Task No. 1
After understanding the issue and angle of the Deletion Bureau, our participants started in on their work. Their first task was to reflect and negotiate about the heritage-items the Deletion Bureau should keep/discard. Accordingly, they also wrestled with how to define and categorise European heritage-items. As part of our setting, they generated variations on abstract, pseudo-code algorithms to search through their collections.
This task began with a mad search through the room where we laid out the items the participants contributed in the priming. They had to find one thing to keep in 1 minute. This task felt impossible to the participants.
While they only had to look at 20 items in 1 minute, it led to the question: what if we had to consider all the items in the 2038 collections in 1 day?
We shared that we had designed tools for them: a special algorithm and accompanying artificial intelligence that could come to their rescue - if they shaped those tools to their needs. We described the positive and negative possibilities of using A.I. in the artistic context, and encouraged them to create guiding structure for our algorithm. At the beginning of the exercise, the A.I. randomly sorted, kept and discarded heritage items without rationale.
Thus, the main problem for this task was to create a rubric, a matrix for the algorithm to follow as it sorted the heritage items. How would it know what these items represented conceptually? The experts sorted and arranged keywords associated with the heritage items they had chosen, finding a common agreement of which concepts (and their associated heritage items) would be most important to keep.
We then put their choices into action, inputting the rubrics into an interactive visualisation that kept or discarded the heritage items the participants had previously contributed.
This direct and active method of provoking our participants to choose what to keep and the values that supported those decisions was the first step to enter into their role in the Deletion Bureau. Because the materials they worked with (heritage items they themselves had contributed in the priming exercise) were familiar and close, they felt those items' deletion or preservation even more keenly, stepping into the emotional dimension of Future Erasure. The narrowness of the algorithm's rubric and "intelligence" prompted our participants into a reflective discussion of how we can better collaborate with tools such as artificial intelligence.