8 Questions for the Future of Inkhaven
If you don’t know, Inkhaven is a residency where you come and publish a blogpost every day. No “Oh it would be nice to blog some day” or “Oh I’m working on something, I’m sure I’ll publish it some day”. No, you have to publish today, otherwise you are asked to leave.
After 10 days into the first ever Inkhaven cohort, here are some questions I have.
1. Is daily the right cadence?
Everyone has been posting daily. There are many amazing things about this. Everyone is actually creating, there is nobody here who talks the talk but doesn’t walk the walk.
The posts are actually getting people’s souls into them. I don’t claim every post is someone’s best work, but I like most posts I read and I think the person is really trying to do something worthwhile.
However, there is a wide variety in how proud people feel about their writing.
I also asked people how satisfied they are with what they’re doing, versus those who want to make a drastic change
That’s 5 people at 3⁄10, toward the end of “1 = I need to make a drastic change to what I’m doing”.
Several fiction-writers have told me that they need a few days to work on a story, which is why they’re writing non-fiction here. And it is often the case that people cut off ideas early in order to publish, rather than letting them expand and develop, or do more research on them.
However, there are real skills being practiced here, and good blogposts being written. People are learning the skill of having an idea and shipping it same-day. I am confident that a lot of these people will take a blogging skill with them.
Nonetheless, I think that some version of this commitment can be extended to other time-scales.
What might a ‘Weekhaven’ look like?
Each week, you write a single, 3,500+ word effortpost
Writers could spend whole days reading and doing research, without writing.
It would have more narrative arc to the week. Currently an idea is brought up today, discussed, and published same day. In Weekhaven, they’d get feedback on it multiple times at different levels of development.
Editing would be more of a process; you could take a finished essay and restructure it, or re-write whole sections.
2. How do I better connect residents with established writers?
I invited a lot of the established writers here to give advice but with relatively little structure.
This is partly overdetermined by the general framing of the event, which is that nothing is mandatory except daily publishing.
In contrast, during the 6-week prestige Clarion Fiction workshop, you spend 9am to 12 noon in feedback circles, discussing the fiction by your peers, every day. And one of the established writers leads it and asks the questions.
I think there are problems with that way of doing things, but it does ensure that everyone actually finds out how their piece lands with other people and absorbs taste and feedback from the established writer.
I can’t do this at Inkhaven! I sorted everyone into 5-person feedback circles on Day 2 and ~half didn’t go. There’s a daily sign-up, yet most days not a single one happens. I think people are wrong about how much feedback they should be having, but I cannot make them do the thing. The lack of mandated activities partly means I’m now left doing environmental optimization. If I put talks outside in the central courtyard straight after lunch, or the office hours on the central deck, then people are more likely to try it out by listening in for a while.
Also, learning from the advisors isn’t primary. The writers can share tricks of the trade, but blogging is typically a very personal expression of one’s soul and should not primarily be imitated. The contributing writers are there to help hone the expression of the soul.
That said, I’d say that for essentially all the established writers, they are hanging around and having casual chats more than half of the time.fa
Ways I could go:
This is fine, all the writers I brought here are enjoying themselves.
Their time isn’t being well used, we should have them around way less.
I should find a clever way to cause them to read the residents’ writing more and for the residents to get more feedback / advice / input.
3. Should I be getting people to relax & do more downtime things?
Everyone is writing until really late, then the ratchet starts again the next day.
From yesterday’s post, here’s the graph of what hour people are publishing.
This means that people are basically busy all evening, and aren’t doing fun things together. There’s almost no board game nights or reliable evening events, because most people can’t reliably be there, and they’re stressing out until ~midnight.
Now, you might think “Why not bring the deadline back to like 8pm, so that people have the night off?”. But that’s kind of antithetical to what Inkhaven is here to offer. Should I just cut out 4 hours of their day where they can’t write? I assure you, they wouldn’t spend the late hours precociously working on tomorrow’s stuff. They selected into being the kind of people who needed an externally imposed deadline to get stuff done. They’d just be losing a good chunk of writing each day.
I just chatted with a visiting artist, and their take was not to try to incentivize this not happening, just to make a better ritual of it. Have people cheer and applaud at the moment of hitting publication. I like that idea, might try more of it.
