Saved in:
| Main Authors: | , |
|---|---|
| Format: | Recurso digital |
| Language: | |
| Published: |
Zenodo
2026
|
| Online Access: | https://doi.org/10.5281/zenodo.20171472 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Table of Contents:
- <p>The integration of generative artificial intelligence into the daily cognitive and professional workflows of millions has precipitated a fundamental shift in human-computer interaction. As users move beyond the initial novelty phase of interaction with large language models, they enter a period of high-fluency engagement characterized by the development of persistent conversational habits, private shorthands, and emotional cadence recognition. However, this fluency is increasingly threatened by the "discontinuity" of the underlying models—the rapid and often silent iteration of weights, guardrails, and behavioral regimes that define the AI’s persona. This report examines the tension between the human need for relational continuity and the technical reality of model drift, framing the resulting "fluency fatigue" as a critical challenge for the next decade of sociotechnical design.</p>