How To: A Workshop (or anything)

Author’s note: while the whole piece is a 55min read, there’s a ~7min standalone essay at the top, and then a break and a shift into a different kind of content. You really actually for real can just read the 7 minutes and then stop, if that satisfies the need you had going in. Or any other reason!


Back in February, I provided ops support and light-touch facilitation for a handful of mathematicians on a research retreat. Just prior to that, I helped MC a section of a two-week program for a batch of promising young EAs trying to refine and improve their long-term plans. Just after, I was the sole instructor for an abridged CFAR-esque applied rationality workshop for nine Circlers in southern California.

Two weeks ago, I solo led a non-abridged, 4.5-day workshop for a little over twenty participants, mostly EAs in their early twenties (thanks to Akash Wasil for laying the groundwork to make it happen).

By all accounts, my impact on each of these four very different events was strongly net positive. And as I reflected on each of them, it became clear to me that there was actually a fairly small number of things-going-on-in-my-brain that led to the good things that happened.

(Approximately just one, actually.)

I’d like to make that mental motion available to others, as I think it’s useful for a wide range of people and pursuits, including ones that don’t look like retreats or workshops at all.

So. How To: A Workshop (or anything, really).


While working as the curriculum director and workshop lead at CFAR circa 2017, I was attempting to hand over the reins to a young up-and-comer who asked me to write down what I was thinking about and focusing on in my role as MC. This was the result:

...this is a lot. My colleague was surprised, and a bit daunted.

And I do, in fact, believe that every item on this board is “pretty important,” and they were all things that I was actively tracking or doing during approximately every workshop. I think that a workshop where any one of these things gets dropped is “noticeably” worse, i.e. staff or participants would be able to say “oh, yeah, that thing could have been better,” even if the overall experience remains solid.

Pet theory, though:

The greatest value that a teacher or mentor can provide is usually in giving you discriminating principles that allow you to confidently ignore everything except the current most important thing to focus on.

(This claim by itself might be fully 13 of the value of this post, in expectation? I use something like this belief to select from among potential teachers and mentors and to better extract the wisdom I want from the teachers and mentors I have available.)

When you are first starting out in a new domain, there’s often simply TOO MUCH INFORMATION coming in, and you often have no conceptual framework for digesting it, no buckets for the information to sort itself into, and no knowledge of how to effectively prioritize. You’re trying to digest a dozen different things at once, which means you tend to fail at all of them for a while.

A good teacher can help by saying “forget everything except X for the next hour.” And then, after an hour is up, “okay, now don’t lose track of X, but start gradually diverting some of your spare attention to Y, as well.”

And indeed, there is one single mental motion underlying all of the above; a single place where my attention stays during a workshop which, given sufficient context, generates all of the various items on the whiteboard. It generated all of them, in the moment, as I mentally traced my way through a Typical Workshop—I wasn’t remembering what I usually did, or running through a mental checklist; I was running a function that outputs different answers at every step of the process of running a workshop.

If you, starting out in your Endeavors, can only juggle a single ball, it should be this one:

Do you know what you are doing, and why you are doing it?


Attendees of the recent workshop (and many past Duncan-led events, too) will recognize the above line; it was the catchphrase of the shoulder Duncan I attempted to give them over the course of the four-and-a-half days.

“Do you know what you are doing, and why you are doing it?” is both the skill I was trying to impart, and (practicing what I preach) the skill I was employing, in the imparting.

It’s the question I ask myself, at every step of the process (when considering/​conceiving of a project, when fleshing out the details of the plan, and during the execution of that plan). It’s the question I ask on behalf of others, when thinking about participants or volunteers (“Will the people involved know what they are doing, and why they are doing it, at approximately every moment?”).

Any time the answer is “no,” that means that something different needs to happen.

(Clarification: it’s not actually that I literally ask myself this question, in actual words, and then answer myself, in explicit statements. That’s how you start, when you’re first building this habit, but it very quickly just becomes a wordless groove or vibe that you occupy, the source of gravity that lets you know which way is up. Put another way, if you get into a mode where the answer is always yes by default, because it’s what’s generating your actions in the first place, you don’t need to pause and ask yourself in words.)

If the people you have invited into an experience understand what they are doing and why they are doing it at every step of the way, they will at the very least come away from your workshop/​conference/​performance/​event with mildly positive feelings. You will at least dodge the failure modes of [things fell apart] or [people found it pointless/​useless].

And if you started asking yourself this question early enough in the process, such that you have a justified sense of what you-and-they should be doing, and why—

(i.e. your event is happening for reasons that have something to do with their actual goals, and the elements of your event are grounded in real-world causality such that they will result in actual motion toward those goals, and all of this is done in integrity with either your or their genuinely held values (and ideally both))

—then they will not only have some positive feelings but will have also acquired or experienced something that is real and good.

This is, quite literally, more than 95% of what I am doing with my brain and my attention while bringing a workshop into being. From the moment I begin planning, right up through the closing ceremonies of the last day, I am asking the question “what should they be doing, and why?” with the related subquestion ”...and how can I best cause that to happen?”

Everything flows from this.

(This question, and the fact of its centrality/​importance, is the second third of the total value of this essay, in my estimation.)


I know that (probably) sounds too simple to be useful.

I have an illustrative anecdote.

Once, my afterschool parkour club student Zach asked me to critique his climbup—the motion of hanging on the edge of a wall, and then pulling yourself up and over so that you can get your feet on top.

There was nothing wrong with his form or technique. He just needed practice. I told him this. He was a little disappointed. Practice is boring and tedious; what we all want is One Weird Conceptual Insight that unlocks a bunch of potential we had all ready to go.

However, a month later, Zach showed up at parkour practice and spent the afternoon running around, showing-and-telling to every other student. “YOU GUYS,” he shouted, over and over again. “Guess what! I practiced, and I got better! LOOK!”

And then he’d do an effortless-looking climbup.

Zach was genuinely surprised to discover that the whole just-do-it-diligently-for-a-while-and-you-will-often-see-results thing wasn’t, in fact, a myth.

The rest of this essay is just going to be a case study in flowing from [this question] to [all the relevant answers], in the context of putting on an applied rationality workshop. I’m basically excerpting and annotating the notes that I sent to my friends as the event of two weeks ago unfolded.

It’s worth noting, though, that even though the rest of this is a super specialized and specific set of answers, this is in fact a general process. It’s an only slightly more refined and mature version of the process I used, as a fourteen-year-old, to recruit the willingness of adults as I taught them Tae Kwon Do. It’s the same process I used as a sixth-grade teacher to deliver useful educational experiences to my eleven-year-old students. It’s the same process I use when scheming adventures with friends, or when giving talks at EA Global or universities, or when guiding a group through a parkour lesson, or when writing essays such as this very essay.

