To mark the launch of a new website for JUST AI (Joining Up Society and Technology in AI), LSE’s Alison Powell gives an update on the project that has built an open research network to identify issues and intervene in data and AI ethics.
Data-based and automated systems are everywhere, raising increasingly urgent ethical questions. Research and innovation processes struggle to take these questions into account – and sometimes even to define them. Conversations and collaborations across the humanities, social and applied sciences and design fields are needed. Otherwise, the big ethical questions – what is good? How should we act? What do we need to allow for flourishing? – can become narrowed into tick-boxes of compliance, leaving room for more “AI unfairness” like the downgrading of student exam results in the UK in 2020, or the punitive cycles of racialized policing associated with tech like PredPol in the US and policies like Stop and Search.
In response to this need, the Arts and Humanities Research Council and the Ada Lovelace Institute have jointly supported JUST AI, a humanities-led research network that identifies issues and intervenes in data and AI ethics. JUST AI has created new mechanisms and practices for defining, understanding and collaborating on ethical questions in data and AI. Our focus has been transforming networking as a critical technology practice – for JUST AI, ‘network’ is both a noun and a verb. Our new website provides detailed documentation of the innovative methodology we have developed for networking and transforming data and AI ethics practice.
Mapping and Transforming
After mapping the field of data and AI ethics, JUST AI sought to transform the people, practices and power that contribute to how ethics are defined and practiced. To do this, we drew on ideas and practices from across the humanities, including creative writing, design, philosophy, and feminist and black studies. We created a ‘reflection prototype’ to help practitioners narrate their approach to AI ethics and transform how they work together. We’ve led workshops with researchers and activists using these tools.
The aim of JUST AI is to transform the ways that people encounter, understand, and work together on issues related to data and AI ethics. We have created mechanisms that foreground voices, perspectives and topics that are emerging into importance.
Foregrounding Emerging Voices
Three emerging areas of ethical interest have generated ongoing work.
- Foregrounding racial justice in data and AI ethics, where JUST AI supported the creation of racial justice fellowships at the Ada Lovelace Institute, including the creation of anti-racist recruitment and assessment processes.
- Our working group on Rights, Access and Refusal brought critical disability studies to AI ethics research, asking questions about how technology and bodies work together to generate new kinds of potential. This has led to an ongoing research program on Crip AI, represented in award-winning research at British Science week and as part of the AI Anarchies Autumn School. The working group has commissioned research as well as artwork.
- JUST AI has worked to define and investigate Deep Sustainability in relation to AI as an extractive practice, where the materials of computing are embedded into global dynamics of labour, resource extraction and exploitation, in addition to the ways that AI systems reproduce inequality through their computational function. The Deep Sustainability working group includes artists researchers Samir Bhowmik, Monai de Paula Antunes, Matterlurgy and Disnovation as well as tech industry members, researchers and experts in environmental activism.
Creative practice and artist research: Convening Futures
JUST AI has generated important dialogues between AI specialists and creative practitioners, as well as foregrounding different kinds of expertise in emerging AI and data issues. In 2020 we commissioned science fiction writers to create short stories based on the work of emerging AI researchers, presenting the process through a series of public salons.
In 2021 we curated a week-long series of events outlining modes of Prototyping Futures – including necessary interventions at strategic, discursive and organizational levels.
Taking humanities expertise seriously, we created spaces of encounter between artistic researchers with substantial expertise on extractivism, and a two-day convening called Beyond Critique To Repair.
JUST AI has introduced radical and urgent new approaches to some of the most significant ethical issues of our time. Many collaborators, partners and participants have made JUST AI what it is – to find out more, please visit just-ai.net.
This article reflects the views of the author and not those of the Media@LSE blog nor of the London School of Economics and Political Science.
