The Commons

Four years ago, I went to the first Build Peace conference at MIT’s Media Lab. I knew next to nothing about the team of young people from Build UP^ who organized the event, but I’ll go to almost anything that involves peace and technology.

The conference and the group of now 30-somethings that put it on changed my life. I have been to all four of their conferences (Cyprus, Zurich, and Bogota in addition to MIT) and plan to be at the next one in Belfast in October. Not only have I learned a lot about peace, technology, and the arts, but I’ve made a few friends for life.

So, when they asked me to help assess their newest project, The Commons, I immediately booked my flight and hotel, although we did not end up at the amazing Media Lab this time. As they put it, The Commons has been a pilot program for addressing polarization in the United States on line, which was led by Build Up’s managing director, Helena Puig Larrauri.

Like many of us, my friend at Build Up^ are worried about the ways life in the United States and Europe has become polarized into what Amy Chua and others call political tribes. As good techies, the Build Up^ team decided to see if they could use ICT tools to reduce the polarization.

This was, of course, ironic. The first time we met, the discussion focused on how we could use the new technologies to promote peace. Barely four years later, we had embarked on a project that might help us reduce the political harm caused by those very same technologies.

So, Helena and her team got a small grant from a Dutch funder to conduct a pilot study to see if they could use IT tools to make a difference. Although this was only a pilot and the group made a few false starts, they did demonstrate that we can use social media tools to address polarization in the United States—and, by implication, elsewhere.

They started by identifying a pool of potential participants all of whom followed and posted material drawn from the 100 American senators using methodologies that reflect the way the two main social media platforms are organized. From that group, they narrowed their focus to smaller subsets of Twitter users who used a set of hashtags that might reflect extreme views and Facebook users whose comments reflected radical points of view. At that point, a self-identified bot (figure out what that means while you’re at it) sent tweets to the targeted Twitter users while ads were placed in a number of particularly divided communities to reach Facebook users.

If users responded to the bots, they were transferred to a live human facilitator who asked if they wanted to participate in a discussion with people who hold different points of view. The Build Up^ team did not want to change anyone’s point of view. Instead, they wanted to see if tensions could be defused if people talked together on line.

The pilot had more success on Twitter, which is not surprising since all tweets are public. Of the 880 people contacted using the bot, 11.7% responded and half of them participated in further discussions. Facebook was harder to use and produced different results. In the end, the team posted general ads as well as on its own page. Of the nearly 25,000 people. Who received the ad, 874 responded positively, and 332 of them took part in further Facebook discussions with one of the facilitators. The discussions on both platforms did not progress very far in part because the Build Up^ team primarily wanted to demonstrate that one could identify and then attract people to this kind of discussion.

At our workshop on the future of the project, the 15 or so people in the room agreed that the team should continue its work and address a number of shortcomings that emerged from the pilot:

  • The team was made up of individuals all of whom were themselves left of center. We plan to include more conservatives in the design and research process from now on.
  • Starting with people who followed U.S. senators may not have been the best approach. We will be trying to find other ways of identifying sets of people whose views are likely to be more polarized than many who focus on mainstream politicians.
  • Explore other ways of conducting dialogues. Neither Twitter nor Facebook were designed to extended, in-depth discussions. While there are other platforms that could be used such as those developed by the late but greatly missed WebLab that could be used here, it is not clear that one can ask people to move from Facebook or Twitter to some third platform.
  • Most importantly, it is not clear how to fund a project that could read tens of thousands of people per month and do so in ways that targets specific issues or places or demographic communities.

Also published on Medium.