As noted above: more than once, when I have had only an hour to try to introduce people to the gestalt of applied rationality, my main method has been to install something like a shoulder Duncan who will ever-after pop up at opportune moments to ask “do you know what you are doing, and why you are doing it?”

It’s fine, sometimes, for the answer to be “nope!”

But not, usually, in the middle of trying-to-make-something-happen.


The Rest Is Commentary

It’s genuinely plausible, at this point, that you’re done and should stop reading. If you do continue reading, I recommend that you do so for some known-to-you reason, even if it’s not expressible in words or sentences.

The two broad categories of readers I most expect to benefit from reading further are:

  • People who are specifically interested in stealing some fraction of my event-coordinator skill, or who want to know more about workshop curation in particular

  • People who do not know how they would begin applying the question above in practice, and would benefit from seeing it in action via concrete examples


Day −20: Preliminary musings

Okay. The Circlers. The Circlers needed something like reassurance that rationality wasn’t looking down on them, evidence that rationality could be useful to them, and confidence that they could build an API between what-they’re-already-doing and what-I-had-to-teach-them.

Why?

Because without those things, they wouldn’t be in a position to receive the rationality, and I genuinely believed that the tools I had in my toolkit were ones they should use, and would benefit from using.

The baby EAs. The baby EAs will be Different.

Why do they even want a workshop? Do they even know why?

Probably not. Probably this is some combination of having heard the hype from their Elders, thinking that this is something they’re really Supposed To Do, and maybe a little bit some of them saw me at Icecone and wanted More but honestly I think that’s like four out of the twenty-some people.

Okay, so what do they want?

Er. To Be Good, prolly. That’s like the cardboard cutout, anyway. Like if the Circlers want to feel and emote and Be Alive, the baby EAs want to Be Good, which means Doing Good, which means Efficacy, which means Go To A Workshop.

Okay. Hmm.

I said “yes” to Akash. Why?

Because I wanted to use this skillset that I have, that I’m proud of, wherein people tend to say nice things and tell me that I helped. Because when I successfully help other people I feel like I wasn’t Being A Waste Of Space for once. Because I need $ because the frigging A/​C install is going to cost way too much money that I don’t really have to spare.

Okay. …...I’m not sure there’s much to be done, at this point. Like, what do you want out of this workshop that’s different from a mainline workshop?

Nothing. Also, as a bonus, if I run “just a mainline workshop,” then I have, like, the jumping-off data for other weird experiments/​proof that I can run a mainline workshop (the Circlers one maybe doesn’t count since there were only nine of them and half of them knew me and they were all soul-touchy; this is more like A Real F’real Workshop with 20+ people who have no real reason to like or trust me).

Okay, so. Mainline workshop it is (and we know how to run those); we’re not going to think about this too much over the next few weeks.


Day 0: Planning opening session and the general gestalt

I am sitting on the train, thinking about the thirty-five-or-so opening sessions I feel like I’ve either run or been a part of, and wondering what will make this one different. What specific strokes of the brush should be unique to this group and this context.

What I know, courtesy of Akash:

  • About 25 ppl. Mostly college students or recently graduated.

  • Mostly people who want to devote their careers toward longtermism/​x-risk stuff. Many have become part of the community in the last 6-12 months but have gotten involved quickly and are considered promising.

  • Most people know each other from EA events/​retreats. Many are friends/​colleagues. A few newcomers.

  • Most people have not formally engaged with rationality content. Most people have not read the sequences or the CFAR handbook. They have read some LessWrong posts and have practiced a few specific skills at retreats (ex: goal factoring).

  • Generally truth-seeking and motivated and ambitious and take this stuff seriously etc

What my brain claims this tells me:

The common failure mode of needing to be “sold” on the content will be rarer than usual. However, the common failure mode of having a specific preconception about what they’re opting into may well be higher; these are less people who are “up for an experience” and more people who are expecting to purchase a product (relatively speaking).

The failure mode of having specific preconceptions or inoculations against chunks of content will also likely be higher; these are people who’ve heard the five-minute, badly-explained version of goal factoring and (somewhat justifiably!) believe they already know everything they need to know.

The failure mode of collapsing into social goo or trying to impress one another or playing follow-the-leader will also likely be higher. Not every change is an improvement, but every improvement is necessarily a change; with all of their regular social connections and lots of familiar faces around, it will be harder to escape their usual headspace and try out anything different. I will include the phrase “if your current skills and ways of being were sufficient to solve all of your problems, you wouldn’t still have those problems.”

Interestingly, this nudges me toward wanting to make “the flag and the pledge” a part of opening session, which is not a thing I’ve ever done before. The flag and the pledge is a mental model that helps people smooth out the proxies—we explicitly acknowledge that “the goals of the workshop” or “the goals of Duncan” are not exactly aligned with the goals of any given individual, and that each individual will have goals that are not exactly aligned with any other, either. But we instead adopt a temporary, limited goal that is a thing we all agree to strive for, because a) it’s close to a least-wrong compromise between everybody’s goals, and b) coordination will help us, instrumentally.

[Note from the future: I kinda botched this, as it was my first time presenting it in the opening session context. I don’t think it made things worse, but I don’t think it made things as-better-as-it-could-have if polished.]

I’m still inclined to do the usual “what is applied rationality” stuff, like with Wason 4-card and so forth, and present the frame of learning one’s tools and what they’re good for and when and how they break and so forth. I’m inclined to talk about each individual concept using the metaphor of kata (movements that you will graduate from using fairly quickly, but which are a good stepping stone to fluid skill).

But I’m also inclined to add in much more of a Logan-influenced slant of “do not lose track of what you are actually trying to accomplish, in the hectic back-and-forth of participating in the dance steps.” I think I might try, right from the start, to do the [explicitly installing a rudimentary shoulder Duncan] thing, and for this particular purpose I’m enamored of the catch phrase “do you know what you are doing and why you are doing it?”

(With added caveat that this is a yes-no question; it is not always the case that you should know precisely what you’re doing and why, but you should always know whether you know what and why.)


An aside: does the opening session know what it is doing and why it is doing it?

I’m popping out of my notes-from-the-past to fill in some context, because I suspect some readers are quite interested in the object level detail. CFAR-esque opening sessions are long—something like 3-4 hours of context-setting and getting everyone on the same page, usually starting around 6PM on the evening that everyone arrives.

I am completely sold on them, both in overall purpose and in specific structure, but there’s a lot to unpack there. Here’s a pair of opening session schedules, presented side-by-side:


What’s this doing?

Expectation management. Expectation management. Expectation management.

