Executive Summary

The Guild of Servants is an organization founded in August 2020 aimed at providing structure and community to rationalists, and others with similar ambitions, in the midst of the COVID-19 pandemic. The Alpha Phase of the project recently concluded. 

The Alpha Phase was an experimental rollout and proof-of-concept involving ~35 volunteer participants, consisting of an online community component and an online course component. We have collected feedback from stakeholders in the form of interviews, and consolidated that feedback in this document.

The courses, as conducted in the Alpha, were determined to be too time-consuming and demanding, both for the instructors and the participants. It is recommended that the duration and intensity of courses be scaled down in subsequent phases. Large and complex courses such as the Character Sheet ought to be broken down into smaller discrete pieces. In this document we make further, more detailed suggestions for changing the format of the courses.

The cohort community structure, in which participants were placed into fixed cohorts of six members, was deemed a success. However, due to a natural expectation of attrition, it is recommended that the starting cohort size be increased to eight members, and that initial cohort sorting be based on mutual timezone compatibility.

The Alpha Phase was a process of learning and experimentation, and each founding Council member played many different roles. For the Beta Phase, it is recommended that the Guild leadership structure be reorganized such that each Guild Council member is assigned a specific administrative role, and that participation and contribution be compensated either monetarily or with ownership equity. It is further recommended that Guild Council members and other administrators not be expected to do both administrative and course-teaching duties at the same time.

The Alpha Phase is considered a success, based on our own stated expectations and goals in the planning. We avoided the pitfalls that we anticipated in our pre-mortem analysis. Before moving into the Beta Phase, it behooves us to recalibrate expectations toward a more ambitious goal, and specify our goals and vision as precisely as practical.


August 2020, a group of five rationalists founded an organization to provide a formal community and development structure for rationalists. Due to the restrictions of the COVID-19 pandemic and the belief that the Future is Online, we aimed to build an entirely-online community structure that was still valuable and meaningful. We named the organization the Guild of Servants, to communicate our intent to make ourselves valuable to the world.

Approximately thirty rationalists responded to our open invitation and took part in the Alpha Phase of the Guild of Servants. We cannot sufficiently express our gratitude to these people for their patience and enthusiasm as we tested out ideas for community infrastructure and course design.

The Guild of Servants Alpha was organized around three courses which were designed and implemented by the founding Council members. These courses were:

  • Fashion, a module which taught principles of personal aesthetics, and provided constructive feedback to participants. The purpose of the course was to encourage participants to pay attention to the ways in which they broadcast information about themselves, and to provide quick and affordable ways to improve in this area.
  • The Character Sheet. This module aimed to guide each participant through a process of uncovering their own unique strengths and weaknesses, clarifying their personal goals and ambitions, and generally aided them in leading a strategic and examined life. It was based on Alex Hedtke’s workshop material.
  • How to Spot a Con was a course rooted in Cialdini’s book Influence and designed to lead participants to be less likely to be exploited by social manipulation.

The courses were conducted via weekly Zoom sessions and relied heavily upon the six-person “cohort” structure. The cohort groups provided a framework for group work and group discussions, as well as a sense of camaraderie.


It was widely agreed that the courses were too demanding and time-consuming, on both the student and instructor side. A good target for the future would be less than one hour per week of total class time and less than one hour per week of homework. In future phases, students may end up taking more than one class at a time, but this should not be expected or default.

Each class should be a maximum of three sessions of one hour each. In general, classes should be short, focused and specific. Big topics should be broken out into multiple discrete classes. Something like the Character Sheet class, for example, could perhaps be expanded into three or more distinct classes.

It was recommended that all courses begin with a questionnaire to determine each student’s background knowledge and competency in the course domain. If possible, in future phases, students can be “tracked” according to their starting knowledge and skill set. Additionally, setting course expectations at the beginning of each course is an indispensable step which we did not always attend to. It was difficult to calibrate course materials without an understanding of where the class attendees were coming from, and difficult for students to calibrate their expectations without explicit course objectives.

