[论文] A paradox of AI fluency

论文概要

研究领域: NLP
作者: Christopher Potts, Moritz Sudhof
发布时间: 2026-04-29
arXiv: 2504.21111

中文摘要

用户对AI的熟练程度在多大程度上塑造了AI实际为他们提供的价值？这个问题对用户、AI产品构建者和社会整体都至关重要，但尚未得到充分探索。通过对WildChat-4.8M中27K段丰富标注的对话样本分析，我们发现熟练用户比新手承担更复杂的任务，并采用根本不同的交互模式：他们与AI协作迭代，精炼目标并批判性评估输出，而新手则采取被动姿态。这些差异导致了AI熟练度悖论：熟练用户比新手经历更多失败——但他们的失败往往是可见的（是其积极参与的直接后果），更可能导致部分恢复，并且伴随着在复杂任务上的更大成功。相比之下，新手更常经历隐形失败：对话看似成功结束，实则偏离目标。这些结果重新界定了AI成功的依赖因素。个人应采取主动参与而非被动接受的姿态。AI产品构建者应认识到，他们设计的不仅是模型行为，还有用户行为；鼓励深度参与而非无摩擦体验，将带来整体更大的成功。

原文摘要

How much does a user's skill with AI shape what AI actually delivers for them? This question is critical for users, AI product builders, and society at large, but it remains underexplored. Using a richly annotated sample of 27K transcripts from WildChat-4.8M, we show that fluent users take on more complex tasks than novices and adopt a fundamentally different interactional mode: they iterate collaboratively with the AI, refining goals and critically assessing outputs, whereas novices take a passive stance. These differences lead to a paradox of AI fluency: fluent users experience more failures than novices -- but their failures tend to be visible (a direct consequence of their engagement), they are more likely to lead to partial recovery, and they occur alongside greater success on complex...

自动采集于 2026-04-30

#论文 #arXiv #NLP #小凯