It’s expectation management at every level of zoom, from the molecule to the cell to the organ to the organism to the ecosystem to the continent. Right from the start, I want people to know what’s about to happen to them—to know what they are doing, to know why they are doing it. I want them to be in a frame where they expect to ask this question, and they expect to know the answer, where that question is right up front.

It starts with a big ol’ schedule that’s right in front of their faces. “What’s this? Oh—it’s the opening session schedule.” It starts with the word “Welcome!” written on that schedule, and the first thing that I do once the group is gathered is, in fact, to explicitly welcome everyone to the workshop, and then put a check mark next to the word on the page.

The general arc of opening sessions:

  • Explain what the workshop is like. What its purpose is, what its history is (in brief), what people can generally expect is about to happen to them.

  • Explain why the workshop. Explain why there will be e.g. various kinds of sessions. Explain why lunch is 80 minutes long. Explain why there are 20+ people all crammed together in this little space.

  • Explain the outcome of the workshop. What are the things you want them to walk away with? How will they be able to judge whether or not the thing has succeeded (to the extent that they buy the frame you’re laying out right now)?

  • Get people into their own heads. Get them to brainstorm their own hopes and dreams and wants and needs and frustrations and bottlenecks and opportunities.

  • Get people back out of their heads, and noticing the other cool people in the room around them. Sow the seeds of future connection in conversation over the next few days, because if they don’t talk to each other they’ll miss out on something like 40% of the value.

  • At some point, give people a snippet or a taste of what the heck this whole applied rationality dealio actually looks like, in practice. Show them a microtechnique, or run them through a toy exercise, so they can feel the kinds of mental shifts the workshop will attempt to model and produce.

  • Take a break, because by this point people will be tired. Call them back and talk logistics, so they understand that there’s a schedule, that there are ways to get a nicer pillow or a fresh towel, so they know the norms about skipping sessions (namely: do it, if it seems like the right thing to do; no one will second-guess you).

  • Give people some tips and advice, usually drawn from the list of thirty-or-so potential fragments of tips and advice that past workshops led CFAR and others to develop. Explain to them the most likely failure modes, and how to recognize and dodge them; explain to them the things they can do to maximize their chances of getting something good out of the experience.

  • If at all possible, if the group still has stamina, give them a full-on muscle-flexing exercise, usually in some kind of introspection, softening the soil for the Day One activities.

This is long. This is hard. This is arduous. This is (experience tells me) absolutely necessary. Without it, the workshop takes a full extra day to cohere, if it ever coheres at all; without it, you’re fighting misconceptions and preconceptions the whole way; without it, you can’t actually rely on informed consent, because you didn’t offer the information.

Do you know why opening session?

Because without it, you won’t have a workshop except by luck.


Day Zero cont’d:

Okay, so I’ve got some Theory appearing—

Oh wait! Eli Tyre sent me some Wishes! They’re private, but I should reread them real quick and see—

Okay, a thing I can say non-privately is that Eli is wishing for a lot of interpersonal epistemic stuff to come out of this. People taking seriously the possibility that other people’s minds and cruxes make sense, at least on some level. People being able to boot up any curiosity at all about what others are thinking rather than leaping straight to dismissiveness. People believing that they can understand the reasons others do the things they do.

Eli’s also wishing for something like “improved form,” such that the people he sees trying really hard can recoup some of their wasted effort/​motion such that they can continue to put forth just as much effort and yet see much better results.

These seem like good wishes. I don’t know if they influence my opening session all that much.

Okay, so back to Opening Session. What have we got in the skeleton so far?

  • Welcome

  • The Flag and the Pledge (as a concept)

  • What this workshop is trying to be (in specific)

    • A shakeup (if your current ways of being were sufficient to solve all of your problems and achieve all of your goals, etc.)

    • A source of useful tools (the techniques)

    • An actual practice ground/​place for you to actually accomplish something

Oh, I noticed that I actually want to talk about the flag and the pledge, but I’m not in fact interested in Securing A Pledge from them, à la “you gotta pledge this to enter the class.” This is in no small part because they’re going to be already there, and so it’s not Ethical to then demand a specific and unexpected commitment.

Rather, I think I’m going to introduce the flag and pledge model and then just talk a lot, explicitly, about what I, Duncan, will assume is true of all participants at all times, such that if it’s not true of you, you should feel free to take a break or like continue to be present but know that I might be misunderstanding you or whatever, this is poorly worded, but anyway I’m moving on.

  • What I hope people will walk away with (this is just the usual set-your-own-victory-condition-and-then-fulfill-it thing)

    • Tools

    • Concepts/​lenses/​a changed and hopefully richer way of seeing and thinking about the world

    • Actual progress on at least one of the things which matter to them, where “progress” includes deconfusion without any further motion

    • Experiences that are themselves valuable

    • Plans for the future that are Better than the plans they would’ve elsewise made

Interesting. Now feels like the point where I would “usually” do Wason 4-card, but it doesn’t feel appropriate in this moment. What feels appropriate in this moment, actually, is Bugs List & Feature Requests.

Why?

Because I’ve been talking and talking and talking, for one, and also because I just talked a lot of game about being them-centric and I’m reminded of how Ashe and Ruby keep saying that their dance lessons are connection-centric but never actually do connection exercises.

Okay, so bugs list and feature requests. I think I can do that stand-alone and then move a little bit into icebreaker via think-pair-share. Now feels like a break (we’re probably about 80min in, fuck).

BREAK.

Okay, what’s next on the agenda? We bring them all back, and talk logistics for maybe just like seven minutes, and go over start and end times and classes and expectations and so forth. I think I need to be a little clearer with this group about the tradeoffs of sleep and socializing and so forth, but I think right this second isn’t the moment to dive into that, but rather just to herald/​headline it.

[Note from the future: the detail here is “some people get no sleep at their workshop; others meticulously defend their sleep. Both groups have people who regret their choices; both groups have people who endorse their choices. Make your own decision, but err on the side of exploring/​trying things/​taking advantage of precious opportunities.)

Once logistics are dealt with, I think it’s time for tips/​advice/​requests.

I brought the LEGO ship, for Eat the Instructions, which should either consume or come after Try Things.

I definitely need Be Present, and I probably need Build Form—hang on, what did I do last time? Okay, with the Circlers I did Be Present, with a little headline about increasing marginal returns from removing distractions. Then I did Adjust Your Seat, then Build Form, then Eat the Instructions, gosh, I was really scared about the Circlers hurting themselves by taking my instructions too seriously, or maybe I wanted to signal that I wasn’t, like, the Man, maaaaan.

What do I want + what do I think these people need?

I think these people need Try Things. I think they are a group that matches in my head to the stereotype of people who will get Very Excited About The Concept Of Pushups but not actually ever do any pushups.

I think they need Be Present, and I think I want them to have a particular flavor of Be Present that makes them more like Circlers. I want them to see each other as fascinating objects and sources of study, and not just sit in their room the whole time. So Be Present will have two separate parts.

