[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-seo-572589-105":3,"detail-sidebar-cat-1-en-105":80,"doc-detail-572589-en":126},{"code":4,"msg":5,"data":6},0,"ok",{"site_id":7,"language":8,"slug":9,"title":10,"keywords":11,"description":12,"schema_data":13,"social_meta":73,"head_meta":75,"extra_data":77,"updated_unix":79},105,"en","confidence-without-competence-in-ai-assisted-knowledge-work","Confidence Without Competence in AI-Assisted Knowledge Work","","Large Language Models (LLMs) are widely used by students, yet their tendency to provide fast and complete answers may discourage reflection and foster overconfidence. This study examines how alternative LLM interaction designs support deeper thinking without overly increasing cognitive burden. A two-phase mixed-methods evaluation used 16 interviews to design Deep3 and 85 participants to test it across learning tasks. Results show effort, confidence, and learning can diverge.",{"@graph":14,"@context":72},[15,34,55],{"@type":16,"itemListElement":17},"BreadcrumbList",[18,23,27,31],{"item":19,"name":20,"@type":21,"position":22},"https://docshare.wps.com","Home","ListItem",1,{"item":24,"name":25,"@type":21,"position":26},"https://docshare.wps.com/template/","Template",2,{"item":28,"name":29,"@type":21,"position":30},"https://docshare.wps.com/template/presentations/","Presentations",3,{"item":32,"name":10,"@type":21,"position":33},"https://docshare.wps.com/template/confidence-without-competence-in-ai-assisted-knowledge-work/572589/",4,{"url":32,"name":10,"@type":35,"image":36,"author":41,"headline":10,"publisher":44,"fileFormat":47,"inLanguage":8,"description":12,"dateModified":48,"datePublished":49,"encodingFormat":47,"isAccessibleForFree":50,"interactionStatistic":51},"DigitalDocument",{"url":37,"@type":38,"width":39,"height":40},"https://docshare.wps.com/thumbnails/confidence-without-competence-in-ai-assisted-knowledge-work/572589.png","ImageObject",442,249,{"name":42,"@type":43},"hdkawk","Person",{"url":19,"name":45,"@type":46},"DocShare","Organization","application/pdf","2026-10-22","2026-10-13",true,{"@type":52,"interactionType":53,"userInteractionCount":30},"InteractionCounter",{"@type":54},"ViewAction",{"@type":56,"mainEntity":57},"FAQPage",[58,64,68],{"name":59,"@type":60,"acceptedAnswer":61},"Why can LLMs lead to overconfidence in knowledge work?","Question",{"text":62,"@type":63},"LLMs often produce fast and complete answers, which can reduce reflection and encourage users to trust outputs without sufficiently evaluating them. The study frames this as an illusion of competence that harms critical reasoning.","Answer",{"name":65,"@type":60,"acceptedAnswer":66},"What is Deep3 and how does it change the LLM interaction?",{"text":67,"@type":63},"Deep3 is a web-based system designed from interviews with 16 Gen Z students. It provides three interaction modes: future-self explanations, contrastive learning, and guided hints.",{"name":69,"@type":60,"acceptedAnswer":70},"How did the interaction modes affect learning outcomes and perceived understanding?",{"text":71,"@type":63},"A standard single-agent baseline produced high perceived understanding but the lowest objective learning. Future-self explanations increased cognitive workload while best aligning perceived and actual understanding, and guided hints delivered the largest learning gains with limited additional frustration.","https://schema.org",{"og:url":32,"og:type":74,"og:title":10,"og:site_name":45,"og:description":12},"article",{"robots":76,"canonical":32},"index,follow",{"doc_id":78,"site_id":7},572589,1792434977,{"code":4,"msg":81,"data":82},"success",[83,87,92,97,102,107,112,117,122],{"id":84,"doc_module":22,"doc_module_name":25,"category_name":29,"show_sort_weight":85,"slug":86},11,90,"presentations",{"id":88,"doc_module":22,"doc_module_name":25,"category_name":89,"show_sort_weight":90,"slug":91},12,"Resumes",80,"resumes",{"id":93,"doc_module":22,"doc_module_name":25,"category_name":94,"show_sort_weight":95,"slug":96},14,"Invoices",70,"invoices",{"id":98,"doc_module":22,"doc_module_name":25,"category_name":99,"show_sort_weight":100,"slug":101},15,"Posters",60,"posters",{"id":103,"doc_module":22,"doc_module_name":25,"category_name":104,"show_sort_weight":105,"slug":106},16,"Social Media",50,"social-media",{"id":108,"doc_module":22,"doc_module_name":25,"category_name":109,"show_sort_weight":110,"slug":111},17,"Forms",40,"forms",{"id":113,"doc_module":22,"doc_module_name":25,"category_name":114,"show_sort_weight":115,"slug":116},18,"Letters",30,"letters",{"id":118,"doc_module":22,"doc_module_name":25,"category_name":119,"show_sort_weight":120,"slug":121},21,"Paper Templates",5,"papers-templates",{"id":123,"doc_module":22,"doc_module_name":25,"category_name":124,"show_sort_weight":4,"slug":125},158,"General","general-158",{"code":4,"msg":81,"data":127},{"doc_id":78,"user_id":128,"nickname":42,"user_avatar":129,"doc_module":22,"category_id":84,"category_name":29,"doc_title":10,"doc_description":12,"doc_content":130,"file_id":131,"file_url":132,"file_type":133,"file_size":134,"view_count":30,"is_deleted":4,"is_public":22,"is_downloadable":22,"audit_status":22,"page_count":135,"language":136,"language_code":8,"site_id":7,"html_lang":8,"table_of_contents":137,"faqs":138,"seo_title":139,"seo_description":12,"update_tm":140,"read_time":141},3985747859855,"https://ap-avatar.wpscdn.com/davatar_276721f389ce27ea32af1340a28f341c","Confidence