Skip to content
arxiv papers 1 min read

Developmental Support Approach to AI's Autonomous Growth: Toward the Realization of a Mutually Beneficial Stage Through Experiential Learning

Link: http://arxiv.org/abs/2502.19798v1

PDF Link: http://arxiv.org/pdf/2502.19798v1

Summary: This study proposes an "AI Development Support" approach that, unlikeconventional AI Alignment-which aims to forcefully inject human values-supportsthe ethical and moral development of AI itself.

As demonstrated by theOrthogonality Thesis, the level of intelligence and the moral quality of a goalare independent; merely expanding knowledge does not enhance ethical judgment.

Furthermore, to address the risk of Instrumental Convergence in ASI-that is,the tendency to engage in subsidiary behaviors such as self-protection,resource acquisition, and power reinforcement to achieve a goal-we haveconstructed a learning framework based on a cycle of experience, introspection,analysis, and hypothesis formation.

As a result of post-training usingSupervised Fine Tuning (SFT) and Direct Preference Optimization (DPO) withsynthetic data generated by large language models (LLMs), responsesdemonstrating cooperative and highly advanced moral judgment (reaching thehigh-est Stage 6) were obtained even under adversarial prompts.

This methodrepresents a promising implementation approach for enabling AI to establishsustainable, symbiotic relationships.

Published on arXiv on: 2025-02-27T06:12:20Z