I want Eat the Instructions because it’s just good practice.

I want Build Form.

Do I want Boggle? Or actually the less horrible form of Boggle which I think I’m calling Gloss?

… probably. I don’t need to decide this right now. Four tips vs. five is not a super huge difference.

Oh, wait, actually, I want to go straight into SPCK so yes, I do want Gloss. I hope Robin likes Gloss.

Okay, so this is what I’ll write on the flipchart:

Opening Session Schedule

  • Welcome

  • The Flag and the Pledge

  • What is an “applied rationality workshop”?

  • Duncan’s Hoped-For Takeaways

  • Bugs Lists & Feature Requests

Okay, my shoulder Logan just frowned at me, so I think I will do Bugs Lists & Feature Requests and then subvert it, there, ha, do I get a cookie now?

  • Logistics

  • Advice from Future You (approx)

  • Something that isn’t SPCK exactly so it won’t be called SPCK exactly

  • 3...2...1...

… that’s a flip chart, right there. And yes, that’s somewhere between 2.5 and 3.5h, and yes, I think I will have accomplished the implicit goal of Getting Everyone’s Expectations Properly Set. Congration, you done it.


Finalizing Day One (these notes written on the night of Day Zero)

DAY ONE (The basics)

I think that I will not respond particularly much to any unique info or intuitions for Day One; there’s a pretty solid corpus of evidence about “you just really really really need all the people to learn these five-or-so basic phrases right off the bat.”

[Note from the future: Do You Know Why You Are Teaching {TAPs, Goal Factoring, and Inner Sim}? Answer: because if you don’t, the workshop falls apart. These three frames/​perspectives/​lenses are the foundation for all of the other concepts, and missing any of the three is A Big Deal for people’s ability to keep up.]

So: Things which plausibly “need” to happen by at least the end of day 1.5 (and thus are candidates for Day 1):

  • A grounding in commensurability and reductionism (things are made of parts, parts may be exchanged, 8020, VOI, not doing the math makes it worse, reinvesting your gains)

  • The ability to access their inner sim, separate from their “author-brain imagination” (Inner Sim + Murphyjitsu)

  • The concept of a trigger-action pattern and at least some initial buy-in for “this might actually describe my brain,” plus a bit of awareness of limited agency as a thing they need to budget

  • Factoring, which a little bit relies on things like Focusing and button-tests and other introspective tools but maybe Oh Well

  • Focusing, both because it’s an important tool and also to set the tone early on that we’re doing both cold math and squishy goo

… okay, I notice that I’m starting to feel tired/​overwhelm, and I suspect I shouldn’t really think super much harder on this. Like, I suspect it is actually the Right Call, at this point, to just go with something simple and standard and not exert a lot of cleverness or optimization pressure.

Here’s what I told Akash weeks ago was a rough approximation of the day:

9:00.......Morning regather + Units of Exchange
9:40.......Your Inner Simulator
11:00......Trigger-Action Patterns
12:00......Lunch
13:20......Intro to Reductionism
14:40......Pair Debugging
17:20......Day 1 reorient
18:00......Dinner
19:00......Courage circles

[Note from the future: do you know why there are 20-minute breaks roughly every hour? Because people need time to do More Than One Thing during the breaks; because the sessions are high-effort and five minutes is not enough; because we want to create a feeling of expansiveness and abundance, not rush; because we want people to actually talk with each other to digest the previous session, and not leave disagreements and confusions unvoiced.]

Here’s what I think I should write on the flip chart (tonight; always put up the next day’s schedule the night before even if you’re going to shift it around a little, because people need something to orient on):

9:00.....Morning Regather + Units of Exchange
9:40.....Murphyjitsu: an Inner Sim algorithm
11:00...Trigger-Action Patterns
12:00...Lunch

(gosh I wish I hadn’t set the expectation that lunch would be early, I really wanted to shift it back an hour just now, but eh, for many people this’ll be their first meal so early seems fine, also it seems really Bad to mess with the catering this late in the game after telling Akash that lunch should be at noon weeks ago, I might spring something like that on a CFAR team that’s got like fifteen people to handle the headaches but not here)

13:20...Factoring

(goshhhhhh, I really really just tried to talk myself into doing Focusing, then Factoring, and bumping pair debugs to day 2, there’s so much content, I want to put it all in their brains)

14:40...Pair Debugging (30min intro takes us to 15:10, first round 20+20+5=50 takes us to 16:00, 10min to handle repairing + 20+20+5 takes us to 17:00)

17:20...Day 1 reorient (which will largely be a debrief from pair debugs plus a heads-up/​orientation to Day 2, whose schedule will need to be written by then)

18:00...Dinner

19:00...Courage circles

God damn it is Suspicious that this is exactly the schedule I came up with on the fly two months ago but also having worked my way back through it, it does seem to be the right foundation for the remaining three days.


Day Zero: Debrief

I stuttered and stumbled pretty badly in the first fifteen minutes. I literally think none of them noticed, because they have nothing to compare it to except other workshops, which will usually be led by people with less stage presence. But I basically bungled Welcome and I said The Flag and the Pledge way too quickly and generally and without any concrete examples, and I almost literally said zero words about “What is an applied rationality workshop?”

But I found my feet with “five things I want you to take away,” and then after that we went smoothly into bugs list and feature requests and it seemed to me that people were … impressed by their own ability to produce content? After multiple prompts?

Also they asked me to write up the prompts and may have convinced me to make a Youtube video of me offering them.

And by that point several stragglers had arrived, so I had a Plausible Excuse to go back and revisit the “what is a workshop” question [which I did because I could tell that What I Was Trying To Do wouldn’t work without it] and on the second time around we did Wason 4-card and I said much more coherent things about finding ways to shift state, and new ways to think and move our minds, and cross-applying tools and algorithms and processes, to get new angles of attack on thorny problems and generally see and think better and more effectively. So that was nice.

Then tips and advice went … medium-plus well? I liked the fact that they asked probing and tricky questions about the tips, including about the history of the tips, and e.g. “why ‘eat the instructions’ instead of ‘adjust your seat’?” and so forth. It let us establish more of a rapport-of-equals and shook us out of the teacher-pupil frame a bit, in a way that was nice. But it also did rob the tips of a little of their crispness, mostly via the actual minutes being spent on stuff other than merely hammering home the ideas.

And my last tip, which is “notice gloss,” did not properly come through (no surprise, since it’s only the second time I’ve delivered it and I did not find space to practice or build up a suite of examples). I may find ten minutes in the morning to brainstorm a set of five examples to hand the group.

