# Triangulating on Possible Futures: Conducting User Studies on Several Futures Instead of Only One

**Authors:** Antti Salovaara, Leevi Vahvelainen  
**arXiv:** [2409.14137](https://arxiv.org/abs/2409.14137)  
**Date:** 2024-09-21  
**Summary page:** https://www.summarizepaper.com/en/arxiv-id/2409.14137v2/

## Key points

- Researchers aimed to explore the use of user studies in investigating two potential futures for AI-augmented knowledge work
- Thirteen participants with backgrounds in entrepreneurship and business were recruited for the study
- Participants represented diverse nationalities and ranged in age from 10 to 31
- Statistical tests were used to analyze differences in attitudes towards prototypes and coping mechanisms for each future scenario
- Participants engaged with tasks and envisioned a future where AI-assisted evaluations could impact stock markets
- Quantitative analyses showed more positive attitudes towards one prototype over the other, but no significant differences in other examined constructs
- Qualitatively, participants found one future less threatening based on observed coping mechanisms during the study
- Research aimed to disentangle specific details from generalizable findings by presenting contrasting scenarios with specific AI elements into each future.

## Summary

In this study, researchers aimed to explore the use of user studies in investigating two potential futures for AI-augmented knowledge work. Thirteen participants with backgrounds in entrepreneurship and business were recruited to evaluate futuristic prototypes through both quantitative analyses and qualitative feedback. The participants represented diverse nationalities and ranged in age from 10 to 31. The study utilized statistical tests to analyze differences in attitudes towards the prototypes and coping mechanisms for each presented future. While some participants expressed skepticism about evaluating financial impacts based on isolated news articles, all were able to engage with the tasks and envision a future where AI-assisted evaluations could impact stock markets. Quantitative analyses showed that participants had more positive attitudes towards one prototype over the other, but there were no significant differences in other examined constructs. Qualitatively, participants found one future less threatening than the other based on observed coping mechanisms during the study. Overall, the research aimed to disentangle specific details from generalizable findings by presenting contrasting scenarios and incorporating specific AI elements into each future. This approach not only made HCI studies more resilient against changing technologies but also allowed for reflective anticipation of potential futures in AI-augmented knowledge work.