How do utopian and dystopian visions shape the discourse around the Partnership on Artificial Intelligence?

Abstract

By performing a critical discourse analysis according to Norman Fairclough’s ‘three- dimensional framework’ this research has provided insights into the way Artificial Intelligence (AI) is discursively constructed on the platform Partnership on Artificial Intelligence (POAI), which was set up in September 2016. Firstly, it has been investigated how artificial intelligence is discursively constructed on the platform POAI. Secondly, it has been analysed how AI is discursively constructed by the founding fathers and thirdly, it has been investigated how the platform PAOI is discursively constructed by American news websites. Four selected texts from the POAI website, four official press releases, nine selected texts from the POAI founding fathers and eight American news articles were analysed. This research has illustrated on the level of the text that the usage of certain words, phrases, metaphors, questions, forms of addressing the audience and pictures influence the way a message is conveyed. The level of intertextuality, or the way the text refers to a broader set of texts effects discourses and power relations. Within the social dimension the media companies which the journalists are writing for influences the ways in which a technology is portrayed. Whereas the POAI platform and their founding fathers set AI in a very utopian spotlight, this differs from the view of American news writers. Based on their own research these authors promote a more negative voice towards the platform POAI and AI. This highlights also the dominance of the technological imaginary in the discourse and the way different actors such as journalists are able to influence discourses. It underpins that a good informed press is a necessity to stabilize the power relations between different discourses.