The final speaker in this session at the ZeMKI 20th anniversary conference in Bremen is Hossein Derakhshan, whose focus is on AI and algorithm studies. He begins by noting an ontological crisis in media studies: the field has looked at the production of, audiences for, and texts of media, but the rise of algorithmic platforms in particular has meant that the media texts, in particular, have now been destabilised – users of media no longer necessarily encounter the same texts in the same forms, formats, and combinations.
Instagram posts and stories, TikTok feeds, Spotify playlists, Meta ads, AI chats are all highly changeable, and algorithms have emerged as the de facto media texts to study instead; with the era of mass personalisation, where production becomes consumption, algorithms are hyper-modulatory, invisible, and inextricable from the media texts themselves, and the platforms through which they occur. How might we study digital media under these circumstances, then?
Such distinct ontology requires a distinct epistemology; we must study what platforms do instead of what they are. One way of doing so is to examine the social figurations they enable: to study them as processes rather than objects, with interdependencies, and as continuously changing.
This means a shift from a focus on datafication to personalisation; from surveillance that connects life events to digits to prediction that connects data to other data; from categorisation of data through data to allocation of data to life events. This can build on ethnomethodology, and Hossein’s current project explores this through a breaching experiment that swaps participants’ Spotify accounts to enable them to become more aware of and reflect on the personalisation of their Spotify experiences.
