Verified65% confidenceBenchmarkExact time
A 2024 study indicates that even when AI-generated content accounts for only 10–20% of training data, measurable quality degradation can occur after multiple generations of training.
3
Sources
65%
Confidence
Long-term
Relevance
8/11/2026
First Seen
Sources
AI时代的公地悲剧：谁在透支互联网的公共资源
Related Claims
UnverifiedIn 2023, researchers from Oxford University and other institutions published a study finding that after only 5–10 generations of recursive training, the degradation in model output quality became visible to the naked eye.80% similarUnverifiedScale AI在2024年的分析显示，部分模型在MMLU上的分数膨胀可能高达5-10个百分点，因测试集已泄露到训练数据中78% similarUnverifiedA single AI inference request consumes little electricity, but because requests number in the billions, the cumulative energy consumption may exceed that of training itself.74% similarVerifiedIn the experimental design, the AI used an 8:2 split to reserve 20% as an independent test set, and applied 5-fold cross-validation on the 80% training set to avoid data leakage.74% similarUnverified斯坦福大学2023年的研究表明，主流AI检测工具误判率可高达20%以上73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/729486API
curl https://kongchang.com/api/v1/knowledge/claims/729486MCP
get_claim(id=729486)