Dr ALAN CHAMBERLAIN alan.chamberlain@nottingham.ac.uk
Principal Research Fellow
Elena Nichele
Mads Bødker
Tommaso Turchi
Abstract
When humans and machines start to chat: beyond anthropocentrism
Digital and embedded artificial intelligent (AI) agents with conversational capabilities have gained significant attention in recent years [1, 2]. Using natural language communication, such systems enable an extensive range of human–machine collaborations and interaction such as information retrieval, customer services, problem-solving, and programming. Moving away from the classical paradigm of interaction in which humans interact with intelligent-but-passive systems, humans today can collaborate with in silico agents that provide designed and data-driven forms of agency and intentionality [3]. Such AI-driven agents can perform tasks autonomously or suggest how to perform tasks in (more) efficient and effective ways, both influencing and being influenced by human users. This shift toward active agents introduces complexities and potential drawbacks, including inherent biases within these systems.
From an analytical point of view, practitioners might decide to overlook this shift by adopting Actor-Network Theory or ANT [4], which assumes a substantial symmetry among all elements within a network. According to ANT, and related theories, such as distributed cognition or sociomateriality, both human and non-human actors hold equal significance, with agency distributed across the entire system. This perspective aligns with the move toward more-than-human and non-anthropocentric frameworks, promoting a balanced consideration of all actors involved in human–machine interaction [5, 6]. Such perspectives focus on mapping relationships and establishing causation and effects among all elements/actors in complex systems [4, 7], for example, in order to produce powerful design changes and experiences by promoting participation, offering multiple perspectives and taking into account multi-species co-design [8, 9].
Citation
Borsci, S., Chamberlain, A., Nichele, E., Bødker, M., & Turchi, T. (2024). Human AI conversational systems: when humans and machines start to chat. Personal and Ubiquitous Computing, 28(6), 857–860. https://doi.org/10.1007/s00779-024-01837-1