Standardization of what a guild “class” looks like would provide a better sense of cohesion both between and within classes. Wherever possible, classes should be the same length, involve the same procedures and follow the same patterns. Deviations will happen, but only with intentionality and purpose.

Multiple council members expressed that the group breakout sessions would have been greatly facilitated by the presence of teaching assistants. These teaching assistants could be guild members who previously took the class and are serving in this role under some incentive, or could simply be cohort members tapped randomly or on the basis of competency.

Some instructors reported a feeling of imposter syndrome and/or inadequacy in the role of teacher. This is a normal reaction to placing oneself in the role of “expert”, and it should be kept in mind that all good teachers are constantly revising and refining their teaching materials. The leadership should focus on keeping up the spirits of course instructors who may be experiencing difficulties.

Purely elective online courses must be regarded as a kind of “infotainment”, with heavy emphasis on the “entertainment” half of the equation. Recommend shortening/tightening classes and punching up their enjoyability level.


This was widely agreed to be the most popular and fun class with the highest degree of engagement. It was suggested that a professional fashion expert be paid to give feedback at specific junctures to avoid groupthink and also enhance perceived credibility.

Character Sheet

All agreed that certain elements of the Character Sheet class ought to be made more central to the mission/vision of what the Guild is trying to do. More focus should be placed on taking inventory of the student’s pain points and weaknesses, and also understanding what their goals are. Something like a “skill tree” concept should be implemented only after understanding what skills the student is actually interested in obtaining.  We can and should provide a “guided wayfinding” approach that locates the inductee and then leads them down a path toward their goals. In short: the Character Sheet content is important to the Guild mission, too important to fit into one class.

People didn’t seem to understand the purpose of certain portions of the Character Sheet. The middle section on strengths and weakness was remarked on as something that people got hung up on.

How to Spot a Con

Regarded as having portions that were fun and hands on and also portions that were dry.

It was suggested that this class could benefit from some social manipulation games or inter-cohort competitions, potentially with prizes. The lifestyle restrictions caused by the pandemic screened off a number of potential “field test” style activities.

Proposed actions:

  • Break the Character Sheet class out into several short modules, with more focus on locating people where they are at the beginning and putting them on track.
  • Consider hiring domain experts for specific parts of classes where expert-level feedback would be valuable.
  • Implement inter-cohort competitions as often as possible. People find competition motivating, and we should take advantage of the cohort structure.
  • Define a fixed template for what all courses should look like, structurally, with allowance for necessary deviations.


All Guild members were placed into cohorts of ~6 people, based on research in optimal task force size. These cohorts were the organizing unit for almost all Guild activities, including class assignments and group discussions.

The cohorts were agreed by all to be a fundamentally good idea in need of a few tweaks. A significant degree of variation between cohorts in terms of involvement and interaction was observed.

There were a small number of minor personality clashes reported within cohorts, which seem to have been addressed and rectified with minimal cognitive-emotional overhead reported by the involved parties. Some council members reported these events as being more stressful than others. Perhaps some people are just better than others at dealing with such conflicts.

This author thinks Discord and other Slack-likes, e.g. Microsoft Teams, are a cancer and would rather encourage people into having scheduled video-calls, phone calls, or practically anything other than just being passively online all day. Rationalists Don’t Let Rationalists Be Online All Day. That said, I recognize the inevitable tension between this stance and the goal of creating an online community space that provides a sense of purpose and cohesion. I am not against the existence of the Guild Discord, I merely recommend that we don’t actually rely on it.

Proposed actions: 

  • Increase the cohort starting size to 8 people to help with attrition and low investment.
  • Implement a policy whereby cohorts are re-sorted and consolidated after the first ~3 weeks, such that no cohorts suffer from low active participation.
  • Cohort members should be assigned by time zone compatibility.
  • Delegate the resolution of interpersonal conflicts such that the Council does not need to get involved unless absolutely necessary.
  • In general, leverage inter-cohort competition more effectively; this concept is discussed further in subsequent sections.


