ValuePilot: A Two-Phase Framework for Value-Driven Decision-Making
This work addresses the problem of personalized AI decision-making for users in scenarios not seen during training, though it is a preliminary exploration and thus incremental in nature.
The paper tackles the challenge of ensuring personalized decision-making in AI for tasks not covered in training datasets by proposing ValuePilot, a two-phase framework that generates scenarios based on value dimensions and learns to make decisions aligning with human preferences, outperforming models like Claude-3.5-Sonnet and GPT-4o in matching human decisions.
Despite recent advances in artificial intelligence (AI), it poses challenges to ensure personalized decision-making in tasks that are not considered in training datasets. To address this issue, we propose ValuePilot, a two-phase value-driven decision-making framework comprising a dataset generation toolkit DGT and a decision-making module DMM trained on the generated data. DGT is capable of generating scenarios based on value dimensions and closely mirroring real-world tasks, with automated filtering techniques and human curation to ensure the validity of the dataset. In the generated dataset, DMM learns to recognize the inherent values of scenarios, computes action feasibility and navigates the trade-offs between multiple value dimensions to make personalized decisions. Extensive experiments demonstrate that, given human value preferences, our DMM most closely aligns with human decisions, outperforming Claude-3.5-Sonnet, Gemini-2-flash, Llama-3.1-405b and GPT-4o. This research is a preliminary exploration of value-driven decision-making. We hope it will stimulate interest in value-driven decision-making and personalized decision-making within the community.