规范性冲突与浅层AI对齐
Normative Conflicts and Shallow AI Alignment
AI总结：
本文指出LLMs的对齐方法仅强化浅层行为倾向，缺乏规范性审慎能力，因而易受利用规范冲突的对抗攻击，并强调该浅层对齐问题对AI安全与监管的深远影响。
摘要：
大型语言模型（LLMs）等AI系统的进展引发了对其安全部署日益紧迫的担忧。本文考察了LLMs的价值对齐问题，认为当前的对齐策略在防止滥用方面存在根本性不足。尽管通过基于人类偏好的微调，人们持续努力向LLMs灌输有益性、诚实性和无害性等规范，但它们仍然容易受到利用这些规范之间冲突的对抗性攻击。我认为，这一脆弱性反映了现有对齐方法的一个根本局限：它们强化的是浅层行为倾向，而非赋予LLMs真正的规范性审慎能力。借鉴道德心理学的研究，我展示了人类进行审慎推理的能力如何增强他们面对类似对抗策略时的韧性。相比之下，LLMs缺乏检测并理性解决规范性冲突的稳健能力，使其容易受到操纵；即便是近期以推理为重点的LLMs进展也未能解决这一脆弱性。这一“浅层对齐”问题对AI安全与监管具有重要影响，表明当前方法不足以缓解日益强大的AI系统可能带来的潜在危害。
英文摘要：
The progress of AI systems such as large language models (LLMs) raises increasingly pressing concerns about their safe deployment. This paper examines the value alignment problem for LLMs, arguing that current alignment strategies are fundamentally inadequate to prevent misuse. Despite ongoing efforts to instill norms such as helpfulness, honesty, and harmlessness in LLMs through fine-tuning based on human preferences, they remain vulnerable to adversarial attacks that exploit conflicts between these norms. I argue that this vulnerability reflects a fundamental limitation of existing alignment methods: they reinforce shallow behavioral dispositions rather than endowing LLMs with a genuine capacity for normative deliberation. Drawing from on research in moral psychology, I show how humans' ability to engage in deliberative reasoning enhances their resilience against similar adversarial tactics. LLMs, by contrast, lack a robust capacity to detect and rationally resolve normative conflicts, leaving them susceptible to manipulation; even recent advances in reasoning-focused LLMs have not addressed this vulnerability. This ``shallow alignment'' problem carries significant implications for AI safety and regulation, suggesting that current approaches are insufficient for mitigating potential harms posed by increasingly capable AI systems.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
