I propose to think with the audience about this question: How can software be written to pay attention to listening, a crucial component of engagement with testimonies and memories?
As a scholar-practitioner, my work is an example of thinking with theoretical/academic spheres as well as with technical/practical spheres across narrative intelligence, software studies and digital humanities. I intend to address this question through my project titled We Are History.
In pursuit of generating communal dialogue in the context of inability to have conversations about our contested history in Lebanon, I set out to build an Artificial Agent that would sift through an oral history video archive of testimonies of daily life with the task of figuring out common threads, sometimes confirming and sometimes contesting each other, and automatically editing many different versions of possible histories. This automatic montage machine addresses two problems in the Lebanese context: first, it circumvents the tiring accusation of being biased since a machine is now the moderator (presenting a multiplicity of stories might be the closest one can get to strategic objectivity) and second, it opens up the possibility of conversation by weaving various and often opposing perspectives in order to start imagining what our histories could look like. The project, which would reside online as well as in booths in public spaces across Lebanon, invites people to listen to an automated montage of oral histories and to then share their own stories and memories. Each newly contributed story is instantly added to the archive, analyzed using new developments in computational corpus-based linguistics, automatic story generation, and social computing and tagged with its transcript which enables the interface to incorporate newly added video interviews into the pool concerning the event discussed, thereby changing the version of history previously compiled.
