Search
Browse By Day
Browse By Person
Browse By Room
Browse By Session Type
Browse By Research Area
Search Tips
Meeting Home Page
Personal Schedule
Sign In
Homes around the world are echoing with conversations between people and their so-called smart voice assistants. Conversing with computers by voice is a mainstay of science fiction, an aspiration of technologists, and has recently been manifest in high profile AI personas Siri, Alexa, Cortana and Google Assistant. In this paper we reflect on the command and conversation relationships established with these systems, and the privileging of speech and voice as supposedly the most natural means of human-computer interaction. This paper draws on cultural history, textual analysis, conversation analysis and technology studies to examine this technocultural form. We analyse the archetypes established by well known conversational computers in cinema from HAL to Her. We examine patents, blog posts, academic books and journal articles on voice user interfaces (VUIs), intelligent agents and the design of personas. VUI design is a new discipline and form of labour involving AI programming and creative authoring, using data and ‘chit chat’ drawn from everyday users. Designers of voice assistants have been influenced by humanities and social sciences work, such as speech act theory, Grice on conversations, Nass and Reeves’ work on ‘the media equation’ and Goffman’s ‘frames’ and ‘social roles’. While designers claim voice is natural, and are localised to national contexts, we suggest these systems remain tailored to dominant cultures, insensitive to speakers with ‘accents’ of all kinds. With only a small number of US corporations dominating intelligent agent services, this creates new forms of cultural imperialism, media concentration and soft power.