It was generally agreed that the demands of simultaneous administration and teaching were untenable. Nearly every Council member reported feeling stressed and overworked. In the Beta Phase, the Guild Council will address these issues by the following strategy:

  • Assigning permanent roles, such that each council member only focuses on one set of administrative tasks, and teaching/course-prep is only undertaken by individuals with no other commitments.
  • Increasing the number of Council-level administrators to spread the load out.

Even with this high degree of subjective overwork, we were generally dissatisfied with our own level of organization and logistical reliability. Specifically, we observe that students were rarely clear on where to look to find important information about courses, expectations, or scheduling. This issue will be addressed by placing one individual in charge of the student-facing organizing content, e.g. calendars and course notes.

The points system which we started out with was essentially abandoned. This was partly because of the lack of clarity in what points meant, partly due to the fact that administering and tracking points was far more time consuming than expected, and partly because points just didn’t seem to be very motivating to people. The point system itself may be a good idea, but it was everyone’s last priority and so its implementation floundered. The points/ranking system should either be dropped, or it should be prioritized explicitly, but it does not function effectively as a vague background element.

The Incentives Problem

The Guild of Servants Alpha actually exceeded all of our expectations, since we all predicted in our pre-mortem document that becoming overwhelmed and/or losing interest would be the most likely failure mode. Most of the Guild Council remained extremely committed and devoted large amounts of time to Guild business on an ongoing basis, and a large fraction of Guild members maintained attendance and interest through to the end.

That said, relying on volunteered time and sustained hard work for no compensation is unlikely to be a viable strategy as the organization grows. It is currently not possible, for example, to simply “assign” someone the job of improving the website, or keeping a calendar updated. The completion of all such tasks is dependent on interest level and willingness to donate time. We are all aware that we need to solve the incentives problem, but there is no broad agreement on how to do this.

A few possibilities include the following:

  • Adopt a standard startup business structure, giving Council members “founder” status and allocating ownership share accordingly, and use ownership shares as an incentive for new members. Voting shares can be separated from ownership shares to give the Council more permanent control, if this is deemed wise. C-Corp startup structures are a tried-and-true framework, and a small share of a business entity that might grow very large is an empirically motivating incentive.
  • Begin collecting Guild dues now, in the Beta phase, and pay out dues in return for Guild service. This is, frankly, unlikely to be very motivating at first, since the dollar amounts will be very low until membership ramps up.
  • Raise capital. Seek out and secure a round of seed funding. These funds could be used to pay people fair wages for their work. Essentially we would reconsolidate the Guild as a venture-backed startup.
  • Set aside the above proposals, and spend time researching monetization strategies for broadly similar organizations, then pick a monetization scheme that we agree could work and reorganize around that.

Beta Roles

We need individuals to take responsibility and ownership for critical administrative functions. Below are listed a set of roles, some of which have been assigned at the time of writing.

  • Scheduling and Calendar manager. (Consult with all stakeholders to schedule meetings and courses. Consolidate all scheduling information into one, unambiguous calendar. Promptly address arising issues, such as instructors calling in sick.)
  • Student advocate, or, Course design director. (Red-team the course design. Oversee and manage the structure, duration, and demands of all courses, to maintain a consistent vision and implementation across the Guild courses.)
  • Technology director. (One person who oversees and makes the final call on issues relating to tech platform choices, and, possibly, handles technically difficult and repetitive tasks such as uploading YouTube videos to the channel, or administrating course surveys.) 
  • Cohort activities director. (Focus entirely on the cohorts. Are people having fun? Are the cohorts the right size? Are we constantly running some kind of inter-cohort competition? Can we host regular Street Epistemology sessions within cohorts? Forecasting tournaments? Good-works projects? There’s a lot of room for possibility here.)
  • Business manager. (Handles payments, billing, fees, invoices, approves reimbursements, officially sets prices and controls monetization streams.)
  • Internal Guild Prediction Market manager. (Of course we should have one of these.)
  • I am almost certainly missing some potential valuable roles; please suggest more.


