ALGORITHMIC AFTERLIVES: THE ETHICS OF REVIVING THE DEAD

Abstract

From oral storytelling traditions to Victorian seances communing with the dead has long been a part of human existence. But as technology advances, our means of connecting with the past proliferate; as Kasket suggests “technology is [now] where the dead live” (2019, 7) and the dead are increasingly online. Websites and apps like PeopleAI and MyHeritage’s Deep Nostalgia utilise deep learning algorithms to evoke, re-frame, re-work and distort the past, and similar tools are now being introduced in cultural and heritage contexts. Indeed, as the capabilities of AI-enabled voice ‘clones’ and ‘deepfake’ technologies improve, working with algorithmic afterlives is fast becoming a mundane proposition. These practices however expose deep ethical questions about consent, legacy, ownership and custodianship which are increasingly important in the context of concerns about disinformation and declining levels of trust in public institutions and the media. We have been working with heritage professionals to better understand and respond to these concerns, in particular when it comes to the algorithmic ‘revival’ of historical figures. In this paper we introduce and reflect upon our recent work – in collaboration with 19 UK cultural professionals and alongside creative studio yello brick – to co-design an innovative toolkit for museum/historic sites navigating the creation of ‘AI afterlives’. We discuss the key takeaways from our research, highlighting the ruptures between past and present that arise in the context of AI and automation and the potential paths organisations and professionals can follow to address concerns around algorithmic revivals.