The debate over AI’s future is happening today. Concerns over generative AI (or GAI) should be the subject of more constructive and nuanced public discourse and consultation in Canada. GAI has provocatively been referred to as a co-opting machine that appropriates and automates labour practices. Even more troubling than these questionable avenues of development is the fundamental critique of GAI systems, highlighted by the major international scandal and eventual closure of Google’s ‘Ethical’ branch, which involved the poignant criticism that general-purpose and large-language models can only repeat back what they’ve already heard, rather than generating new information or truly “thinking.” At worst, these “stochastic parrots” can only repeat back stereotypes, racist tropes, and the other structural inequities encoded within the Internet.
Yet, the continued belief in GPAI as either a social good or as an existential event for civilization is a key school of AI governmentality, funding GPAI as a subfield and problematically focusing AI regulation on long-term issues. Outputs from this project will improve public discussion of GAI and improve futures literacy.
This research was supported by the Social Sciences and Humanities Research Council of Canada.