Unverified50% confidenceOpinionExact time
对AI能力的静态认知会导致低估和高估两种典型误判
1
Sources
50%
Confidence
Long-term
Relevance
9/11/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedGary Marcus、Timnit Gebru等AI批评学者指出过度拟人化和去人化两种认知偏差都阻碍了对AI系统的客观评估与有效监管79% similarUnverifiedTejal认为外界系统性地低估了AI模型的真实能力，而非过度炒作78% similarUnverified事后可证明性（审计与追溯）是最被低估的智能体挑战维度76% similarUnverified吴恩达指出社交媒体上广泛传播的AI翻车案例导致人们对AI错误率产生过高估计，这是一种可得性启发认知偏差75% similarVerified过度依赖AI可能带来思维惰性、技能退化和信息茧房等潜在风险75% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/905459API
curl https://kongchang.com/api/v1/knowledge/claims/905459MCP
get_claim(id=905459)