This second workshop in our series continues our goal of identifying a line or lines of research that could lead to a longer funded project. As we noted then, social and cultural research on machine learning, large language models, and related technologies has inspired enormous interest and remarkable results. At the same time, it is poorly funded by comparison with AI research itself, and public frameworks lag way behind technology investment and its corporate and governmental advocacy discourse. At the ISRF, we would like to do our part to develop the social and cultural frameworks that allow AI to reflect human interests and enhance human capabilities rather than determining and replacing them.

At our March workshop, we identified 18 themes that in the aftermath of the workshop we reduced to three. These are the current candidates for continuation:

• Humanities Requirements for AI

• New Cultural and Learning Infrastructures

• Prototyping an Embodied Human-Machine Learning Approach

After an introduction of new delegates for this workshop, we will discuss each of these via some supporting readings. One question is whether we should keep these, pursue all three at different scales, shrink this list further, or also pursue a new lead. We have some sessions set aside for collaborative writing of some or all of the final proposal.

We may also want to discuss the relation of this project to ISRF’s other special topic projects: two examples are Redesigning Finance for Climate Justice and the Future of University Research. We remain very interested in ensuring that the cluster of technologies now called AI helps address our societies’ various knowledge crises rather than making them worse.

Participants to include:

Chaired by Baindu Kallon and Chris Newfield.