From promises to practice : unraveling users’ mental models about the role and implications of large language models (LLMs) for individuals and society
Bibliographic record
Abstract
The growing prevalence of large language models (LLMs), such as ChatGPT, calls for a deeper understanding of their role and implications for individuals and society. This study explores users’ mental models to understand the role and implications of LLMs for individuals and society. Through qualitative research using the critical incident technique (CIT), insights were gathered from 76 LLM users, resulting in 234 narratives about their interactions and experiences with LLMs. The findings reveal a nuanced ambivalence, with users comprehending LLMs as offering both significant opportunities and considerable risks. At the individual level, users’ mental models depict LLMs as helpful assistants, tools for personal connection, conditional partners, suspicious and disempowering actors, and dangerous opponents. At the societal level, users hold mental models that characterize LLMs as societal aids, conditional allies, disruptors, ethical hazards, threats to human sociability, and forces dividing society. This study identifies the mental models users hold regarding LLMs, providing valuable insights that can inform human-centered AI design and robust regulatory frameworks. By uncovering these mental models, this research lays the groundwork for balancing the opportunities and risks of LLMs. It contributes to the AI adoption and use literature by offering practical recommendations for designing and integrating AI-enabled services into everyday life in a way that aligns with users' mental models.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.031 | 0.056 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.003 | 0.001 |
| Science and technology studies | 0.006 | 0.028 |
| Scholarly communication | 0.017 | 0.028 |
| Open science | 0.002 | 0.011 |
| Research integrity | 0.004 | 0.005 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".