Alex: Hello Iwan! Welcome to this 5-minute interview. I’m Alex, the Communications Coordinator at Kineviz, and I’m excited to delve deeper with you on the art of creating space for polarized conversations. Speaking of interesting conversations, can you tell me a little bit about yourself and what you’re doing at Polarizer?

Iwan: Thanks Alex, super happy to talk with you! As you said, my name is Iwan Itterman. I’m the founder of Polarizer which is a community-based app visualizing the relationships between opinions expressed on any given topic. Once an opinion is expressed, participants can connect with others on that topic and engage in one-on-one discussions. People on both sides of the discussion can rate the quality of conversation and have these ratings determine the position of their opinions on a large social graph, meaning that if you have a lot of bad discussions you might be positioned on the fringe and if you have a lot of good discussions, then you’re closer to the center.

Alex: So, what is considered a good discussion?

Iwan: We strongly believe only those involved can answer this question. That is why user self-ratings are a core feature of Polarizer. There are two dimensions that affect a conversation’s rating: how closely both opinions are in agreement, and how respectful the discussion is. A good discussion in Polarizer is, therefore, one that both participants rate as disagreeing and respectful.

Alex: That is fascinating. How did you get to this idea and what’s the backstory?

Iwan: It started back in 2016, which I think a lot of us can remember as the time when problems with social echo chambers really hit the fan. The two things I observed going awfully wrong were on..

How affective polarization skyrocketed when social media platforms started personalizing news feeds based on interests.
How radicalization rose when they soon after used AI to maximize engagement.

I wanted to address these issues with Polarizer by…

Sorting, not filtering information. Polarizer wants to enable people to have transparency about other information out there and how it relates to them.
Using meta-data instead of personal data. The platform does not collect any identifying information. Opinions are standalone with each one provided a hash id so as to preserve anonymity.

Alex: How are you bridging opposing views online?

Iwan: Like how are we solving the social dilemma? It is not an easy feat but we’re doing our best by allowing people to explicitly see where their opinions are across a random pool of participants. Opinions are placed on a real-time social graph based on self ratings from one-on-one discussions. People are able to see the evolution of echo chambers and their distance to other people’s opinions. While most people might engage with like-minded people, they do have the ability to engage with outside opinions safely. So if they manage to hold more respectful discussions based on how the conversation is rated on either side, they will also receive more attention.

Alex: Can you elaborate on what kind of technology you are using to structure these one-on-one discussions?

Iwan: In a graph, the opinions are the nodes and the one-on-one discussions are the edge values. We are using GraphXR on top of a Neo4j database to create views of opinions across topics. This is currently in progress. The Polarizer graph, therefore, is working on how discussions — the edge values — can shift the view based on ratings provided by participants.

Alex: So, what are the next steps for you and Polarizer? What are you looking forward to?

Iwan: Next steps involve launching a second prototype of Polarizer which is going to be mobile and graph-based. I’m very happy to be collaborating with Kineviz on this front! Once we have our second prototype, we are looking to do a crowdfunding token sale.

Alex: How can people get involved and learn more about this work?

Iwan: Of course! Feel free to engage in the conversation with the links below or check out our Polarizer explained in 90 seconds video. 👇

Alex: Great conversation today! Thank you for taking the time to share with us what Polarizer is all about. We’re looking forward to continue this conversation in upcoming gallery posts soon. Stay tuned!

Originally published at https://www.kineviz.com on July 27, 2021.

Social Network AnalysisSocial DilemmaData VisualizationResearch And DevelopmentPolarization

“Meet people you don’t like.”