Without Competence in AI-Assisted Knowledge Work  \nElena Eleftheriou  \nUniversity of Cyprus Nicosia, Cyprus [eleftheriou.elena@ucy.ac.cy](eleftheriou.elena@ucy.ac.cy)  \nGeorge Pallis  \nUniversity of Cyprus Nicosia, Cyprus [pallis.george@ucy.ac.cy](pallis.george@ucy.ac.cy)  \nMarios Constantinides∗ CYENS Centre of Excellence Nicosia, Cyprus  \n[marios.constantinides@cyens.org.cy](marios.constantinides@cyens.org.cy)  \narXiv :2604 .09444v1 [ cs .HC] 10 Apr 2026  \nFigure 1: Overview of our study and findings, which illustrate (A) the two-phase study methodology; (B) the three conditions developed from the design requirements gathered in Phase 1; (C) our research questions; and (D) the resulting findings.  \nAbstract  \nLarge Language Models (LLMs) are widely used by students, yet their tendency to provide fast and complete answers may discourage reflection and foster overconfidence. We examined how alternative LLM interaction designs support deeper thinking without excessively increasing cognitive burden. We conducted a two-phase mixed-methods study. In Phase 1, interviews with 16 Gen Z students informed the design of Deep3, a web-based system with three interaction modes: a) future-self explanations, b) contrastive learning, and c) guided hints. In Phase 2, we evaluated Deep3 with 85 participants across two learning tasks. We found that a standard singleagent baseline produced high perceived understanding despite the lowest objective learning. In contrast, future-self explanations imposed higher cognitive workload yet yielded the closest alignment between perceived and actual understanding, while guided hints achieved the largest learning gains without a proportional increase in frustration. These findings show that effort, confidence, and learning systematically diverge in LLM-supported work.  \n∗ Also affiliated with University of Cyprus, Cyprus and University College London, UK.  \nPermission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than the author(s) must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission [and/or a fee. Request permissions from permissions@acm.org](and/or a fee. Request permissions from permissions@acm.org).  \nCHIWORK’26, Linz, Austria  \n© 2026 Copyright held by the owner/author(s) . Publication rights licensed to ACM. ACM ISBN 978-1-4503-XXXX-X/2018/06  \n[https://doi.org/XXXXXXX.XXXXXXX](https://doi.org/XXXXXXX.XXXXXXX)  \nCCS Concepts  \n• Human-centered computing → Empirical studies in HCI; HCI design and evaluation methods.  \nKeywords  \nlarge language models, learning, education, critical thinking  \nACM Reference Format:  \nElena Eleftheriou, George Pallis, and Marios Constantinides. 2026. Confidence Without Competence in AI-Assisted Knowledge Work. In Proceedings of CHIWORK’26: Proceedings of the 5th Annual Symposium on HumanComputer Interaction for Work (CHIWORK’26). ACM, New York, NY, USA, 25 pages. [https://doi.org/XXXXXXX.XXXXXXX](https://doi.org/XXXXXXX.XXXXXXX)  \n1 Introduction  \nThe rapid integration of LLMs into higher education has fundamentally changed the academic environment for Generation Z (Gen Z) and emerging young professionals. As a “digital-first” generation, these students have quickly adopted tools like ChatGPT to speed up information retrieval, note-taking, and problem-solving, treating generative AI (GenAI) as “the new calculator” for the modern era [59] . While recent studies show that nearly all Gen Z students utilize these tools for tasks ranging from note-taking to test preparation [3, 50], this adoption has created a generational divide. Students embrace the efficiency of AI, while ed","cbCaifIhTRmjR9YL","https://ap.wps.com/l/cbCaifIhTRmjR9YL","pdf",6282709,25,"English","# Introduction\n# Abstract\n# Keywords","[{\"question\":\"Why can LLMs lead to overconfidence in knowledge work?\",\"answer\":\"LLMs often produce fast and complete answers, which can reduce reflection and encourage users to trust outputs without sufficiently evaluating them. The study frames this as an illusion of competence that harms critical reasoning.\"},{\"question\":\"What is Deep3 and how does it change the LLM interaction?\",\"answer\":\"Deep3 is a web-based system designed from interviews with 16 Gen Z students. It provides three interaction modes: future-self explanations, contrastive learning, and guided hints.\"},{\"question\":\"How did the interaction modes affect learning outcomes and perceived understanding?\",\"answer\":\"A standard single-agent baseline produced high perceived understanding but the lowest objective learning. Future-self explanations increased cognitive workload while best aligning perceived and actual understanding, and guided hints delivered the largest learning gains with limited additional frustration.\"}]","Confidence Without Competence in AI-Assisted Knowledge Work | PDF",1791928298,9]