The AI Memory Gap: Users Misremember What They Created With AI or Without
Tim Zindulka, Sven Goller, Daniela Fernandes, Robin Welsch, Daniel Buschek
Abstract
Ideas: For a given problem, come up with 5 solution ideas (as 1-3 keywords each).
On your own [noAI] or with a chatbot [withAI] Elaborations: Write one sentence about each idea. On your own [noAI] or with a chatbot [withAI] Phase 2 For each problem + solution text: • Did you work on this? [yes / no] • Source of idea? [you / AI] • Source of elaboration? [you / AI] ... and also for distractors (unseen problems/solutions) one week later... Negative impact of AI on source memory overall Mixed workflows harder to remember than never/always using AI People tend to be overconfident about their own performance
How might we solve the problem of plastic waste in the ocean? ( ) Figure 1: Overview of our experiment and findings: In Phase 1 (left) participants were presented with problems and asked to come up with five ideas each (as 1-3 keywords per idea), with or without assistance by a chatbot (alternating across problems).
They then wrote on these ideas (1 sentence per idea), again with or without AI. After a week, Phase 2 (centre) presented the problems and solutions again, asking participants if they worked on this (item memory), and if so, what the source was for both idea and elaboration (source memory). We also showed distractors (unseen problems/solutions). The key findings (right) show that AI had a negative impact, mixed workflows were particularly difficult, and people tended to be overconfident in their own performance.
Ask about this paper
Your agent reads all of it.
Lune indexed this paper to the last equation, along with the top-tier papers that cite it. Ask a question and the answer quotes them.
Your agent calls
Luneget_paper_fulltext
Free to start. No credit card required.
Terminal
Install the CLIlune papers fulltext 981376c8-9622-4553-87fb-9f69be0be409Builds on24
- CoAuthor: Designing a Human-AI Collaborative Writing Dataset for Exploring Language Model CapabilitiesMina Lee, Percy Liang, Qian YangCHI 2022 · 340 citations
- The Metacognitive Demands and Opportunities of Generative AILev Tankelevitch, Viktor Kewenig, Auste Simkute, Ava Elizabeth Scott et al.CHI 2024 · 279 citations
- Co-Writing with Opinionated Language Models Affects Users' ViewsMaurice Jakesch, Advait Bhat, Daniel Buschek, Lior Zalmanson et al.CHI 2023 · 249 citations
- Design Principles for Generative AI ApplicationsJustin D. Weisz, Jessica He, Michael J. Muller, Gabriela Hoefer et al.CHI 2024 · 221 citations
- The Effects of Generative AI on Design Fixation and Divergent ThinkingSamangi Wadinambiarachchi, Ryan M. Kelly, Saumya Pareek, Qiushi Zhou et al.CHI 2024 · 186 citations
Related papers
- IdeaBot: Investigating Social Facilitation in Human-Machine Team CreativityAngel Hsing-Chi Hwang, Andrea Stevenson WonCHI 2021 · 76 citations
- Exploring The Impact of Proactive Generative AI Agent Roles In Time-Sensitive Collaborative Problem-Solving TasksAnirban Mukhopadhyay, Kevin Salubre, Hifza Javed, Shashank Mehrotra et al.CHI 2026 · 2 citations
- Timing Matters: How Using LLMs at Different Timings Influences Writers' Perceptions and Ideation Outcomes in AI-Assisted IdeationPeinuan Qin, Chi-Lan Yang, Jingshu Li, Jing Wen et al.CHI 2025 · 19 citations
- Semantic Commit: Helping Users Update Intent Specifications for AI Memory at ScalePriyan Vaithilingam, Munyeong Kim, Frida-Cecilia Acosta-Parenteau, Daniel Lee et al.UIST 2025 · 5 citations
- Navigating Neurodivergence with AI Chatbots: Benefits, Tensions, and Implications for HCIDeepak Giri, Erin Brady, Megh MaratheCHI 2026 · 1 citation
