What the pol.is appears to do well is provide a topology of opposing viewpoints- identifying the main "teams", finding the contours of agreement and disagreemen

What the pol.is appears to do well is provide a topology of opposing viewpoints- identifying the main "teams", finding the contours of agreement and disagreement, "a technology for survey research." The HN URL describes something more ambitious: a technology for "Large Scale Discussions." This requires iterative steps where representatives from opposing viewpoints ask clarifying / challenging questions of other viewpoints, and collapsing "consensus-achieved" points as summary discussions in a knowledge base. I've long had `pol.is` as a tool to be used in a 21st century platform for discourse, so the HN title has me excited that you're taking steps in this direction already. Where should I look to understand "what happens next" after that initial sorting of viewpoints has occurred?