Rationalize: Shared Semantic Reasoning for Human-AI Alignment

2026-05-28

AI Summary

This study addresses the misalignment between humans and AI in data-driven cognition, particularly regarding semantic understanding and reasoning intent. To bridge this gap, the authors propose a bidirectional alignment framework grounded in complementary Explorer-Guide roles. The framework enables both parties to explicitly articulate goals, hypotheses, evidence, and inferences within a shared reasoning space, distinguishing mechanisms for “AI aligning to humans” and “humans aligning to AI.” It integrates theories of human–machine collaboration, models of critical thinking, and the semantic reasoning capabilities of large language models. The work contributes an interpretable and collaborative paradigm for cognitive alignment and advocates for fine-grained, role-specific evaluation metrics, thereby advancing a new agenda for the design and assessment of human–AI collaborative alignment.

Abstract

We introduce Rationalize, a role-pair framework for shared semantic reasoning between humans and AI models in data-driven sensemaking. Building on ideas in human-machine teaming and critical thinking, we conceptualize human-AI interaction as a series of complementary role pairs (Explorer-Guide, Investigator-Informant, Teacher-Student, Judge-Advocate) operating in a shared reasoning space. In this space, human analysts and AI models (such as LLMs) make purposes, questions, assumptions, evidence, inferences, and implications explicit, facilitating alignment not only at the output level but at the level of rationalization of intent and action by each side. We relate these role pairs to the bidirectional human-AI alignment framework, illustrating how "aligning AI to humans" and "aligning humans to AI" differ by role, and sketch a collaborative research agenda for alignment design and assessment using element-level and role-specific approaches.

Problem

• human-AI alignment
• shared semantic reasoning
• rationalization
• role-pair framework
• sensemaking

Innovation

• shared semantic reasoning
• role-pair framework
• human-AI alignment
• rationalization
• collaborative sensemaking