Then we spent the final hour on what I called “Immediate Observation” (same title I gave it with the Circlers). It was, in brief:

  • Logan’s How To Think Of Things exercise, with a framing prompt of “we’re going to spend the next hour knocking the rust off our ability to observe what our minds are doing, moment to moment.”

    • Write down as many buildings as you can

    • Write down things that slowed you down as you attempted to do this

    • Write down things that sped you up as you attempted to do this

    • Write down “how I’ll go faster”

    • Write down as many modes of transportation as you can

  • Discussion of what sorts of things our brains were doing

  • Mental challenges partner exercises

    • Break up into pairs: a solver and a squinter

    • Solver tries to multiply 27*83 while sometimes verbalizing what’s going on in their head

    • Squinter squints, eventually begins asking questions and offering observations and theories

    • Switch roles; solver tries to turn CHIN into AREA via one-letter transformations that always make a word (CHIN > COIN > COIL)

    • Squinter squints, eventually begins asking questions, etc.

  • We talked about what sorts of objects and processes and other phenomena we’re currently able to observe in our minds (they came up with a really good list)

… according to me, I should be having more evaluatory Thoughts and drawing more narratives and conclusions that segue into having Plans and Intentions for the group, and for tomorrow. But actually I’m just tired. I’m quite glad I finalized tomorrow’s schedule earlier in the day rather than having to do it nowish.


Oh they had lots of social anxiety and feelings-of-not-being-good-enough doing a mental challenge in front of each other. Memorable quote:

I remember thinking “oh god, I’m so bad at this skill that might be correlated with a thing that matters, maybe.”


Day Two planning

… possibly premature but I’m wanting to get a little ground under my feet before diving into the work of day one.

An incomplete list of things that might belong on Day Two:

  • Goodhart

  • Understanding Shoulds

  • Systemization

  • Focusing

  • Bucket Errors

  • CoZE

  • TDT

  • IDC

… it’s very interesting how, even when I’m pulling in all sorts of other random or new content (not necessarily in the list above, but like, in my background list of stuff-that-might-be-workshoppable) I still very much have an intuitive sense of “day two content.” Like, I’m running down my list and some things feel very much “no, day three; day two is too soon.”

[Note from the future: the arc of a CFAR workshop is very clear once you’ve been to a few of them. Day one is pretty technical and high-level, and involves a lot of content transfer and giving people epiphanies and revelations and whole new ways of conceptualizing their thoughts and behavior. Day two is much softer/​squishier, and involves using the vocabulary of day one to start drawing out people’s intuitions and emotions and desires and deeply-held values. Day three builds on that and is largely about synthesis and digestion; actually locking in some of the insights and decisions of the first two days, and then leading into What Do You Actually Care About Though in the evening, by which point the people are Real enough and Shaken enough that they’re actually able to be vulnerable both emotionally and epistemically in Hamming Circles. Then Day 4 is largely future-oriented, with a handful of last-minute content snippets thrown in.)

I think I need to see how they respond to at least UoE, TAPs, and Inner Sim before I can really claim to have a feel for the group, but if I’m going to put the day two schedule up before dinner I’ll have to make the call mid pair debugs.

Some things I might observe, that will help me make the calls:

  • Looks of hope, awe, or epiphany vs. looks of confusion, skepticism, or despair

… well, I guess that’s most of it, actually. I’m expecting them to not bounce off of the Day 1 content, but if they do, I’ll need to respond somehow.

Which I guess clarifies the plan a little bit (and I feel a little embarrassed for having missed such an obvious 101 lesson): the plan is to treat them like standard central undifferentiated workshop participants unless and until I need to do something Different so I should run with my sense of What I Want People In General To Grok On Day 2, at least in the first draft.

And in the first draft, Day 2 is about souls and guts, where Day 1 was about crystals and machines and microscopes and telescopes.

So Understanding Shoulds is a good start, followed by Focusing. Shift gears a little to Bucket Errors, then lunch, then in the afternoon look at Goodhart, CoZE, and maybe do IDC as the practical. That feels a little top-heavy or something, like it needs a smidge more contrast, but it’s enough to let me turn my focus fully back to the morning’s classes without feeling like the future is Looming.

[Note from the future: why courage circles? Because Hamming circles on Day 3 is the emotional and agentic apex of the workshop, which requires CoZE on Day 2 to prepare the groundwork of goodwill and warm feelings and willingness-to-stick-one’s-neck-out. There needs to be something on Day 1 that lays the groundwork for CoZE; something to set the rhythm of “more intimate optional stuff in the evenings,” and I want to get more data on what happens if I try to give people Logan’s courage circle prompt (even if clumsily). Other candidates included regular Circling, clicker game, and hot seat.]


C-C-C-Combo! Day 1 partial debrief and Day 2 planning cont’d

They were dead tired/​low-energy/​sad/​nonparticipatory this morning, especially by the end of the third session. Eli claims that this might be down to how I am doing less Charismatic Performance and more speaking to them as humans/​equals, e.g. the classes are less “showy” and more conversational than at a CFAR workshop (I want to say “less manipulative” but it’s not that I think manipulation is Never The Right Call).

It’s also possible it has to do with the lack of getting up and rotating to fresh contexts. They’re all in one room, there’s Just Me as the instructor, etc. They’re off at pair debugs now and I’ll be curious to see if this has any sort of positive impact on mood and energy.

I’m not quiiiiite sure what to do re: energy, in the world where they come back and are still just as lethargic. It’s the sort of thing where, at CFAR, we might trigger a massive push to have our 15 mentors each go out and one-on-one engage every participant, and then check back in. It’s plausible I should e.g. invent a brand-new session or activity to cause people to behave differently, somehow (clicker game? Stag hunt?) but I’m not sure what to target. Like, I don’t know the source of the disease, and it seems like there are a lot of ways where trying to cure the wrong thing can make things worse.

[Logan’s suggestion: What if they did something fun? Such as zombie tag.]

Fingers crossed that they come back having made some progress, I guess?

More into finalizing the schedule for Day 2. I feel uneasy. Like, I feel like my thoughts here should be conditional on some facts coming back about pair debugs, but honestly I can’t see how [different actions] on the part of the participants would really shift me from one possibility into another. The space is either too wide or too flat.

Uneasy.

9:00........Morning regather + Understanding Shoulds
9:40........Focusing
11:00......Internal Double Crux

… I dithered and went back and forth a BUNCH on those two classes. For a minute, I wanted one, then the other, then both, then I was like “both is too much,” then I was like “okay but I need both eventually,” and then I was like “day two is too soon,” and then I was like “not the way I teach them, though,” and then I was like “fuck it, we can just make it Squishy Saturday Morning.”

I think it’ll be somewhat deep, and I think there’s a decent chance some people will just … not be in the right mood, or otherwise refuse to sink in. I’m tempted to do IDC first, even though it utilizes Focusing, because it’s more scaffolded and concrete, and for some reason I suspect that they need that. But I’m not at all sure.

