Unverified70% confidenceFactExact time
A single AI inference request consumes little electricity, but because requests number in the billions, the cumulative energy consumption may exceed that of training itself.
2
Sources
70%
Confidence
Long-term
Relevance
7/10/2026
First Seen
Sources
谷歌加码弗吉尼亚：数据中心扩张背后的能源与就业博弈
rss6/11/2026
Related Claims
VerifiedA 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.74% similarUnverifiedAI生成假设的边际成本趋近于零，一旦模型训练完成每增加一条假设几乎不产生额外成本74% similarUnverifiedThe threshold for AI systems regulated under the draft is training compute exceeding 10^26 FLOPs or parameter count exceeding two trillion.72% similarUnverifiedSome argue that current AI lags behind humans by approximately 6 orders of magnitude on the efficiency dimension.70% 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.70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/461984API
curl https://kongchang.com/api/v1/knowledge/claims/461984MCP
get_claim(id=461984)