Today’s Reading Club session was a thought provoking discussion inspired by an article from Amber Cushing and Giulia Osti in the Journal of Documentation - “So how do we balance all of these needs?”: how the concept of AI technology impacts digital archival expertise (https://doi.org/10.1108/JD-08-2022-0170).
The article summarized the thoughts and expectations of a focus group of archival practitioners around Artificial Intelligence (AI) and the impact on expertise within the sector. After a brief round of introductions and some background on our own individual experiences relating to use of AI in our digital preservation work, we turned to a more general discussion of the article and I’ve tried to sum up a few key points below:
-
We were interested in the parallels that were drawn in the article around AI being a disruptive technology similar to the mass digitisation of archives in the 1990’s. We discussed in particular the expectations of users in this context. Like digitisation, AI has the potential to change how users interact with archives.
-
Managing expectations was a strong theme in the article and we talked a little about whether having to manage the expectations of senior colleagues reflected our own experiences of AI.
-
We had some discussion about fear of digital disruption within the archives sector, and fear of AI more specifically. Lack of trust in the tools came up in our discussions and it was noted that the adoption of AI may take us backwards in terms of commitments around the decolonisation of archives. The importance of representative training data (and of course careful monitoring of processes and outputs) was thought to be particularly important in this context.
-
There wasn’t a strong feeling that AI is likely to take over our jobs in the very near future, rather, there was agreement with the article around the importance of the human-machine partnership. AI can help us do our job, but the final curatorial decisions should be made by us. This reminded me of a great presentation from the World Bank Group at a DPC event earlier this year about the successful use of AI to help digital archivists to appraise large quantities of AV material.
-
The article mentioned the incorporation of AI technology into Preservica’s digital preservation solution, and there was some uncertainty about what this did and whether attendees had any experience of using this yet.
-
The question was raised around what we as a community want AI to help us with. It was recognised that it would be useful for helping us to assess, sort and describe large quantities of digital content and also for helping flag up potentially sensitive records.
-
We discussed what skills digital preservation practitioners really need to be able to move forward with AI and related technologies. The article had mentioned that skills would be needed for relating to the upkeep of the tools and systems in use for AI. More broadly than that, we thought that communication skills would be a priority and in particular the ability to communicate with technical colleagues and service providers.
-
In terms of what participants felt they needed in order to make progress in this area, in the first instance it was just to have some basic understanding of what all the terms related to AI mean and what initial steps could be made to get started. It was noted that the Beginner's Guide to Computational Access released by the DPC last year was a good starting point for some of these definitions.
The article gave us a good opportunity to chat about AI and its implications for digital preservation practitioners. This topic is one that we will definitely be returning to over the next year with further DPC events and activities planned.
Our next Reading Club session will be held next month and will be on the topic of the accessibility of data, so do sign up here if you would like to join us for a chat.
Read more...