Social and epistemological bases of technology transfer : the case of artificial intelligence
This thesis addresses a problem in the literature on technology transfer of understanding the local appropriation of knowledge. Based on interpretive and analytic traditions developed in Science and Technology Studies (STS) and ethnomethodology, I conceptualise technology transfer as involving communication between discursive communities. I develop the idea of 'performance of community' to argue that explanations of research and technology, and readings of those explanations, are sites for the elaboration of the identity of a discursive community. I explore this approach through a case study in the field of artificial intelligence (AI). I focus on what I call 'explanatory practices', that is practices of describing, identifying and explaining Al, and trace the differences in these practices, according to location, context and audience. The novelty of my thesis is to show the pervasiveness of performance of community within these explanatory practices, through showing the differences in the claimed identity and significance of Al, associated with different locations, contexts and audiences. I draw out some of the implications of my approach by counterposing it to a theory of technology transfer as the passing of neutral units of information, which I argue is implicit in a complaint made by Al vendors that the Al marketplace had been damaged by overselling or hype. In particular, I show that disclaimers of hype (more than the perpetration of it) had always been associated with the marketing of Al. More generally, my claim is that it is politically important to understand that neutral information is not available even as an ultimate standard, and that the local appropriation of knowledge is not an aberration to be controlled, but a component of both successful and unsuccessful communication between discursive communities.