H&D Meetup 1: Algorithmic Consensus


This first H&D meetup of 2021 focussed on tools for decision making. More specifically we looked at consensus as a concrete example, and explore the potentials and limitations of consensus practices for community organization, but also the ups and downs of algorithmic decision making platforms and technical infrastructures that are built upon the consensus model.

Top CTA 02.jpg

We asked participants of our meetups to read the H&D Code of Conduct before attending our events.

Location: Framer Framed, Jitsi and H&D livestream (received upon rsvp)

At the time of the meetup Framer Framed finalized their exibition 'A Funeral For Street Culture'. A Funeral for Street Culture is a special project that digs deep into the world and sorrows of contemporary street culture, its influencers, shape-shifters and failures. Together with artists, poets, designers, thinkers and hustlers, Metro54 and Rita Ouédrago look into these questions and explore the ways street culture both interweaves with and departs from design, performativity, queerness, fashion, activism and cultural appreciation. A Funeral for Street Culture is a critical celebration of street culture and will take the shape of meetings, installations, performances, conversations, mournings and workshops. The group show will be open to the public when the corona measures are lifted. Till that time there will be an online public program.

Part 1

1st Part: livestreamed from Framer Framed open to larger audience via the H&D livestream

  • 14.00 Introduction about the H&D meetup series on tools for self-organization by Margarita and Anja
  • 14.15 - 15.15 Histories and practices of consensus by Angela Jerardi (+ Q&A)

Angela Jerardi contributed to the meetup by giving a contextual talk about the roots of consensus, as well as her experience of consensus practices in the context of large activist groups, and the Quaker movement.

Talk recording

You can watch back the introduction by Anja and Margarita as well as the talk by Angela.

Watch -->.

Q+A period

Sepp (he/him) :

The value of consensus is clear when it concerns a defined and localized common pool of resources. But how does it work in the network society? How do we do consensus on the internet? Are there promising experiments on this level?

lila (she/her):

What is the relationship between community building and consensus? Can consensus be a regenerative practice rather than tool a for decision-making? I am thinking of how to make it more of a practice of care for each other rather than something that feels bureaucratic?

Sepp (he/him) :

Are all common goals best reached through consensus, or are there certain types of decisions that are best reached in a different way (in your experience)?

Or the other way around: are there certain common goals that particularly require consensus?

emirhakin (he/him):

Who decides what is urgent or what is trivial?

emirhakin (he/him):

Can two or more opposed groups use this consensus process?

Anja: regarding the notion of 'working on oneself' ... the idea that a group doing social justice work is not only concerned with their main goal but also simultaleously reflect on their own functioning. consensus seems to put an focus inwards in that sense. self-reflection of a collective. how to find the balance here? like not to turn into an introverted non-welcoming group... i wonder if consensus can be perceived also as alienating?

Anja: You touched on this a bit in relation to time and in the slide about when not to use consensus. Do you have experience thinking about how to work through cultural differences in terms of consensus? what if people are approaching consensus from different base points of understanding?

Interesting practical consensus guide

Part 2

2nd part: Online on Jitsi for a smaller group of 10-15 participants

  • 15.30 WINWIN workshop by James and Nienke (60 min)
  • 16.30 Reflection with Angela, James, Nienke, and the participants of the workshop

Following Angela's talk, the conversation took an interactive form with a smaller group of people, who represented different self-organized grassroots initiatives. This second part of the meetup was be led by James Bryan Graves and Nienke Huitenga who built a digital consensus platform WINWIN. James and Nienke tried the platform in different contexts to see if consensus models could help to create a less polarized (online) debate culture. By testing this platform in action participants were able engage in a conversation and exchange about consensus – while using a platform that builds upon an consensus algorithm.

Notes from the meetup

For the second part of the meetup we split into two groups, one that was watching the consensus algorithm in process, and the other group was taking part in it. After the discussion concluded the groups switched.

These were the statements we tried to reach consensus on:

  1. Consensus decision making is a more inclusive way of reaching agreement between all members of a group, than for instance majority models. (group one statement)
  2. Consensus decision making processes can help to create a less polarized online discourse (group two statement)
  3. Consensus decision making is an inefficient way to come to an agreement in a group.

There were three moments of consensus on the platform. Everyone agreed with:

  1. "One might say all online discourse has established it is polarization. Consensus might carry potential to create more nuance"
  2. "Homogeneous groups might have an easier time agreeing on issues"
  3. "A homogeneous group doesn't necessarily reach consensus, and a heterogeneous group doesn't necessarily have to be at odds."

