Orienting AI Toward Peace
21 Nov 2023
On November 10th, 2023, CHAI senior scientist Jonathan Stray presented a talk called “Orienting AI Toward Peace” at the Beyond Moderation: How We Can Use Technology to De-Escalate Political Conflict conference at Stanford. His talk outlined a three part strategy to ensure that AI systems do not inadvertently escalate political conflict as a result of misaligned optimization and are resistant to bad conflict actors. This strategy is outlined as such:
１. We have to define what kind of conflicts are desirable and undesirable, because non-violent conflicts are an important part of how societies change for the better.
２. We have to develop indicators and metrics which can distinguish between good and bad conflict at the level of individual discussions and items of content.
３. We need to incorporate such feedback into the objectives of AI systems. There is ongoing research on all of these topics happening at CHAI.