The Council is open to revisiting everything about Guild outreach and branding, including the name “Guild of Servants”. It has been agreed that we ought to hire a PR expert for a consultation, although we have not actually priced this out, nor determined where the funds would come from. (This author cautions that we ought not to expect magic; the purpose of a consultant is often to simply say obvious things with great authority.)

The twee approach (silly hats, cryptic Cyrillic letters) served to filter a certain kind of person for the Alpha. This seems to have worked well, but is unlikely to work in exactly the same way as we grow. As discussed in the next section, we shouldn’t be afraid of seeming weird, as long as the weirdness is intentional.


The following is an attempt to consolidate the different but often overlapping statements of vision and mission provided by the Council:

“The mission of the Guild is to serve the Guild members by providing a structure which facilitates their growth into impressive human beings, according to their own needs and desires. A secondary but also important mission is to provide public benefit by encouraging/incentivizing Guild members to do “good works”, improving the world and raising the sanity waterline.” 

Most council members agree that we lack focus and strictly lack a solid elevator pitch. If I had to distill everyone’s remarks down to something like a pitch, it might look like this:

“The Guild is an organization of people who have the same meta-goal of making themselves more rational and generally impressive, and it is innately valuable to push such people in contact with each other, providing a support structure.”

Nearly every council member expressed something like a wish for the Guild to serve as a monastery-school for rationalist-warrior-monk-Bene-Gesserit-Mentat-Freemasons. This is a good “a computer on every desktop” sort of inspirational target, but doesn’t contain many gears.

Sometimes in order to define what a thing is, it is important to define what it is not. Now is the time to really focus on this. If we don’t define the vision clearly and succinctly, then we will simply pursue the path of least resistance, and our failure will not be a failure of execution but rather a failure of direction; we will look up and realize that we have traveled a thousand miles on the wrong heading. The Beta Phase will be an opportunity to more carefully explore and define the Guild vision, mission and elevator pitch.

A Possible Model

This is my little corner to editorialize. I’m attempting to find a golden path that connects what we have, and where we want to be. 

In my vision, the Council transitions to being administrators with defined roles. The function of the Council is to:

  1. Serve and support the cohorts.
  2. Facilitate and empower the teachers.

Teachers would be volunteers who come to us with a plan for a course. The Council helps them hammer out their lesson plan, their schedule, their assignments, and integrates their course with the Guild resources, calendar, website, YouTube channel, etc., including Mentat (an extensible teaching app developed by yours truly).

The course materials would exist as static YouTube videos, notes, and Mentat modules. “Class sessions” cease to be live lectures, and are instead short, scheduled check-ins with teachers/TAs to give feedback on assignments, ask questions, etc.

These course materials (or at least the YouTube videos) would be freely available to all, serving as advertising which drives/upsells people into paying for Guild membership in order to get the full benefit.

The Guild thus becomes an ever-growing library of curated, well-designed and thoroughly tested course content, and a superstructure of cohorts who are supported in their own ambitions, and in taking the courses. In this way, one can be “in the Guild” and take no courses at all, instead simply participating in the cohort activities (competitions, projects, etc.).

The business funnel consists of free YouTube videos that aim to convert people into paying Guild members, and then provide ongoing valuable support and other services to those Guild members to retain and develop them.


The Guild of Servants Alpha Phase is a strict success based on initial stated ambitions and expectations. The primary noted points of failure were in administrative inadequacy, and excessively high expectations for the courses. Other concerns include expected difficulties in scaling the current design up to serve more Guild members, and a lack of specificity in monetization and incentivization going forward. All things considered, we are highly optimistic that the Guild can be sustained and grown if appropriate steps are taken. 

The Guild of Servants Beta Phase is set to begin in the summer of 2021, and will include a greater course variety and expanded cohort activities. We plan to scale up for many more participants this time, as well. Please contact us at guildofservants@gmail.com if you’re interested.