Someone asked, do you feel more together after aligning on these points of consensus?

Cristina noticed that having her own statement come up and back to her felt like a bug at first, but in the end felt like it was a positive thing since it made her question her own statement Frederique (in the second session with Nienke) was asking if, in this context, 'consensus' means 100% agreement Nienke responded that in this context it is 100% agreement, but their thesis was about whether you could achieve more nuance by having smaller moments of consensus instead of just one big decision Margarita thought it was interesting that it is not known who agrees and disagrees. James said they had thought about that. This is quite a different approach from Angela's understanding of consensus Sepp commented the time you get to write the answers on the platform is quite short, so it is difficult to write something nuanced. He wondered if that was the intention? Nienke's intention was to create nuance and rich responses, but it is difficult when the responses are very long, almost like essays Nienke furthermore noted that one of the vulnerabilities of this digital experience is that you might lose track of the previous statements and where the supplementary statements came from

Nienke introduced the idea of the consensus platform as a 'double checking method'. After a meeting it could be interesting to check if actually everyone was heard sufficiently. She also mentions that it is important to consider who 'designs' the statement that are put up for discussion.

Generally participants thought it was not enough time to make statements and write them Mario preferred short statements Juliette thought it was hard to keep track of the tree of statements, to understand what were the statements reacting to? Because there was little time to write, statements were very short and hard to contextualize" Cristina added that it depends on the discussion, and proposed – if this were a tool for groups or collectives it would be great if they could set the time themselves. Valentina thought the risk (with the quickness of it all) is to subtly force people to come up with truisms rather than articulate complex statements A lot of people in the chat agreed with that.

Comments from the chat:

  • emirhakin (he/him)> there were very vague statements
  • Valentina (she/her) > responded later in the process that "it would also be helpful if we could sometimes respond with "maybe" which would basically be a request for further nuance or details, since some of these statements were vague or unclear"

More comments and questions from the chat:

cristina (she/her) asked "which groups is this tool for?

Sepp (he/him) felt like we should maybe problematize the notion of 'group' a bit in this whole conversation, but that's for another time

frances (she/they) was struggling to spilt their mind between artist collective + union as they operate very differently yet are both groups

emirhan wanted to be able to reconsider previous statements... going back is not possible or reiterate

Sepp (he/him) noted he would also be able to type faster on a keyboard, but on a phone it just takes a bit more time.

juju she/her asked if it would be possible do this asynchronously and have a word limit? so instead of having limited time. James responded that they do have another version of the project where and it lasts several days

emirhakin (he/him) really liked the first gesture - waiting for everyone to arrive, and being aware of each other

Nienke said the moments of consensus are very important, but they are still working towards what to do with the final outcomes (perhaps through a visualization)

Looking into ways to form a group is also important... as part of the onboarding process.

Valentina (she/her) wondered if the result is clarifying what the main values/assumptions of a given group are. Perhaps those results to be valuable the moments of non-consensus should be highlighted too" +1 from Angela

James explained the technical idea behind WINWIN is that, the consensus should be maintained no matter the number of participants, and that even if the group changes or shifts, the system will be trying to maintain this. There seems to be a friction between the human side of consensus and the technological side of consensus

Would there be ways that groups could determine themselves what they want to discuss and set conditions for consensus? Super majorities for instance: 75% might be be sufficient for some groups. This would be going deeper into how is it that we are finding common ground.

cristina / ccl (she/her) remarked that the anonymity made her think this the platform is intended for a collective of people who do not know each other very well. Knowing who said what could also lead to further conversations outside of this application and the possibility to follow up on disagreements.

To join H&D meetups send an email to: info@hackersanddesigners.nl

The H&D meetups are informal gatherings for anyone with an interest in technical and artistic explorations of open-source tools and infrastructures that are meant to faciliate collaboration. The bi-monthly gatherings will specifically focus on unintentional use of such tools and question them in terms of their durability, accessibility, openness, and creative output. What is technology's partaking in collective organisation? How can we explore/build/hack tools that help small self-organized initiatives in their work?

H&D Meetups are kindly supported by Amsterdam Fund for the Arts and the Creative Industries Fund. Special thanks to Framer Framed for providing us the space, resources and infrastructure.