Adaptive Alignment: Dynamic Preference Adjustments via Multi-Objective Reinforcement Learning for Pluralistic AI
This addresses the challenge of pluralistic AI alignment for users with varied needs, though it appears incremental as it builds on existing MORL methods without demonstrating new performance gains.
The paper tackles the problem of aligning AI systems with diverse and shifting human preferences by proposing a dynamic approach using Multi-Objective Reinforcement Learning (MORL) for post-learning policy adjustments, but it does not provide concrete results or numbers.
Emerging research in Pluralistic Artificial Intelligence (AI) alignment seeks to address how intelligent systems can be designed and deployed in accordance with diverse human needs and values. We contribute to this pursuit with a dynamic approach for aligning AI with diverse and shifting user preferences through Multi Objective Reinforcement Learning (MORL), via post-learning policy selection adjustment. In this paper, we introduce the proposed framework for this approach, outline its anticipated advantages and assumptions, and discuss technical details about the implementation. We also examine the broader implications of adopting a retroactive alignment approach through the sociotechnical systems perspective.