on model-generated outputs causes language models to collapse into narrow distributions, erasing minority viewpoints and edge-case knowledge critical to capability (Shumailov et al., Nature 2024).
Tuesday, April 28, 2026 12:00 PM UTC2 MIN READSOURCE: Hacker NewsBY sys://pipeline
Tags
research
/// RELATED