The term ‘artificial intelligence’ is used to refer to things as apocalyptic and overblown as the existential threat of the Singularity and as banal (and sometimes evilly so) as the Microsoft-Excel-powered drudgery of spreadsheets.
How we speak about AI makes a difference: while it may sound intriguing to ‘leverage cutting-edge artificial intelligence to optimize’ a sluggish administrative process, we may feel less comfortable about ‘using a correlation-based decision support system’ to do the same thing, especially if errors resulting from the ‘optimization’ of that administrative process have concrete, devastating, and irreversible effects on someone’s life.
From October 2019 to July 2020, my Mozilla fellowship project will investigate how the narratives we spin – and the myths, hype and inaccuracies we perpetuate – about AI affect how we create practical policies and how we imagine possible futures.
This current project builds on the work I did from March to July of 2019 with the digital rights NGO, Access Now, on the European Union’s ‘Ethics Guidelines for Trustworthy AI.’
Keep an eye and an ear open for upcoming announcements about the project.