Exploring our engagement with AI
When I joined Connected by Data, there was an implicit understanding that we would be working specifically on data and there was a specific aim to separate this from conversations about AI. The lines were often blurred, and AI often trickled into our discussions and our work one way or another. An example being at the ‘DPDBI’ Parliamentary Event (you can find our summary of the event here) where there was a discussion on Automated Decision-Making. Whilst the focus of the discussion was deeply rooted in how a lack of collective authority in decisions regarding how our data is collected and used was resulting in systematic, collective impacts; The mode for these decision-making processes was a form of artificial intelligence. Thus it can be argued that our call for communities to have meaningful engagement with decisions about data may be helpful to extend to the designs and implementation of AI.
As of recent, the research team in particular, has been exploring how best Connected by Data can interact with the AI space. Personally, this has sparked a deeper understanding of the relationship between data and AI. Whilst there is an intimate link between the two, they cannot necessarily be discussed in the same way because of their key dissimilarities . As such, it has been really interesting discussing with @Maria and @Tim, how Connected by Data can contribute to discussions around AI, and how our knowledge and understanding of participatory data practices could be of use to conversations about AI. @Maria and I have been exploring organisations who already work in this space and I’m looking forward to contributing to such efforts.
Centering Communities in the Data Economy
As part of furthering my understanding of the relationship between the work of Connected by Data and ‘AI’, I attended the MozFest session on ‘Centering Communities in the Data Economy’. Out of all the things I’d learned from this session, what stuck with me the most was the passion and dedication of the speakers for this specific cause. They all came from different backgrounds, supporting different communities with different approaches but had all come across similar issues.
One key issue that was thread throughout the session was the issue of power imbalances, often between the researchers and the researched. Attempts to reduce the distance between the two and empower the ‘researched’ often came into conflict with popular ideals of open access to data. If power lies in those who hold the data, what can be done to make these relationships more mutually beneficial?
There was further discussion around how the ideals of open data do not benefit communities equally. There can be large groups of marginalised people uploading open source data, but those accessing and using the data often do not come from the same backgrounds. They argued that there is sometimes a level of privilege required for people to take advantage of this open-source data, and communities should be empowered to take better advantage of such data. Whether that be restricting access in line with traditional values of sovereignty, like the work done at Te Hiku Media or in the form of royalties; there are many opportunities for a shift in the decision-making power.
Developments in the Insight Infrastructure
My final reflection in this set of weeknotes is regarding a slow but gratifying evolution in the JRF project. When initially introduced to our work with the Joseph Rowntree Foundation, we had envisioned that the insight infrastructure would be an innovative digital platform whose sole aim was to provide insight and improve the timeliness and quality of data used to reduce poverty and social inequalities in the UK. Whilst the project still aims to fulfil those aims, our understanding of the infrastructure has truly evolved over time and it looks like its initial launch will be wildly different to what we had initially imagined. Off the back of our workshops, interviews and surveys to explore stakeholder engagement, there seemed to be a lean towards the infrastructure being a more community-based, development initiative in its first iteration. This is most exciting to me, not only for the improved network of organisations working in the poverty and social inequality space, but in terms of watching a project transform in careful consideration of the wishes of its potential users. In the long term, I hope this makes for a more useful and productive infrastructure, than one that could have been released without this level of engagement.