Thanks to Alex Hedtke, David Youssef, Errol Highberg and Gray Morrow for their service to the Guild, and thanks to each and every participant in the Alpha Phase for their patience and enthusiasm. Thanks to Steven Zuber for his feedback on this document.

Thanks for reading,

Matt Freeman

New Comment
8 comments, sorted by Click to highlight new comments since: Today at 9:51 PM

This is great, thank you for writing this up and I am looking forward to where this goes! 

(I probably have more detailed thoughts, but not sure whether I will get around to writing them up, so it seemed better to leave encouragement instead of nothing)

I was in the alpha and found value in both the community component and the modules. I had the pleasure of seeing people bring problems they were facing to the group and watching what it's like to have a group of Rationalists coordinate and descend on it together.  It's hard to overstate the value of having a dozen people with practiced thinking skills work on a problem with you, but I suspect the reader of the comment can imagine what I'm talking about. 

The modules were varied and valuable in their own way. Like Matt indicated above, there were some areas where they could be improved, but from my perspective I think The Council is being too hard on themselves. Many of the issues with practicing the fashion and How to Spot a Con modules were due to lockdown. While I suppose one could make the case that The Council could have seen this complication coming and planned for it, none of us really knew how long we could expect the Plague Months to last when this project was being scoped. I suspect the lack of clarity with the Character Sheet will be soundly cleared up by the time the beta launches. (I strongly recommend Alex's Tsuyoku Naritai podcast, which I suggested he make required prerequisite learning for that module. It outlines the goal and methods of the Character Sheet brilliantly.) 

I explain all of this because I think that our community as a whole will benefit from both pillars that the Guild is looking to build. Having a coordinated group of similarly-trained allies to collectively work on problems will be invaluable and I can easily imagine a future where the resultant successes are a huge selling point going forward. Likewise, having a sort of training camp where the techniques (in these modules and future modules) are actively taught can rapidly reduce the time it takes newcomers to become adept in the Methods of Rationality. I took away actionable lessons from the modules that I didn't glean in my almost-a-decade of being somewhat involved in the community. 

I would go on, but then I'd worry that the word count would warrant its own post. I also want to extend a sincere thanks to both the Council and my fellow Guild members. I very much look forward to watching this project grow. 

Would a module on how to overcome the deficits of autism without losing its strengths or becoming a different person be useful for the Character Sheet section of this project, or useful to Rationalist-aligned autistic people in general?

Briefly, the Character Sheet module didn't focus on any particular deficit for the entire cohort, but rather allowed each participant to identify their own weaknesses through some helpful guided exercises. Overcoming these completely was not in scope for the duration of the alpha, but we put together some docs and spreadsheets to track our progress going forward. 

If you would like to attend a Guild mixer to meet the Council and some of the students, come join us saturday! We expect to do this on a monthly or quarterly basis.


You abused passive voice too much. In the following examples, you made it unclear who suggested these ideas and whether the guild council agrees with them and plans to implement them.

For the Beta Phase, it is recommended that the Guild leadership structure be reorganized such that each Guild Council member is assigned a specific administrative role, and that participation and contribution be compensated either monetarily or with ownership equity.

It was recommended that all courses begin with a questionnaire to determine each student’s background knowledge and competency in the course domain.

It was suggested that a professional fashion expert be paid to give feedback at specific junctures to avoid groupthink and also enhance perceived credibility.

It was suggested that this class could benefit from some social manipulation games or inter-cohort competitions, potentially with prizes.

I do not love passive voice either, but the nature of this document is:

We have collected feedback from stakeholders in the form of interviews, and consolidated that feedback in this document.

If there is ever a place for passive voice, it is a document whose purpose is to consolidate the opinions of multiple people while explicitly not implying definite consensus among those people.

In a document that consolidates opinions of multiple people I think it's useful to be specific about how many people propose an idea. Instead of "It was recommended" you could say "One person recommended" or whatever is the most appropriate. That would make it much more clear to the reader how much the recommendation is backed.