In a recent freelance facilitation gig, where I’ve been running quarterly retreats for a remote-working organisation, I’ve been emphasising the mantra that ‘Less is more’ to help focus work, and equip the team for the marathon, rather than sprint. However, as I look back over the last month for week month notes, I’m realising I might need to get better at taking some of that advice myself. That said, as the school holiday’s start (meaning reduced working hours and some blocks of holiday), it does feel like we’ve got some balls rolling ready to hopefully nudge along in the summer, and pick up in earnest come September. More on those in a moment, but first, some reflections on the latest case I’ve added to our library of participatory data governance in practice.
Meta Community Forum
At the end of June, the Digital Democracy Lab at Stanford announced the results of, what they describe as, the first ever global deliberative poll: involving over 6000 deliberators in 19 languages across 32 countries to discuss and respond to proposals on how bullying and harassment should be handled in private virtual reality communities on hosted platforms.
The project is interesting on a number of levels:
- This Community Forum (and the follow up announcement from Meta that they will hold similar activities to explore “how people would like [generative AI] models to behave for nuanced topics”) underscore the increasing interest of large technology firms in democratic and deliberative processes to help in developing, or legitimating, their approaches to platform and data governance.
- The approach taken, using the Stanford Online Deliberation platform, and a structured set of polling questions, was able to achieve an impressive scale (more than 6000 deliberators) and was paired with a control trial to generate evidence on the impacts of deliberation.
- Both the quantitative and qualitative data presented start to point to some of the subtleties of how governance might need to be designed for different countries and contexts.
However, there are also some big limitations:
- As yet, the materials used to inform deliberation do not appear to have been published, and there is little information in the reports about how these were translated / contextualised for different countries;
- Neither the write-up from Stanford, not linked content from Meta, have any information about what Meta might do as a result of the Community Forum.
- The construction of randomly sampled groups, with their interaction carefully managed through a digital platform, does not appear to have allowed any effective formation of a ‘community’ or communities amongst members of the Community Forum.
In other words, while there is a demonstration here of the potential and practicality _of deliberating about digital governance at scale, there isn’t evidence (yet) that this has given the community a _powerful voice in shaping or making decisions.
I found using our (Participedia derived) case study tool to document this example was super helpful for exploring these details - and I’ll be exploring with Maria whether we might do some further digging into this case in coming weeks.
Open AI? Closed conversations?
And while on the topic of Silicon Valley firms exploring ‘democratic’ inputs to their governance, the proposal I worked on with Iswe in response to the Open AI ‘Democratic Inputs to AI call’ didn’t make the final cut (out of 800 applications), although it seemed it was under consideration until the final stage. Unlike the Meta Community Forum model, that sought to deliver ‘consistent’ facilitation across every single participant group, our proposal was based on supporting local groups to host ‘distributed dialogues’ that would respond to common questions and to local priorities.
The selected grantees are not announced yet - but it will be interesting to see the approaches that have backing.
However, the work on our proposal hasn’t been wasted. A few hours after hearing the outcome of the grant process, we started to explore how we might build on the concept of a re-usable toolkit for meaningful local deliberations on data and AI - and will be putting together some updated plans with the Iswe team over the coming weeks. (If we were in touch with you about the bid before - look out for a note from me soon!)
Open Government Partnership Summit
Helena and I have been firming up plans for our next design lab - building on the framework from the Policy Lab in Costa Rica - to focus on the more detailed design of policy commitments on transparency, participation and redress in data and AI governance.
We’re going to host this as a fringe meeting on the side of the Open Government Partnership Summit in Tallinn on the 4th September and have just started sending out invites. If you might be interested in getting involved - do drop me a line.
September US trip
I’ve also been booking my travel to then head onwards to Boston for 6th September where I’ll spend a few days at the Berkman-Klein Centre, which will be celebrating its 25th anniversary. I’m also looking forward to 10-year reunion of my fellows cohort: which will be a fantastic opportunity to reconnect with old friends (albeit sure to be rather overwhelming given the diverse and amazing work they all do!). I’m around until at least end of the 11st Sept, so if you happen to be in New York or Boston and want to connect - let me know!
Other bits
We’ve been thinking quite a bit about our fundraising plans, and how to get involved in research projects that can help both advance our learning, and contribute directly to stronger participatory data governance practice. That has coincided with a few approaches from researchers putting together UKRI and ESRC funding bids, which has provided the opportunity to learn about some fascinating work.
One such conversation was with Dr Erica Nelson who is exploring race-based inequalities in access to digital health services, and posing key questions about how to make public-involvement in health data governance (which is relatively well developed in the UK, at least in formal terms) much more inclusive and responsive to marginalised groups.
Last week we also had a workshop as part of our Measuring Data Values project to explore strategies for using indicator data in change advocacy. More to write-up on that soon.
And it’s been a month with lots and lots of reading (and a lot still on the list to read). A few highlights:
- David Eaves on the promise and pitfalls of digital government - providing some critical background for our G20 engagement around the concept of Digital Public Infrastructure;
- A brief history of collective intelligence, democracy, and governance from the Routledge Handbook of Collective Intelligence for Democracy and Governance - which we’ve also selected to discuss for our Connected by Data fellows reading group in August.
So much more I feel I should write-up, but out of time for now. Perhaps next time I really will take my own less-is-more advice!