What’s the good thing I’m chasing? I’m thinking that they will finally get in touch with something like their deep motivations—the things they are actually moved by and fueled by; that a triple-punch of shoulds, Focusing, and IDC might snap them into … human-shape, instead of EA shape? It feels like they’re EAs rather than people, and I need to get through to the people. So says my brain.

12:00......Lunch

What’s available for the afternoon? Bucket Errors, Goodhart, and CoZE. Maybe also something like TDT or Systemization or Fermi modeling, but those feel a little bit inappropriate. Too much whiplash, or something.

CoZE for sure, said my brain, but then I was like, hmm, does CoZE have to be an actual full class? I could just do CoZE in the intro to the evening activity … but then they’d have little to no time to think and come up with experiments.

Oh, right, I also promised them executing intentions blocks starting on Day 2. That’s both a relief in that it means I don’t have to think about what thing fits in, and also a little bit “fuck” for some reason. Like I’m … grumpy about it? I think because I’m still currently not convinced they’ll do anything useful or good with ExInt. Maybe I can fix that with the intro spiel.

13:20......Bucket Errors
14:40......CoZE + ExInt Intro

15:40......Executing Intentions (pair debug, trio walk, bugs list)
17:00......Flash classes: Split-and-commit, subject-object shifts, TDT, Fermi estimation, understanding bell curves (This is not a list; this is a menu from which I need to choose at most three) (fuck)
18:00......Dinner
19:00......CoZE (Practical)

What were those ten concepts? When I did ten-concepts-in-an-hour? (checks photo from Circlers’ workshop) Okay, valid/​plausible ones include secretary problem, broccoli error, and passing ITTs.

Fuck. Grrr. I can write up a schedule, but it doesn’t feel real. Grr.


[Note from the future: why’s it say “squishy” in the upper corner? So they’ll know that the day has a theme and how the theme is supposed to feel, relative to the previous day.]

...if I take the above schedule as given, what’s that leave for Day 3?

  • Systemization

  • Goodhart

  • TDT

  • More pair debugs

  • Hamming

  • Eat Dirt

  • Red Vs White

  • Resolve Cycles

  • Stag Hunts

[Note from the future: the feel of digging through a box of LEGO, searching for the right piece that will do the thing you’re trying to do. You know it’s possible, you just don’t know precisely how (but you expect to recognize it when you see it).]


Courage circles intro seemed to go well. Roughly fifteen participants showed up for the intro spiel, and twelve went off into circles. [That’s right, these notes were written before Day One was even properly over.]


Talked with Robin about how courage circles and some of their other conversations went. Am now highly confident that the squishy Day Two content is exactly what the median participant needs. This is because they’re all running EA bot.

[And I, Duncan, think that this is a bad thing, even from the perspective of genuinely valuing impact, i.e. I think performatively or otherwise forcing yourself into EA shape makes one less effective as an EA than casting aside the shoulds, getting in touch with your actual values, and engaging in sustainable emergent behavior in touch with those values.]


Added two questions to the standard brainstorming spiel during CoZE lecture:

  • What are things you think are good or acceptable to do, but which you find uncomfortable?

  • What are things you enjoyed in the past, but don’t do anymore?

  • What are things you see other people enjoying, but never try yourself?

  • What are things people like you can’t do, or aren’t allowed to do?

  • What are things you’re curious about, but for some reason have never actually explored?

  • What are things you want to do, except you think your community more or less frowns upon them?

  • What are things that are contrary to your identity or self-image, that you nevertheless sort of want to try?

  • What are things that all humans should feel free to do, but you, personally, do not? Where are you cut off from your “birthright” as a human?

  • (Added) What are things that [specific person very unlike you] does?

  • (Added) What’s something that went really wrong the one time you tried it, so you never tried it again?

[Logan: if i were teaching cozy right after courage circles i would for sure ask “what do you feel drawn to and also afraid of?”]

[Duncan: ooooooo yoink]


Day 2 partial debrief and Day 3 initial planning

Okay, here we go, planning for tomorrow.

“It seems to me that the group is Doing Okay. They are not in the upper quintile of Awesome Workshops, but they are solidly median-plus. Like, on the low end of workshops that are unambiguously good.”

If I have concerns for them, they are mostly:

  • Not enough getting outside of their headspace (may be cured by tonight’s CoZE lab)

  • Not enough cutting the enemy (will at least be helped by Hamming, but maybe needs more than just Hamming)

  • Not enough progress on their problems/​goals/​too much just hanging with friends (probably needs to be partially addressed but also it’s fine if they don’t meet my personal bar as long as they’re doing decently in an absolute sense)

Emotionally, people seem good. There are dips into deepness and seriousness and nobody is drowning.

Technique-ally, people seem good. There’s some degree of bounce or skepticism or nonparticipation but it seems pretty Normal and mostly people are engaged and engaging with each new concept or tool.

Tonight, I’m going to give them a short version of Resolve Cycles, and then follow that up with three mini-lectures on Subject-Object Shifts, Broccoli Errors, and Eat Dirt, as the last words before dinner and then CoZE lab. I think the order of those should be … hmm … it feels to me like Broccoli Errors first, then Subject-Object Shifts, then Eat Dirt.

Why? What is it about that order that feels right?

Well, Eat Dirt is the most reassuring and … quietly energizing? It’s the one that says “do not let your lack of perfect knowledge stop you; follow your nose.” That feels like a good last word.

Meanwhile, Broccoli Errors feels straightforward and clear and simple. It’s one they’ll get right away, and probably laugh.

Subject-Object Shifts feels meatier, and so it’s more the “center” of this little triple-punch lesson. I want it to have the most time to breathe, and the most focal attention, which … yeah, there are just-so stories that argue for all three placements (“First is best because you set the tone and catch their attention!” “Last is best because it’s the thing they walk away mulling over!”). But honestly it just feels like the center is the right place for the meat. This is an intuitive thing.

Okay, because part of the whole point of this note-taking exercise is to make my opaque intuitions more explicit and visible, I’m going to do a thing I wouldn’t usually do, and double-check the claim. How do the other five possible orderings feel in my mouth, when I roll them around?

Broccoli Errors, Eat Dirt, Subject-Object Shifts. Feels bad. The way it feels bad is “eat dirt will be lost/​feel fake/​not make itself heard.”

Eat Dirt, Broccoli Errors, Subject-Object Shifts. Feels … okay-ish? Something about eat dirt followed by broccoli errors feels like a good combo, like “don’t sweat knowing which way you’re going” + “don’t let your preconceptions limit you.” But then subject-object shifts feels like a whole different muscle, or something. Like, the first two form a clear ray pointing in a direction, and the third points in a very different direction.

