With ChatGPT’s rapid adoption, concerns regarding generative artificial intelligence (AI) have shifted from theoretical to practical. Drawing upon the “algorithmic imaginary” framework from critical algorithm studies and the anthropological concept of “ordinary ethics,” we analyzed Twitter discourse during ChatGPT’s initial deployment, examining 368,359 tweets. Our analysis identified five topics reflecting functional and critical aspects of ChatGPT. We specifically point to two topics with a critical perspective: “Ethics” and “Concerns.” The first aligns with scholarly discussions in AI ethics on fairness and transparency, while the second focuses on ChatGPT’s generative capabilities. This highlights an emerging trend: While the academic discussion on AI ethics has gained popularity, especially in scrutinizing ChatGPT, the conversation is now expanding to more nuanced ethical deliberations. We analyzed the posts’ engagement and sentiment over time, demonstrating the AI ethics community’s influence in addressing the potential and harms of generative AI systems.

中文翻译：

---

公众眼中的 ChatGPT：社交媒体讨论中的道德原则和生成性关注


随着 ChatGPT 的快速采用，对生成人工智能 (AI) 的关注已从理论转向实践。借鉴关键算法研究中的“算法想象”框架和“普通伦理”的人类学概念，我们分析了 ChatGPT 初始部署期间的 Twitter 话语，检查了 368,359 条推文。我们的分析确定了反映 ChatGPT 功能和关键方面的五个主题。我们特别指出两个具有批判性视角的主题：“道德”和“担忧”。第一个与人工智能伦理中关于公平和透明度的学术讨论相一致，而第二个则侧重于 ChatGPT 的生成能力。这凸显了一个新兴趋势：虽然关于人工智能伦理的学术讨论已经流行起来，尤其是在审查 ChatGPT 方面，但对话现在正在扩展到更细致的伦理审议。我们分析了这些帖子随着时间的推移的参与度和情绪，展示了人工智能道德社区在解决生成人工智能系统的潜力和危害方面的影响力。