Anyway, I think at a minimum, I need to help people stay embodied and get enough sleep. We’re bringing in a masseuse and I’ll think about some exercise stuff to do.
4. How do I help readers find signal in the noise?
We’re going to produce about 41 * 30 = 1,230 blogposts over the month of November.
I think they’re generally pretty good. Interesting ideas, some nice choice of phrasing, real soul in a lot of them.
Now, most people shouldn’t just be doing a random sampling, they should be sent to some posts that give them something they’re looking for.
However, the naive way, which is to just list some of the “best” ones, puts some things on an axis of goodness, which isn’t what I want to happen here. The things happening are more incommensurable than that.
One person I know writes extremely funny semi-fictional freudian psychoanalyses of himself. One person writes practical advice about statistics. One person has written C.S. Lewis fan-fiction of the Screwtape Letters. One person has done a replication of a very popular psychology study (that did not replicate!). One person is writing a sequence of analysis of how language works on a group level, how it constrains and coordinates action.
These are the more successful things, but many people are trying experimental things that are niche or strange. There’s not a single axis of what counts as ‘good’ here, of ‘quality’. Any way of highlighting things must steer clear of that.
I think the best ideas for now are for me to (a) encourage established writers to link to some pieces that they resonate with from their blogs, and (b) highlight things in different genres on the website. But I am not sure how to do this without having read everything (and having the established writers read everything).
I suspect there’s some way to get the residents to do a substantial amount of reading and highlighting, but I haven’t come up with a proposal I like yet.
5. What kind of podcast or video footage should I record here?
I have a lot of high-quality recording equipment. Here’s an example where Patio11 recorded a podcast here with my boss.
I have a lot of interesting people about. Should I do interviews? Have them read their blogposts and discuss them? Make an Inkhaven YouTube channel? Make a documentary?
We’ve got a few interesting interviews on a theme, where Adam Mastroianni interviews people like Aella and Slime Mold Time Mold and GeneSmith about where the ideas for their blogposts come from. I’ll probably find a way to release those.
6. Can I make more money with Inkhaven?
So far I am making some money from the paying residents (many have needed financial support to be able to be part of this program), and from sponsorships (thank you WordPress.com for your generous support of the program!). Nonetheless it’s still in the red at the current time, by somewhere in the $50k–$100k range.
This is substantially due to the opportunity cost of not running lots of weekend conferences at Lighthaven in November.
I think the obvious thing is to try harder to figure out whether we can take conferences in that period. We’re currently accepting one, we’ll see how annoying it is to move all the residents out of their rooms, but of course there’s definitely prices that make it worth it.
Natural options:
Charge more
Sell more resident spots
Get more sponsors
Get more weekend conferences
Have fewer staff (currently 5 full-time coaches and one full-time ops person)
My plan is to work on 3 & 4. I also suspect that after doing it once, it’ll be a more known quantity, and demand will go up next time (2). 1 and 5 seem bad.
Of course, if none of these work out, then it’s up to our broad support base of donors to decide if this thing is worth supporting in the non-profit sector.
7. How long should Inkhaven run?
We did the month of November for two reasons:
It is the same as NaNoWriMo, which was an inspiration
We could make it work with the calendar of Lighthaven
Here are some reasons to change:
Not everyone has the same month free. Perhaps if we let people pick a single month in the summer, they can pick the time that works for them.
Perhaps people should do different lengths? 3 weeks of daily publishing, then 3 weeks of weekly publishing, then 3 weeks working on a single essay.
The relevant variables are how much value is there from starting simultaneously; and whether people will burn out from doing it for more than a month.
I will know more about burnout after 30 days; I’m writing on day 12.
8. Should I have more mandatory stuff? More structure?
These are the sorts of people who have self-selected into a program that puts strong external demands on them in order to produce creative/intellectual outputs. Perhaps there are other things that would be worth me mandating?
Getting more painful & scary feedback from their peers
Getting more painful & scary feedback from established writers
Doing an editing class
I have a sense that there may be something else in this space, but I’m not convinced of anything in-particular currently.
