Shaping Human-AI Interactions to Provide Improvement Pathways and Balance Competing Objectives
Keziah Naggita
Why It Matters
What makes this one worth your time
Understanding and improving human-AI interactions is crucial for developing systems that are both effective and aligned with user needs, which is increasingly relevant in various AI applications.
The paper explores how to optimize human-AI interactions for better outcomes and system alignment.
Summary
The thesis investigates the design of human-AI interactions to foster accurate beliefs about AI systems, discourage gaming behaviors, and ensure AI systems meet their intended objectives, employing a mix of theoretical analysis and empirical evaluations.
Key contributions
- Development of principles for designing AI systems that align with human needs and values.
- Methodological framework combining theoretical analysis, data-driven modeling, and human-subject experiments.
Notable insights
- The integration of both evaluated individuals' perspectives and AI system objectives offers a holistic approach to interaction design.
- The use of real-world and semi-synthetic datasets for empirical evaluation may provide insights into practical applicability.
Possible limitations
- Not stated in the abstract.
Abstract
arXiv:2608.05710v1 Announce Type: new Abstract: When an AI system is deployed, the individuals who use and or are evaluated by it form beliefs about how the system operates and use those beliefs to strategically present their preferences, behaviors, or attributes. The system then responds with feedback or a decision outcome, thereby creating a human-AI interaction loop. This thesis studies how to design and shape such interactions to achieve three goals: (1) help individuals develop accurate beliefs about the AI systems so they can improve and or secure favorable outcomes at minimal cost, (2) encourage improvement and or discourage gaming behaviors, and (3) ensure that the AI system continues to achieve its intended objectives, such as maximizing accuracy. To address these goals, the thesis is organized into three complementary parts that examine and study human-AI interactions from the perspectives of both evaluated individuals and AI systems. Together, the work presented in this thesis advances human-centered machine learning by providing principles and methods for designing AI systems that align with human needs, values, and capabilities. Methodologically, this thesis integrates theoretical analysis, data-driven modeling, human-subject experiments, and empirical evaluations on real-world and semi-synthetic datasets.