過去社會學中的重要主題,大多關注的是行動者與社會的互動,無論是將二者切開來看,或者是將二者視為一種辯證關係,或者是強調二者的相互獨立且相互依賴性,其前提似乎都是將二者視為各別獨立存在的實體,劃分二者的那條線似乎一直都存在。許多具有社會學意涵的主題,例如性別、族群、種族、勞動、資本主義、社會正義、團體、組織等,也都是在此一框架下來進行論述。這樣的論述之所以依然有效,部分是因為人類至今還是一個獨特的物種,無論是在思想或行為上,都有別於其他生物有機體。人類並未真正與其他可與之媲美的智能進行互動。然而,人類智能(human intelligence,以下稱HI)的獨特性,其中很大部分表現為具有一定程度的彈性或可塑性,其會受到環境的影響,這個獨特性可以說是透過人與非人(其他有機體以及技術物件等)的互動所共構出來。從科技人類學的觀點來看,這個獨特性便是「人類將自身打造成「人類」這個物種的過程」,換句話說就是「人之所以為人」的性質(例如,Stiegler, 1998)。然而,此一情況在1960年代開始有了些許變化——尤其是在學術研究場域。當人類開始想像有可能打造出一種具有跟人類一樣思考與行為能力的機器時,「人之所以為人」的獨特性也開始受到挑戰。這個挑戰或許也對行動者與社會之間的那條界線提出了新的問題。本研究將透過俗民方法學(ethnomethodology)的觀點及方法,將人工智能(artificial intelligence,以下稱AI)重構為一種社會現象,進而探究如何可能將其放在社會學理論中來進行討論。在實證上,本研究將以具有人工智能形式的聊天機器人(AI-based chatbot)為例,透過俗民方法學中的對話分析法,以錄影的方式紀錄HI與AI聊天機器人的對話過程,藉由分析該對話內容,進一步探究人類與另一種同樣被視為具有所謂「智能」的非人之互動樣貌,並且提供對社會學觀點如何探討人工智能的未來可能性。 The research project will focus on the topic of artificial intelligence and its relation to sociological theory. There is no doubt that human intelligence is still unique in almost every way, comparing to other organisms. The uniqueness of human intelligence presents itself in some aspects such like adaptability, plasticity, and creativity. Through perspectives from STS, whose contributions to sociological studies of science and technology, this uniqueness emerges from the interactions between human and non-human objects. Things may have been changed since 1960s when people not only start to imagine that they can build a machine thinking like humans, but also realize that they are capable of doing it. “What it means to be human” became a challenge to human beings. This challenge also questions the boundary between social actor and society, which has been a traditional and presumptuous idea in sociological theory. The research project will adopt viewpoint from ethnomethodology and reconstruct artificial intelligence as a social phenomenon, and put it into discussion of sociological theory. In practice, the research project will design a situation in order to observe and record how people converse with AI-based chatbot. With ethnomethodology and conversation analysis approach, the project expects to investigate how two kinds of intelligence interact and what it may reframe the oldest questions of sociological theory: “what it means to be human" and "what it means to be social.”