AI Beasts
The Bestiaries Project is a collaboration between the WEH, TORCH and Rachel Douglas Jones at the ETHOS Lab from IT University of Copenhagen.
The project's goal is to bring publics' attention to aspects of Artificial Intelligence (AI) that require oversight and regulation. We achieved this by creating an AI bestiary, a collection of AI beasts or monsters representing areas of AI technology that are in need of governance.
What is a Bestiary?
A Bestiary is a collection of beasts as documented in medieval manuscripts. Bestiaries were made popular in the middle ages. They are illustrated volumes that described various creatures. Each beast was usually accompanied by a moral lesson. In this activity we aim to bring to life the many beast-like qualities of AI that need controlling and monitoring.
Why are we doing this?
In the middle ages, bestiary beasts were made to express a moral lesson or theme, and this underlying message was key to their creation. It influenced what the beasts looked like and how they behaved. Therefore with our “digital bestiary” we want to raise awareness of the “monstrous” aspects of AI that require human oversight and regulation.
Teams of Oxford researchers who work on AI from various perspectives, sociologists, medievalists and ‘monster study’ experts created the ‘hero’ monsters of the digital bestiary during a facilitated workshop.
This collection of ‘hero’ AI beasts were used as the basis for workshops in Oxford primary schools, where students created their own beasts and as part of the 2022 Oxford IF Science festival interactive explorer zone.
The full bestiary was then displayed in an exhibition at the Oxford County Council Library.