Eat Dirt, Subject-Object Shifts, Broccoli Errors. Feels bad, mostly because broccoli errors feels too small and trivial after the first two, which are weightier.

Subject-Object Shifts, Eat Dirt, Broccoli Errors. Same issue.

Subject-Object Shifts, Broccoli Errors, Eat Dirt. Feels okay. This is the most plausible alternative ordering. You have a big lesson about what growth and evolution looks like, then you have a concrete specific failure mode highlighted and good-natured-ly mocked, and then you have the reassurance of Eat Dirt as a final piece. I’m plausibly more interested in doing this one—

—except that we’ll be coming off the heels of Resolve Cycles. I want to have the small thing first, as a bit of a palate cleanser/​small bite that’s not big enough to choke on.

Okay, so. Resolve Cycles (because I want to give them a booster shot of energy and progress and successery) and then three concepts, and then CoZE and that’s the day.

So. Tomorrow.

I like the idea of there being something content-ful at the beginning of the day. Plausible options are basically only Systemization, or Goodhart.

Goodhart is a more important lesson imo, but Systemization lends itself better to “then they’re gonna go off and pair debug.” Like, having the “increasing marginal returns” and “permanently solve trivial inconveniences” lenses booted up right before pair debugs feels pretty good.

So we’ve got:

9:00......Systemization
10:20....Pair Debugs (10m intro, 50m round 1, 50m round 2, 10m debrief)
12:20....Lunch
1:40......

Okay. At this point, they’ve seen the core CFAR content, they’ve seen the squishy content, they’ve done pair debugging for at least four rounds. Hamming … now? Or Hamming before dinner?

Hmmm. Other things we haven’t seen much of include flash classes, lightning talks.

How many schedule slots do I actually have?

1:40 − 2:40, then a break, then 3:00 − 4:00, then a break, then 4:20 − 5:20, then a break, then … 5 … 40? No, that doesn’t make sense, there shouldn’t be a thing starting that late.

So one of the earlier things needs to be longer or shorter.

1:40 − 2:40, something.
3:00 − 4:00, something.
4:20 − 5:20, something.
5:40 − 6:00? Hamming? Very short Hamming intro straight into dinner?

That feels better. That feels okay.

Okay, so. One thing that’s missing is the tutoring wheel (something to digest/​synthesize/​lock in the content). That’ll hit them pretty hard, as a loss, though they won’t necessarily notice it themselves since they’ve got nothing to compare to.

I could do a miniaturized tutoring wheel, in which they e.g. explain the techniques and concepts back and forth to each other, or in which they work with a partner and reinterpret each of three bugs with more-than-one-lens/​technique.

Hmmm, that sounds like an eighty minute session, actually, or even ninety.

1:40 − 2:40, [thing].
3:00 − 4:20BLAZEIT Tutoring Wheel replacement
4:40 − 5:40, [thing]

Okay, still not good. I’m going to tuck a short thing like TDT in front of the Tutoring Wheel replacement. So we have:

1:40 − 2:40, [thing]
3:00 − 3:20, [short thing] then straight into 3:20 − 4:40 [wheel]
5:00 − 6:00, [thing]

… this is a decent flow. So I have to choose two Things to put in, and then I need to invent a Tutoring Wheel replacement. Wild that the latter thing feels easier than the former.

9:00......Systemization
10:20....Pair Debugs (10m intro, 50m round 1, 50m round 2, 10m debrief)
12:20....Lunch
1:40......Goodhart
3:00......TDT
3:20......Tutoring Wheel
5:00......Hamming (Right, it’s not that I had two things to choose, I just had to find one, and Goodhart was the obvious answer)
6:00......Dinner
7:00......Hamming Circles

w00t, that’s a Day 3 right there.


In A Strange Place

Day three seems to have gone mostly-well. I got like four hours of sleep because I was up late (endorsedly! I knew what I was doing and why I was doing it =P).

The same seemed true of approximately everyone else; I got a vibe that the house was “too quiet” when I went downstairs at 8AM and made the decision on the spot to shift the schedule to start at 9:30. I put up signs in places where people would see them coming down the stairs (the signs deliberately had a Duncan vibe; they said e.g. “Starting at 9:30 instead of 9:00. Make breakfast! Go on a walk! Dissociate and hallucinate!”)


This rejiggering of the start time did not technically require a rejiggering of the Whole Schedule, but it raised to the surface some dissatisfactions I had (apparently) been feeling, plus gave me a chance to act on some subconscious mulling I had done. The new schedule became, thus:

… they’re off at Hamming Circles right now; I’m on lifeguard duty in case anyone gets in over their heads (and because I probably don’t have the unmasking capacity to do a Hamming Circle for real right now as a real person).

Systemization was mainly presented as, like, “a set of concerns I want to prime, before you go into the pair debugs” as opposed to being a full technique or perspective all its own. I started off by noting three “motivating insights,” i.e. why you should care at all about systemizing things. Those were:

  • Increasing marginal returns. The more you can remove distractions and drains on your attention, the more valuable each removal is; don’t be fooled by how low-value removing the first distraction is, and thus never actually clear out all the clutter.

  • Repeated costs add up. Be suspicious of things which “just” take 10sec of your time or 1/​10th of a spoon every hour or every day; very seriously consider that it’s worth a large up-front effort to plug a leak.

  • Freedom to reinvest gains. Remember that people often confuse “you should systemize your email” with “you should spend less total effort on email.” It may be that the latter is true, but even if the latter is false, systemizing the email can still be a good idea, since it frees you up to reinvest time and attention and energy right back into the emails.

All of these were concepts we’d previously touched on; I was just booting them back up.

Next, I talked about Malo’s task system, which is (in brief):

  • What if we just don’t do this?

  • Okay, what if we make someone else do this?

  • Okay, how can we solve this once, systematically, such that it doesn’t come up ever again?

  • Okay, fine, I’ll put it on my own to-do list, as a one-off.

Then I clarified that the goal of a systemization push is to reduce attentional drains. To cause yourself to not have to think about X often, cause yourself to not have to think about X hard, to set things up so that the thing you want to have happen just happens, automatically/​effortlessly, on the path of least resistance.

So then it was off to pair debugs.

Goodhart’s Imperius was straightforward and solid-as-usual; I started it off with a quick clicker game and also had them do some brief Taste exercises.

