The Effects of Fine-Tuning on Agentic AI; The Struggle Over Remote Office Work 105 views

How is task management and productivity used in evolving modern professional work? White collar workers face complex tasks in multiple settings that challenge old work models. The pressures require modern tools and a deeper understanding of work environments. While LLMs can perform general tasks well, they lack the specificity that is required for individual workflows. How can AI systems be optimized for specialized task management capabilities beyond basic general-purpose prompting? Fine-tuning can refine models for specialized use with special training. This project fine-tuned a model to decompose large tasks in a web app. Curated input-response pairs composed the training and validation datasets. A study comparing responses from a baseline system prompt to the fine-tuned model showed that fine-tuning was effective; for over 90% percent of use cases participants preferred the responses generated by the fine-tuned model instead of the baseline prompt responses. Hybrid work has emerged as the standard for white-collar work. This paper explores this development by viewing the conflicting narratives under the lens of the Social Construction of Technology framework. Employers pushed for return-to-office mandates, to enforce control. Workers fought for the benefits of remote work with data and social pressure. Technology vendors created platforms for remote work and employer surveillance. Hybrid work emerged from power relations, not efficiency, as employers captured the concession of three days in the office, while workers continued to retain some access to working remotely.

BS (Bachelor of Science)

Fine-tuning; AI; Remote Work; Hybrid Work; SCOT

School of Engineering and Applied Science Bachelor of Science in Computer Science Technical Advisor: Rich Nguyen STS Advisor: Peter Norton

English

All rights reserved by the author (no additional license for public reuse)

2025-12-16