mlsanityData to 7 Oct 2026

A paradox of AI fluency

arXiv 2604.25905 ↗First posted 28 Apr 2026cs.CLcited by 0

Not on the map. Why?

Abstract (arXiv)

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 tasks. Novices, by contrast, more often experience invisible failures: conversations that appear to end successfully but in fact miss the mark. Taken together, these results reframe what success with AI depends on. Individuals should adopt a stance of active engagement rather than passive acceptance. AI product builders should recognize that they are designing not just model behavior but user behavior; encouraging deep engagement, rather than friction-free experiences, will lead to more success overall. Our code and data are available at https://github.com/bigspinai/bigspin-fluency-outcomes

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 18 of its 41 references that are in the map. First posted on arXiv 28 Apr 2026.

Builds on

18 of 41 references.

References · 18

  1. 2022Emergent World Representations: Exploring a Sequence Model Trained on a Synthetic Taskcited by 417
  2. 2024Evaluating the World Model Implicit in a Generative Modelcited by 107
  3. 2024Collaborative Gym: A Framework for Enabling and Evaluating Human-Agent Collaborationcited by 60
  4. 2024The Widening Gap: The Benefits and Harms of Generative AI for Novice Programmerscited by 46
  5. 2025What Has a Foundation Model Found? Using Inductive Bias to Probe for World Modelscited by 46
  6. 2025Future of Work with AI Agents: Auditing Automation and Augmentation Potential across the U.S. Workforcecited by 33
  7. 2025How Do AI Agents Do Human Work? Comparing AI and Human Workflows Across Diverse Occupationscited by 17
  8. 2025Potemkin Understanding in Large Language Modelscited by 14
  9. 2025Can LLM-Simulated Practice and Feedback Upskill Human Counselors? A Randomized Study with 90+ Novice Counselorscited by 13
  10. 2022Pragmatics in Language Grounding: Phenomena, Tasks, and Modeling Approachescited by 12
  11. 2025AI Safety Should Prioritize the Future of Workcited by 11
  12. 2024GLAT: The Generative AI Literacy Assessment Testcited by 10
6 more
  1. 2026A clinically validated framework for auditing AI chatbot behavior in mental health interactionscited by 9
  2. 2026Invisible failures in human-AI interactionscited by 3
  3. 2025From G-Factor to A-Factor: Establishing a Psychometric Framework for AI Literacycited by 2
  4. 2025Calibrated Trust in Dealing with LLM Hallucinations: A Qualitative Studycited by 2
  5. 2025The Agency Gap in AI-Supported Writing: How Reactive and Proactive Agent Designs Shape Multimodal Reasoningcited by 1
  6. 2026On Emergent Social World Models -- Evidence for Functional Integration of Theory of Mind and Pragmatic Reasoning in Language Modelscited by 1

Built on by

Not cited yet.