Last week, I attended the Archivists in the Loop symposium in Ann Arbor, Michigan. The highly interactive gathering, which was sponsored by the Bentley Historical Library and the University of Michigan School of Information, brought together practitioners, technologists, and scholars to develop a deeper understanding of how generative AI will impact the work of archival reference services. I found the two days we spent together to be incredibly generative, in ways that I haven’t experienced in conversations around AI and archives.
Structuring Input
A big part of what made this successful was who has part of the meeting. The meeting organizers (Meghan Courtney and Jesse Johnston) did a fantastic job of pulling together a group with a wide range of expertise, experience, and perspectives. Not only were students, reference archivists, developers, metadata experts, and archival scholars in the room, the gathering was designed to ensure that everyone’s perspectives were heard and valued through a series of structured activities. In many ways, this continues to build off of the Lighting the Way Forum, which introduced these kinds of facilitation techniques to a broader archives-focused audience, and which was itself grounded in Liberating Structures methodologies, which we’ve used heavily here at the RAC over the last ten-plus years to ensure we are hearing from everyone, and are building consensus and buy-in as part of the change process. If you know me, you know that I’ve been banging on about how archives need to take facilitation more seriously as a core skillset, so it’s gratifying to see these techniques being used as a foundation for challenging discussions.
Continuing Discussions
Many, if not all, of the conversations that happened over the course of the two days we spent together were not new conversations. This is perhaps to be expected for a gathering that billed itself as being at the intersection of archival reference and AI. Taken one way, that means that the gathering was part of much larger conversation about the role of AI in archives writ large. As any archivist can tell you, the reference services are very much shaped by all the other archival functions: accessioning, arrangement and description, and preservation, so any conversation about AI in archival reference is really a conversation about AI in archives. Taken another way, placing the discussions at the intersection of archival reference and AI means that they were about what search interfaces for archives should be and do, and what kinds of research methodologies they should support and privilege. This is a debate that has been going on for a long time (see all the “our search interfaces should look like Google” takes from the early 2000s) and will likely continue as new technologies emerge.
This is simultaneously reassuring and also requires us to think more deeply about our work. For better or worse, the expression of core archival principles and practices (for example the anathema of item-level description) are often grounded in the realities of chronic under resourcing and lack of technical expertise. Because AI can automate some activities that were previously very difficult or impossible to automate (as one forum participant put it, “we can now scale previously unclimbable mountains”) the conversation in many ways is shifting from what can be done to what should be done. Personally, I welcome this shift, as it asks us all to be more intentional about how our work is tied to our values, but I also recognize that means more conversations like the ones that took place during the symposium.
The challenge is where those conversations are going to take place. Over the past few years I’ve observed that we seem to be missing some vital connective tissue that allows us to work together across organizational boundaries. In the past, this kind of thing has been supported by grants, but with that funding drastically reduced or eliminated entirely, we seem to be lacking the social infrastructure for getting things done. Our national and regional professional networks, while useful in their own ways, lack the specificity of purpose and accompanying rewards and accountability structures to move initiatives forward. In the context of resource scarcity, it’s hard for anyone to make the case for the work involved in building and maintaining community.
All that professional existential dread aside, I thought the symposium was a really great step forward in producing positive and nuanced conversations around a controversial topic. I also thought the inclusion of reference practitioners as valued partners in these conversations was especially important, given that these folks are often missing from conversations about technologies which impact their work.