I grew up with my parents raising me in the liberal Quaker tradition. Even now, looking back from an atheist perspective, I believe there are valuable lessons in the Quaker practice. Especially around how to get a small (<50) group of people to really listen to each others viewpoints and come to consensus (or as close as can be reasonably expected in the case of strong fundamental disagreements).
Although I enjoy the practice of Meeting, I actually really disagree with you about Quaker practices around decisionmaking. My local Meeting had some huge disagreements around COVID that weren't resolved at all well; from that and how disagreements are handled in general, it almost seems to me to be more of a Tyranny of Structurelessness[1] kind of situation, where conflict is handled via backchanneling and silently routing around disagreements and leaning on people who disagree to let it go.
Frankly I just don't think consensus is a good decisionmaking method at all.
[1] https://www.jofreeman.com/joreen/tyranny.htm
Well, I do think it has significant weaknesses, such as being vulnerable to falling into bad patterns such as you have discussed. I've also seen it go poorly sometimes. But when it works well, I've found that it works really well at getting people to understand each others viewpoints and genuinely build empathy and make people feel happier about compromises that must be made, and make people do a better job at searching for novel win-win solutions.
But I agree that it can't be used 'as is'. I just think that the core elements are something worth keeping in mind for designing novel decision-making systems.
For instance, what if the participants each had an AI advocate who was trained on their own point of view, and these AI advocates went through a simulated consensus process at 1000s of times human speed, and then the result was a personalized report for each human summarizing the others points of view in a way designed for the particular person to be able to understand, and suggesting a compromise maximized for win-win solutions.
Related: https://arxiv.org/abs/2211.15006
Fine-tuning language models to find agreement among humans with diverse preferences
Michiel A. Bakker, Martin J. Chadwick, Hannah R. Sheahan, Michael Henry Tessler, Lucy Campbell-Gillingham, Jan Balaguer, Nat McAleese, Amelia Glaese, John Aslanides, Matthew M. Botvinick, Christopher Summerfield
Recent work in large language modeling (LLMs) has used fine-tuning to align outputs with the preferences of a prototypical user. This work assumes that human preferences are static and homogeneous across individuals, so that aligning to a a single "generic" user will confer more general alignment. Here, we embrace the heterogeneity of human preferences to consider a different challenge: how might a machine help people with diverse views find agreement? We fine-tune a 70 billion parameter LLM to generate statements that maximize the expected approval for a group of people with potentially diverse opinions. Human participants provide written opinions on thousands of questions touching on moral and political issues (e.g., "should we raise taxes on the rich?"), and rate the LLM's generated candidate consensus statements for agreement and quality. A reward model is then trained to predict individual preferences, enabling it to quantify and rank consensus statements in terms of their appeal to the overall group, defined according to different aggregation (social welfare) functions. The model produces consensus statements that are preferred by human users over those from prompted LLMs (>70%) and significantly outperforms a tight fine-tuned baseline that lacks the final ranking step. Further, our best model's consensus statements are preferred over the best human-generated opinions (>65%). We find that when we silently constructed consensus statements from only a subset of group members, those who were excluded were more likely to dissent, revealing the sensitivity of the consensus to individual contributions. These results highlight the potential to use LLMs to help groups of humans align their values with one another.
My take after skimming the paper: this focuses on figuring out what opinions people do hold, not what opinions people should hold given their values. For instance, asking people things like 'Should we increase taxes on X with the intent of causing Y?' gives you information about how people weigh the downsides they perceive in proposed solution X versus the upsides they expect. They could be totally wrong!
This isn't so much a critique of the paper, but a note of caution that the topic the paper focuses on is a small and insufficient part of what making an actual governance decision should depend on. To me, a more interesting question which could be focused on would be trying to figure out the underlying values and relative weightings of different costs/benefits that lead people to have the opinions they have.
I wrote this up for a meetup I ran recently, where we held a Quaker-style Meeting for Worship session. I recommend doing this in a group setting, but you can also do it by yourself if the ideas appeal to you.
Why Quaker meeting
I’ve had depressive episodes throughout my life and one of the few things that seems to consistently help is going to Quaker Meeting.
Some of this may be the community (religious practice of any kind is supposed to be helpful). But I’ve found it useful even when I participated very little in the community around it, even just doing Quaker practice by myself.
I don’t think this is true of everyone who does it. But it works for me, and I have some idea why. For this meetup, I want to try to convey some of that to you, and teach you how to sit in silence in a way that makes your life better.
I have an ulterior motive here, which is that while I do care about and enjoy the Quaker community, there are certain deep divides between me and it that sometimes make it hard to feel comfortable. I think I’m more in sync with most rationalists, and I imagine there are parts of Quaker practice that would be a lot easier to do with a community I feel more generally at home in. So I’m hoping I can get some of you interested in doing this more.
What we’ll be doing:
The usual Quaker advice
Disclaimer: There are lots of different explanations or suggestions for what to do during Meeting for Worship. Quaker practice has a long history and I’m frankly not an expert in it. But I went to Quaker schools growing up and developed my own approach, based on the stuff my teachers and other adults told me. I’m going to share that approach here.
Second disclaimer: This is NOT meditation. There are definitely some similarities, especially in the “settling in” part that I’m going to describe, but the important parts are pretty different, from what I have read. (However: Quaker meeting is ultimately a practice of “sit in silence” and I can’t police what’s going on in your head! And plenty of Quakers historically and currently have done stuff more akin to meditation. But it’s not what I’m going to tell you about.)
