Challenge on AI and participatory engagement


Diverse AI faces the challenge of wanting to re-work existing AI tech to better serve our diverse communities of users. We are looking to explore the potential for implementable processes and methods that offer genuine participatory engagement; that demonstrate a possible way to re-imagine, re-create, and re-tool existing deployed AI technology to become democratic, user-centred, participatory and inclusive.

As Generative AI technology outputs are designed to mimic human outputs, the quality of the output is by definition, subjective and what counts as correct outputs varies between different communities. Instead of an expert-led effort to determine how to assess these outputs for factuality, efficacy or harm reduction, our challenge is to explore ways of achieving these aims through participatory and community driven exercises. Methods that go beyond current metrics (e.g. – thumbs up or thumbs down buttons) to measure ‘success’.

Our challenge is to imagine ways in which AI can be utilised as a collaborative tool to amplify values and lived experiences of marginalised voices, in the context of technology development, as well as a tool to be used to build and create community and digital cohesion and empowerment movements.

Our present focus is on Generative AI technologies and finding ways to guide positive uses and impacts of these technologies in the context of community building and digital advocacy and empowerment movement building.

In this context, we invite research proposals that address how to assess whether a Generative AI technology aligns to specific community values. (challenge-code DAI-C)

Additional Context

As part of the project, we are also going to partner with community organisations based in the global South. Our aim is to ensure that our work be guided and informed by participatory inputs drawn from communities not just restricted to the Global North –  reaching people who will inevitably be impacted by the deployment of AI tools, but have not historically been involved in designing or evaluating them.

Our team members are either from countries and communities in the Global South or have strong links to and/or experience in community organising to help facilitate this aim. For example, we are partnering with an organisation that works in training community based teachers to further computer literacy in rural West Bengal and Bangladesh – to understand how AI tools can best be utilised to help in educational efforts that meets the requirements and access needs of this particular community.

We invite research proposals that can explore or contribute to various aspects of the challenge, such as:

  • how to determine what value alignment of technology with humans in a distributed and participatory context means.
  • research activity in developing new metrics and / or procedures for evaluating community specific uses of LLMs within our project.
  • how this exercise of data collection and values discovery in the context of the use of AI can inform data and AI literacy efforts, as well as digital and community building.
  • how the results of this project can inform a more democratic, user-centred approach to AI development, as well as AI regulation and policy.

Working Arrangements

Diverse AI is a registered community interest company and a non-for-profit membership organisation that aims to drive positive change through supporting and growing diversity in the field of AI. We are run by a team of volunteers including an AI ethicist, a social scientist working on social aspects of big data, a public policy and big data researcher specialising in inclusive AI and a marketing and communications lead.

How it will work

  • If the application is successful, Diverse AI will work with the fellow to refine the project plan and agree on shared goals and outcomes.
  • Diverse AI will support the fellow in terms of onboarding, providing a research contact and collaborator, and engaging in regular meetings. These meetings will comprise weekly virtual 1:1’s with the key research contact as well as virtual biweekly project update meetings. We believe in celebrating our successes however small and communicate regular updates to our wider community through bi-monthly project updates.
  • As Diverse AI is a volunteer-led organisation, the project will need to be conducted remotely, with opportunities to meet in-person in London with members of the time from time to time.
  • We expect the fellow to factor into their budget proposal travel, accommodation and subsistence costs and any specific research costs they envision.

(Image: Anton Grabolle / Better Images of AI / Human-AI collaboration / CC-BY 4.0)

Scroll to Top