from Polis of the discussion in Mr. Hoshi's PR
hazuma How does broad listening decide in this brilliantly polarized situation? nishio I wrote a long post because I wanted Mr. Azuma to have a good understanding of Polis.
First of all, broad listening is not a "mechanism for group decision making". So this tool "Polis" does not have a "final decision making function", and I did not create this place with the intention of making a decision with Polis.
Then what is gained by Polis is "visualization of the feelings of a large number of people". For example, at this point in time, 100 people have expressed their approval or disapproval a total of 4000 times. Polis visualizes people's voting behavior with statistical processing. Polis calculates and visualizes the structure of how the pros and cons of each opinion correlate with the pros and cons of other opinions.
By enabling people to recognize what they could not recognize before, some people gain new insights. For example, "I thought everyone would agree with opinion A, but surprisingly not so much", "I thought anyone who disagrees with opinion B is an enemy and we cannot understand each other, but surprisingly they also agree with opinion C", "I thought opinion D, which is my focus, is the important topic, but surprisingly many people don't care (judgment I thought that opinion D was the important topic that I was focusing on, but there are a surprisingly large number of people who don't care about it (judgment pending)" etc.
Based on this realization, a few percent of the 100 people will try to bridge the divide or change the way they express their opinions, adding new cards that are not mere opposition to the "A should v.s. B should" conflict structure. Newly added cards are less likely to be displayed if they have the same voting tendencies as existing cards. New statements that are highly redundant with existing discussions will have a lower presentation priority. This mechanism increases the frequency of displaying cards that change the composition of the conflict. It does not guarantee resolution of conflicts, but it allows for communication that is different from the type of social networking that maintains or reinforces the composition of conflicts.
One interesting point is that Polis does not require all participants to be rational people or to have a calm discussion. 90% of the people can vote for or against the displayed opinion without thinking it through, and Polis just visualizes it as it is: "this is how people feel". I think this "visualizing how many people feel in aggregate and improving discussion based on that" is quite similar to General Will 2.0, don't you think?
Hazuma asked: “In this sharply polarized situation, how would Broad Listening make a decision?”
Nishio replied: First, Broad Listening is not “a mechanism for group decision-making.” The tool “Polis” therefore does not include any “final decision-making” function, and I did not set up this space with the intention of making decisions using Polis.
So what do we get from Polis? It’s “a visualization of how a large group of people feel.” For example, at this point 100 people have cast a cumulative total of 4,000 up/down votes. That volume is beyond any individual’s cognitive span; left as raw data, no one can form a bird’s-eye understanding. Polis applies statistical processing to people’s voting behavior and visualizes it. It computes and displays the structure of how agreement/disagreement on one statement correlates with agreement/disagreement on others.
By enabling people to perceive things they previously couldn’t, some participants gain new insights. For instance: “I thought everyone would agree with statement A, but maybe not,” or “I’d assumed people who oppose B are enemies we can’t understand, but it turns out they also agree with C,” or “I believed my pet topic D was crucial, yet quite a few people either don’t care or withhold judgment.”
If even a few percent of those 100 people act on these insights—trying to bridge divides or changing how they express their views—new “cards” (statements) get added that are not merely “A should vs. B should.” Newly added cards that share the same voting pattern as existing ones are shown less often. In other words, redundant new statements are deprioritized. This design increases the display frequency of cards that can shift the polarized framing. It doesn’t guarantee conflict resolution, but it enables a different kind of communication than the sort of social media that tends to preserve or intensify divisions.
Interestingly, Polis does not require that all participants be perfectly rational or that the discussion remain cool-headed as a precondition. It’s perfectly fine even if 90% of people cast emotionally driven votes on displayed statements without thinking too hard; Polis simply visualizes, as-is, “this is how people feel.” In this respect—“aggregating and visualizing how many people feel, then improving the discussion based on that”—isn’t it quite similar to the idea of “General Will 2.0”? What do you think?
This page is auto-translated from /nishio/東さんに対してのPolis解説 using DeepL. If you looks something interesting but the auto-translated English is not good enough to understand it, feel free to let me know at @nishio_en. I'm very happy to spread my thought to non-Japanese readers.