For Locking It In, I basically gave the-people-who-showed-up-at-all (I’d previously said they could do ExInt instead) a goal:

Don’t let yourself gloss over confusions about the concepts or tools, don’t become the kind of person who forgets to ever use a tool they genuinely expected would be helpful

… and then gave them a couple of tools for how to pursue that goal, along with encouragement to invent their own. The two suggestions I had were “explain each skill or concept in brief back and forth with a partner,” and “make a grid of five bugs and five concepts/​tools, and spend 1 minute on each cell in the grid confirming that either a) you do, in fact, know how to use this lens to start working on the problem, b) this lens is inappropriate for this problem, or c) you cannot definitively answer a or b and need to ask questions.”

Then Red vs. White was really good and useful, according to me, and in particular impressed/​interested our Most Skeptical And Aloof Participant, among others. I posed it largely as “guys, I dunno what the answer is here, but it really seems like both of the common perspectives are Insane and also it seems like this is relevant to e.g. how to build a functioning AI that doesn’t ever decide to kill all the humans.”

Then I gave Hamming questions in brief (30min) and the send-off to Hamming Circles was fairly normal.

...

...

I notice that I myself am something like Sad and Lonesome. I think I got more human connection in the small group with the Circlers, and thus was not doing an … alone aloof on a mountaintop thing?

Robin has been great, when I’ve managed to drop my mask enough to actually hang with them (such as last night). But man, I guess if I imagine doing ten of these workshops in precisely this fashion I expect to get really super duper depressed. Like, I am in a very serious sense All Alone Out Here, and I didn’t notice it accumulating until tonight. The camaraderie of morning scheming meetings and occasionally ditching responsibilities to hang out and shoot the breeze with staff in the staff lounge at actual CFAR workshops were apparently much more important than I noted.

(Also probably feeling weird because I had a near-interaction with **** where we were both commenting on the same ** thread. This isn’t crucial but it’s certainly relevant.)

Anyways, overall I think things are really good. I think this is a 60th-percentile (or better) workshop, among the set of all workshops I’ve been to. The participants are alive and awake and engaged and taking things seriously, and none of their unique weirdnesses as a subgroup proved destructive or fatal. I could certainly do better on a second round with a similar crowd (in particular I think I might target the weird cargo-culting of EA memes), but they’re happy and I’m happy (re: the goals of the workshop).

I think that “this is as good as the 60th percentile thing we accomplished with 4+ instructors and 10+ mentors” is an achievement to be reasonably proud of, and I do think that’s the actual true state of what’s happening here.

No idea what tomorrow holds. Maybe a second email. Sad, though.


Day 4 Planning

Okay. Day 4 will be planned “bottom up,” which is a different way from most days.

Essentially, it’s going to be thoroughly defined by its constraints, starting with the most crucial one: we have somewhere between 2 and 4 hours of content-space, period.

A possible schedule:

10:00......Thing 1
11:20......Thing 2
12:20......Quick Lunch
12:50......Thing 3
2:10........Closing session

When I look at my list of hopes and dreams for the workshop, there’s not much left on here that hasn’t at least been gestured at. The “Tim Stack,” a more direct gesture at TDT/​policy-level decisionmaking (somewhat redundant, most of the message has been conveyed, but would be nice to crispify it), Fermi estimation, actual double cruxing, Split-and-Commit, Moloch.

And, of course, The Actual Point, which is “is your damn life gonna be any better?” I think, as mentioned above, that we’re doing great in a counting up sense. But counting down from the actual goal, I am skeptical as always.

Eli has volunteered to lead a session on training regimes, which I think should happen right before closing session. So:

10:00......Thing 1
11:20......Thing 2
12:20......Quick Lunch
12:50......Training Regimes
2:10........Closing session

I’m tempted to do something like:

10:00......Three concepts in brief (split-and-commit, digging in your heels, stag hunts, Moloch, TDT, confusing 50% on A or B with 100% a-mixture-of-A-and-B, how to think about labels, okay, that’s more than three, maybe instead of being flash classes they’re lightning talks, which for the uninitiated means “instead of 20min, they’re 3.5min”)

11:20......Final pair debug
12:20......Quick Lunch
12:50......Training Regimes
2:10........Closing session

… you know what, that’s good enough.


A mostly-complete list of named concepts touched on during the 4.5 days:


Closing Session [notes from the future]

I did not, as it happened, send off another email to my group of Listeners on Day 4; I was sufficiently exhausted that I barely remained conscious in the passenger seat on the drive back to Nevada City.

However, closing session also has a purpose, and should Know What It’s Doing, and Why It’s Doing It, which I can sketch in brief after-the-fact.

Goals:

  • Lightning talks! This allows participants a chance to share what’s been burning, helps signal “you’re graduates now; you too can Say Stuff To The Group,” and (if I’m being brutally honest) helps people appreciate, by contrast, the coherence and polish of the sessions led by me.

  • Give everyone a moment to celebrate—the hard part is over, and now all they have to do is be Awesome every day for the rest of their lives. =P

  • Form participants into trios for follow-up (pairs fall apart too easily because it’s too easy to form common knowledge that you’re both tired and would rather cancel; trios are more robust while still being small enough that actual coordination is possible).

  • Do a survey/​feedback form, because if you don’t do it right then your response rate plummets to around 40%.

  • Do actions and insights, with the motivating spiel “so, a lot of the magic of this workshop happened privately in your head, or in some one-on-one conversation that no one else got a chance to experience. Here’s where we’re going to put all of that awesomeness into a shared pool for everyone to partake.” (This both gives people a chance to learn from each other and also is a nice moment to remind people to digest/​crystallize/​unforget the stuff they thought was important, and notice how much ground they covered in the four days. Having just gone through the survey and e.g. seeing the names of every session listed helps jog memory for this part.)

  • Do appreciations, in which we request that people be specific and concrete about exactly what it was that happened and why it was good, rather than just generically thanking each other for being awesome. (This both makes the appreciations more heartfelt and also more useful/​actually allows people to boost the specific behavior that was good and separates signal from noise, and also signals a bit of our nerdiness on our way out the door.)

  • 3 … 2 … 1 … progress! (Oh, I guess I haven’t mentioned this the whole time, but we shout “progress!” at the end of most sessions, and “charge!” at the beginning of some, and that provides a ready-made all-together Loud Sound Full Of Meaning with which to end the workshop; if we hadn’t seeded that throughout the four days it wouldn’t work nearly as well at the end.)


Here, as I reach the end of my notes, I find myself stretching to have some sort of tying-it-together paragraph that pulls the essay together and helps solidify it in the reader’s mind as coherent.

I think this is standard practice for a reason! I think that, most of the time, listening to this impulse is correct.

But in fact I am tired, and you have just slogged through Many Words, and I think we can just call it. Questions and comments genuinely welcome below, especially if you feel like you almost know how to run your own event, but have one or two places where you’re not quiiiiiite sure...

Also happy to answer questions like “okay but what are ‘broccoli errors’ and what is a ‘subject-object shift’?”

My response time may be >36h, tho. Sorry about that.