PROGRAMME BACKGROUND
& PROCESS
Background
Advances in AI don’t automatically benefit those who need them most. In the case of a scientific breakthrough like AlphaFold, for example, it took intentional partnership with the Drugs for Neglected Diseases Initiative (DNDi) to understand how and where AlphaFold’s predictions could help those suffering from neglected diseases, which disproportionately affect people in the Global South.
Our experience of working together to share the benefits of breakthroughs like AlphaFold has, in part, contributed to our practice of meaningfully engaging with civil society, experts, practitioners and advocates so that we can learn from one another. This kind of collaboration is essential to designing, building and deploying safe and beneficial AI in different contexts and geographies.
It was with this in mind that, in 2022, we worked with the Aspen Institute to convene two multi-disciplinary discussions about how to think about and enable equitable AI. A summary of those conversations was published as the report ‘A Blueprint for Equitable AI’ in early 2023. Our goal in exploring this concept was to see how AI might facilitate and drive inclusion, as opposed to reinforcing bias and historical patterns of injustice. The roundtables we worked on this year all build on questions and themes that emerged from collective consideration of how the benefits of AI might be distributed equitably.
The Process
As AI will ultimately affect all sectors and societies, there is a vast range of topics to explore. With a focus on learning through policy discovery, we were guided by the organisations we worked with to prioritise sectors where AI holds significant promise and where collaboration and thoughtful policy design are needed to ensure benefits are distributed equitably. Additionally, we sought to understand pressing governance questions that must be answered in order for AI to have a positive impact in the world.
For each roundtable, we worked with an organisation with deep expertise in the selected subject and connections to a diverse range of experts and practitioners, including those working on the ground. Designing the roundtables was a collaborative process, though partner organisations took the lead given their embedded expertise.
While each conversation took a slightly different format, the programme overall sought to explore the following questions:
How should AI be built?
Where should AI be used?
How should AI be governed?
Who should be accountable for the use of AI?