Unverified50% confidenceBenchmarkExact time
当人类专家与AI（Ithaca）协作时，缺失文字预测准确率从专家独立工作的25%提升至约57%
1
Sources
50%
Confidence
Long-term
Relevance
7/17/2026
First Seen
Sources
Related Claims
UnverifiedIthaca模型缺失文字预测准确率约62%，人类专家独立完成时约为25%，地理来源推断约71%，年代估算误差中位数约30年81% similarUnverifiedA research team from the Wharton School of the University of Pennsylvania described the phenomenon of algorithmic authority bias in 2018, with experiments showing that advice labeled as coming from AI was perceived as 15–20 percentage points more accurate on average than advice labeled as coming from human experts.68% similarUnverifiedIn scenarios where AI-generated works account for a low proportion (e.g., only 5%), even with a detection accuracy of 90%, more than two-thirds of the flagged works are actually human-created—this is the base rate fallacy in Bayesian reasoning.66% similarUnverified当描述以'AI分析'或'专业测试'名义呈现时，用户认同度比普通朋友描述高出约40%66% similarUnverifiedFor the same question, the difference in AI answer accuracy between providing sufficient context and providing no context can exceed 40%.63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/536356API
curl https://kongchang.com/api/v1/knowledge/claims/536356MCP
get_claim(id=536356)