I am delighted to share the details of the CNI Fall 2023 Membership Meeting plenary talks and to announce that the preliminary project briefing list is now live. In the coming weeks, we will release the final schedule along with full session details. Closer to the meeting, we’ll also share the traditional meeting “roadmap” guide. 

I will deliver my customary December opening plenary with a survey of recent key developments and trends in the networked information and research landscapes alongside an update on CNI's programs, strategies, and plans for program year 2023-2024 and beyond. It’s been an extraordinarily eventful year, and there’s lots to talk about. As always, there will be time for questions and discussion.


Recent issues surrounding the use of the scholarly literature as training data for machine learning-based systems, particularly large language models (LLMs) and various generative artificial intelligence (AI) applications driven by these systems, are raising new questions about what the community seeks to enable and accomplish through commitments to open access and open scholarship. These questions are shaped in part by assumptions about the extent to which these new machine learning-driven systems will be centralized and controlled by a few large technology firms, as opposed to being diffused and distributed much more widely through open source developments.

The emergence and adoption of ideas surrounding open access and open scholarship have largely been driven by visions of desirable futures and values-based choices for institutional investments; in recent years, this has enjoyed a relatively broad consensus despite some ongoing disagreements about specific tactics to advance these goals. Suddenly, as we see examples of where generative AI may take us and the highly extractive nature of these generative AI systems, the consensus on ideal futures seems less clear. What will most effectively advance scholarship and the creation of new knowledge?

Historically, there has been clear consensus extending from the freedom to read scholarly literature to the freedom to compute on this corpus through text mining and natural language processing technologies. This had been recognized in funder open access policies and mechanisms like Creative Commons licenses. Newer machine learning-based technologies create persistent and reusable computational artifacts, and there seems to be less comfort with a powerful, broad-based set of mandates to enable these developments. Yet the implications of blocking or impeding these developments also remain unclear. And whatever is chosen must scale.

The panel, which I will moderate, will consider the current state of developments, particularly in light of the seemingly insatiable demand for training data by LLM-driven systems, and will focus on both what the desirable outcomes should be in light of the values and goals shaped by open access and open scholarship and the mechanisms that might be adopted to advance these outcomes.

Panelists include Rachael Samberg, Scholarly Communication Officer & Program Director at the University of California, Berkeley and Heather Sardis, Associate Director for Technology and Strategic Planning at the Massachusetts Institute of Technology. We will announce the complete list of panelists soon.

As a reminder, the meeting registration deadline is Friday, November 3rd; registration instructions were sent to member representatives in September. 

I look forward to seeing you in D.C.!

-Clifford Lynch, CNI Executive Director