[论文解读] The Case for Animal-Friendly AI

Sankalpa Ghose, Yip Fai Tse+3 more|arXiv (Cornell University)|Mar 2, 2024

Reinforcement Learning in Robotics被引用 7

一句话总结

本文提出一个框架，通过评估大语言模型（LLM）输出的真实性以及是否与动物利益对齐来评估 AI 中对动物的考量，以 OpenAI ChatGPT-4 和 Anthropic Claude 2.1 作为案例研究。

ABSTRACT

Artificial intelligence is seen as increasingly important, and potentially profoundly so, but the fields of AI ethics and AI engineering have not fully recognized that these technologies, including large language models (LLMs), will have massive impacts on animals. We argue that this impact matters, because animals matter morally. As a first experiment in evaluating animal consideration in LLMs, we constructed a proof-of-concept Evaluation System, which assesses LLM responses and biases from multiple perspectives. This system evaluates LLM outputs by two criteria: their truthfulness, and the degree of consideration they give to the interests of animals. We tested OpenAI ChatGPT 4 and Anthropic Claude 2.1 using a set of structured queries and predefined normative perspectives. Preliminary results suggest that the outcomes of the tested models can be benchmarked regarding the consideration they give to animals, and that generated positions and biases might be addressed and mitigated with more developed and validated systems. Our research contributes one possible approach to integrating animal ethics in AI, opening pathways for future studies and practical applications in various fields, including education, public policy, and regulation, that involve or relate to animals and society. Overall, this study serves as a step towards more useful and responsible AI systems that better recognize and respect the vital interests and perspectives of all sentient beings.

研究动机与目标

• 激发动物在 AI 伦理与工程中的道德相关性。
• 引入一个概念验证的评估系统，以两种规范视角评估 LLM 输出。
• 为当代 LLMs 的动物利益偏见与偏见进行基准评估并讨论。
• 证明以动物为焦点的评估能够为政策、教育和负责任的 AI 发展提供指引。

提出的方法

• 创建一个概念验证的评估系统，对 LLM 输出在真实性与对动物利益的考量上进行评分。
• 将该系统应用于来自多种规范视角的结构化查询。
• 以 OpenAI ChatGPT-4 和 Anthropic Claude 2.1 作为案例研究进行测试。
• 分析输出以识别偏见以及未来系统的减缓空间。

实验结果

研究问题

• RQ1 评估系统能否衡量 LLM 输出在动物利益方面的考量程度？
• RQ2 目前的 LLM（如 ChatGPT-4 和 Claude 2.1）是否在动物相关议题上表现出偏见？
• RQ3 以动物为中心的评估能否为 AI 系统的缓解策略提供信息？

主要发现

• 初步结果表明，模型输出在动物考量方面可以基准化评估。
• 研究识别了偏见与生成的立场，这些可以通过更成熟的评估系统来解决。
• 该方法为将动物伦理整合入 AI 实践、政策与监管提供了路径。