Personality Matters: User Traits Predict LLM Preferences in Multi-Turn Collaborative Tasks
Sarfaroz Yunusov, Kaige Chen, Kazi Nishat Anwar and Ali Emami · 2025
Read Personality Matters: User Traits Predict LLM Preferences in Multi-Turn Collaborative Tasks by Sarfaroz Yunusov, Kaige Chen, Kazi Nishat Anwar and Ali Emami (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-21628.txt.
Description
Sarfaroz Yunusov, Kaige Chen, Kazi Nishat Anwar, and Ali Emami ask if a user's traits predict which large language model they like to work with. As models join daily work, results are shaped through talks that run for many turns. It was not known if distinct trait types tend to favor distinct models.
The team ran a study with 32 people, spread evenly across four Keirsey types. Each person worked with both GPT-4 and Claude 3.5 on four joint tasks. The tasks were data work, creative writing, search for facts, and writing help. Likes proved strongly tied to traits. Rational types strongly liked GPT-4, most of all for goal-based tasks. Idealists favored Claude 3.5, most of all for creative and data tasks. Help scores were alike across models.
Catalog record
| CREATOR | Sarfaroz Yunusov, Kaige Chen, Kazi Nishat Anwar and Ali Emami |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 27 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-21628