General Quaker advice is that there is “the light of God in everyone”, and that during Meeting for Worship, your goal is to listen to that “inner light.” Or to listen to a “still, small voice” inside you. Or to sit in a state of “expectant waiting,” not thinking about anything in particular, but just staying in readiness for divine prophecy to come to you.
I’m not kidding about the “prophecy” part. It is a part of Quaker belief that every person has it in them to be an instrument of communication from God. And if you receive such divine communication, and it feels very important, then you share it, as one is usually supposed to do with divine prophecies.
This probably seems like nonsense from an atheist perspective. Without the God stuff, the bare bones are: you sit in silence, and wait for ??something?? to happen, and if it does, maybe you speak.
This is still pretty vague. So let me try to describe what actually happens inside my head during Meeting for Worship.
What I do
During the first few minutes, I usually take some time to “settle into the silence.” Whatever thoughts have been buzzing around in my head will take some time to resolve. Maybe I’m thinking about an art project I’m working on. Or something I need to buy. Or that task that I need to get done. Or that thing that person said to me last week.
If I’m being honest, sometimes this stage lasts the whole hour. It depends on how I’m feeling that week, how tired I am, and how much stuff is on my mind. I do my best to just watch the thoughts go by during this period. (One meetup participant reported that this happened to them, but they still felt it was a good use of time. “One hour of not being on Twitter.”)
On good days, though, this phase elapses after a while, and I start to be able to go deeper.
To get there, my current practice is to let the thoughts drift by and for each one, notice it, then ask myself “Why is that important?”
Oftentimes the answer is, “It isn’t,” and I let it drift by. It’s okay. Thinking about things that aren’t important is the natural human condition. But when I find one that is important, I focus down on it.
What makes it important? What do I care about related to this thought?
When it feels like I have enough mental space between buzzing thoughts to think more deeply, I also have a couple general prompts I use for myself:
What’s important in my life right now? How are things going in general?
This could be seen as a form of Hamming Questions, or a kind of self-one-on-one if you want to use corporatespeak. But it’s less judgmental and pointed than the Hamming Question of “... and why aren’t you doing it?” Instead, just have curiosity about what’s important and why.
As I notice things that feel uncertain, or unresolved, or confusing, or that are bothering me, I try to make more effort to focus on those and ask myself questions about them. What is going wrong there? What’s the problem? And then usually not, “how do I solve it,” but instead, “how can I think about it differently?” (If you’re familiar with Gendlin’s Focusing, you might recognize this as being very similar.)
I find that I regularly have important and useful insights about my life this way. Sometimes I naturally end up doing some thinking or planning that I’d been putting off, or coming to some decision that I was trying to avoid.
And the feeling of being in tune with myself just makes me feel better about everything, even when I’m upset and miserable. Maybe it’s just the silence and the space. Whatever it is, it seems to work for me.
I think it really matters that I have a long time in silence to think about these questions, and also that there are no distractions at all. I could maybe imagine doing this process with a pencil and a blank piece of paper to help, but certainly no more than that. And even that feels like it would be unhelpful, at this point, for my focus. It’s too easy to get distracted by what you have down on the paper so far, instead of what’s in your mind.
Speaking
The second part of Quaker practice, which I haven’t talked about yet, is to speak when you feel moved to. Or “when the Spirit moves you,” if you are a theist Quaker (which is most of them).
I’ve spoken very rarely in my life. There seem to be very different social norms around how often people speak in different Quaker meetings; at the Meeting I go to, it’s most common for no one to speak at all, and after that for only one or two people to speak. Usually the windbags (affectionate). I’ve heard that at some Meetings several people speak every Sunday.
For me, there are a lot of prerequisites before I want to speak. Usually I have to have reached an important insight, and decided it’s something that applies to other people and not just myself, and then taken the time to formulate it a little bit in my head before I talk. If it’s powerful, I do usually try to say something, but it often just doesn’t happen anyway.
I do want to encourage people to speak, because I think it’s probably better for the Meeting as a community practice if people are sharing more. So if you had an internal bar you were thinking of for “how often I would talk,” maybe lower it a little. Unless you’re someone who talks a ton all the time, then maybe raise it a bit.
If you do speak, unless you’re someone who talks a lot, you can probably speak for as long as you want. Usually it feels longer to the person speaking than to everyone else. But don’t recite entire speeches or sing entire songs or anything. In olden times Quakers would give quite long speeches during Meeting, but our time here is limited, so let’s not deliver entire lectures.
If someone else speaks, don’t speak right after them. Leave some time for everyone to think about what they said and settle back into the silence. Part of the goal here is that Meeting for Worship is not supposed to turn into a discussion or debate; it’s a time for reflection. Oftentimes people will end up talking on the same topic, because they are inspired to, but please don’t give any point-by-point rebuttals. You should be sharing your own insights, not just responding to someone else’s.
Queries
Regional groups of Quaker Meetings publish a Faith and Practice document every few years, which contains information on what the Quakers in that group believe and practice. They also typically contain a section called Queries, which are question prompts to get you started during Meeting for Worship. Specific Meetings will often have a practice of reading a few of them before Meeting once a month or so.
Here are some from Baltimore Yearly Meeting Faith and Practice that I thought might serve as helpful jumping-off points for this group.
I don’t agree with the premises of all of these, so take them as you will.
Housekeeping for Meeting
I will be keeping time on my watch and will let you know when it’s over. We’ll go for one hour.
If you need to go to the bathroom or anything, please do that beforehand.
Please turn off any glowing rectangles or demon boxes you may have. (One person asked if they could use one if they got too understimulated during the hour. I requested they go to a different room if they needed to do that, but said fidget toys were fine.)