Getting past our post-truth paralysis: Communicating across the divide became nearly impossible once everyone decided they were entitle to their own facts. We at least need to agree upon the evidence that defines a problem before we can effectively discuss collaborating on a solution. Artificial intelligence has an amazing ability to “average out” conflicting information into an objective “truth.” In this project, we let reds and blues take turns adding books and documents to a chatbot’s reference library, and then ask it to take both sides into consideration when answering tough questions on policy. Because those answers are based upon a balanced set of reference documents, the answers are balanced, giving red and blue a solid place to start their dialogs. We also conduct the first few rounds of conversation with that chatbot’s as facilitator to keep things civil and fact based. Net result: constructive conversation that incorporates everyone’s point of view, and builds trust between the parties.