arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2306.03097cs.HCcs.AI
收藏

在杂草之外看见种子:对生成式AI进行绿色组队以发掘有益用途

Seeing Seeds Beyond Weeds: Green Teaming Generative AI for Beneficial Uses

University of Minnesota(明尼苏达大学);Carnegie Mellon University(卡内基梅隆大学)

Logan Stapleton, Jordan Taylor, Sarah Fox, Tongshuang Wu, Haiyi Zhu

AI总结:

本文提出“绿色组队”概念,通过绕过生成式AI内容过滤器支持自杀干预培训、调试教学和政治批判等有益场景,反思过滤机制所嵌入的伤害观与价值观。

摘要:

像GPT和DALL-E这样的大型生成式AI模型(GM)经过训练,可用于生成服务于广泛通用目的的内容。GM的内容过滤器经过泛化,以过滤掉在许多情况下具有伤害风险的内容,例如仇恨言论。然而,被禁止的内容并不总是有害的——在某些情况下,生成被禁止的内容可能是有益的。因此,当GM过滤掉内容时,它们在排除有害用例的同时,也排除了有益用例。哪些用例被排除,反映了嵌入在GM内容过滤中的价值观。近期关于红队(red teaming)的工作提出了绕过GM内容过滤器以生成有害内容的方法。我们创造了“绿色组队”(green teaming)一词,用以描述绕过GM内容过滤器、为有益用例进行设计的方法。我们通过以下方式展示绿色组队:1)将ChatGPT用作虚拟患者,模拟有自杀意念的人,用于自杀支持培训;2)使用Codex有意生成有缺陷的解决方案,以训练学生进行调试;3)考察一个Instagram页面,该页面使用Midjourney生成反LGBTQ+政治人物的变装图像。最后,我们讨论这些用例如何表明绿色组队既是一种实用的设计方法,也是一种批判模式,它对当前关于生成式AI中伤害与价值观的理解提出质疑并加以颠覆。

英文摘要:

Large generative AI models (GMs) like GPT and DALL-E are trained to generate content for general, wide-ranging purposes. GM content filters are generalized to filter out content which has a risk of harm in many cases, e.g., hate speech. However, prohibited content is not always harmful -- there are instances where generating prohibited content can be beneficial. So, when GMs filter out content, they preclude beneficial use cases along with harmful ones. Which use cases are precluded reflects the values embedded in GM content filtering. Recent work on red teaming proposes methods to bypass GM content filters to generate harmful content. We coin the term green teaming to describe methods of bypassing GM content filters to design for beneficial use cases. We showcase green teaming by: 1) Using ChatGPT as a virtual patient to simulate a person experiencing suicidal ideation, for suicide support training; 2) Using Codex to intentionally generate buggy solutions to train students on debugging; and 3) Examining an Instagram page using Midjourney to generate images of anti-LGBTQ+ politicians in drag. Finally, we discuss how our use cases demonstrate green teaming as both a practical design method and a mode of critique, which problematizes and subverts current understandings of harms and values in generative AI.

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。

↑