The Art of the “Prompt”
Beth Fisher-Yoshida, co-director of AC4 at Columbia University, shares advanced narrative tools to improve the digitization of listening processes promoted by ALC. The criteria we establish for analyzing information (prompts) will determine the quality and impact of artificial intelligence in addressing complex social challenges.
Co-creating meaning
The CMM (Coordinated Management of Meaning) approach holds that communication is never neutral and that meaning is co-created through interaction and context: “the moment you walk into a room, you already change the situation.” From this perspective, the three dimensions of the model were analyzed: Coherence (how we organize meaning), Coordination (how we act together despite differences), and Mystery (the acceptance that complex systems cannot be fully controlled).
Fisher-Yoshida recalled a key maxim for social transformation: “Data doesn't change our minds.” It is narratives and relationships that drive profound change. The ALC team was able to delve deeper into narrative complexity using the LUUUUTT Model. This technique allows us to distinguish between public narratives that are shared (told and lived) and those that are not shared (unheard, untold), those that are unknown, and those that are not permitted (untellable).
Practical application in the K-tool
The workshop allowed the participants to connect these tools (along with other models such as Serpentine and Daisy) to ALC’s current work, identifying applications for automating narrative analysis within the K-tool.
This will result in concrete improvements to the listening processes through the design of more thoughtful questions and prompts that will enable us to train language analysis models and simulate different response scenarios.