I’m not observing the residents very closely, but I tentatively roll to disbelieve that most of the residents who are publishing last-minute are making full use of the day to write. My guess is that having e.g. a 10 pm deadline wouldn’t reduce “active writing time” by anything like 2 hours for most residents; they would simply “get down to business” earlier in the day. Someone mentioned that this would substantially reduce the amount of time that residents have to integrate feedback from the feedback circles, which happen right before dinner. That seems true and somewhat difficult to avoid given the current structure. I still think that forcing the last-minute writing to happen two hours earlier has a lot of benefits: residents are less tired when they’re doing their last minute writing & editing, and they have more time to socialize, unwind, and spend some time doing less “pressured” research/writing/ideating/etc.
Yeah, I agree. An earlier deadline would cut out procrastination time, or more accurately, it would just make people write more efficiently per unit time. I’d honestly be pretty excited to try having a deadline even as early as 6pm.
Seconding this, I try to get my posts out by 6pm so my homies on the east coast can read them, but sometimes I faff about. I think it would be really good if the deadline was 9 or 10pm, and then I feel like I can chat to people, do more slow collaborative pieces with others, and talk about their posts with them. I feel like this might help with, like, idea generation and stuff as well? If there’s a few hours in the day where you can talk about your posts and the posts others want to write.
Quick Thoughts:
Daily pace is excellent. I get a lot of looks at what I want my writing to look like. I had some very concrete ideas coming into Inkhaven about how I expected to improve, and I’ve already overshot those goals. I expect Weekhaven to be weaker growth stimulus overall.
There should be one day off per week. This would allow people to rest, plan more ambitious & fun events, go on hikes and actually decompress, etc. It would also give residents doing technical work the chance to focus on research rather than writing.
More mandatory, scary feedback from peers and residents sounds good. I’m not making myself get feedback. This is low-agency behavior on my part. At the end of the day, it’s up to the residents themselves to get the most out of the experience. That being said, yes limiting the action space of the residents a bit here would probably help.
Even though I have heavily underutilized them, I’ve gotten a lot of value out of the contributing writers. I overheard a snippet of feedback Andy Matuschak was giving someone else that I’ve been thinking about for days now.
I’ve read 70% of the submissions, and median writing quality is much higher than I expected coming in. There are definitely posts that seem both low insight and low effort however.
I feel pretty well socialized. I don’t expect to form lasting friendships, but many people have already left very deep impressions on me.
Maybe scheduling two, two-week Inkhavens a year will reduce the opportunity cost, and allow you to iterate & filter participants more.
Thanks, v helpful!
Re (2): I am not sure about the one-day-off-per-week. I think it’s healthy; also I’m not sure that, looking back in a year or two, whether most residents will think “I wish I took it all a bit more measured” or “I’m glad I went all out” that month.
Re (3): Perhaps next Inkhaven, the mandatory things will be:
Publish 500 words every day.
Share your writing with a minimum of one feedback circle per week, and with one advisor per week.
I think some alternate form of (2) could be interesting, where instead of the requirement being 500 words/day it’s 1000 words/2 days or 1500 words/3 days. I feel kind of hampered by daily posting because there’s a few longer posts I’m working on and what I want are some 8 hour deep work sessions on them and instead I’m doing them in these terrible 2-3 hour bits.
A related question would be, what would be the right number of great posts? The kind that might become shorthand or establish a new idea, or be quoted years from now.
I would say that I expect something like 1 in 100 posts to be great. (Imagine a blogger who writes 1 post a week for 2 years. Wouldn’t you expect 1 or 2 really awesome posts? So then that’s a 1% rate.) Then that would imply ~12 great posts from Inkhaven.
A related question is whether we can try to estimate whether Inkhaven helped. Perhaps we could go back over the edge cases in admission, which prompted some debate and were not clear accept/rejects, and pre-register their names now, before Inkhaven is over, and then have someone blinded look over their writing trajectories or something?
As a reader, I wish there was more filtering or signal boosting to help bring some Inkhaven posts to my attention.
There are a few ways that could happen. It could be something reddit-like where there’s a centralized place which at least has links to all the Inkhaven posts and people can upvote them. It could be something like LW curation where some moderators pick a few posts to curate (possibly some of them could even be cross-posted and curated on LW). It could be a linkpost style thing (as Vaniver has been done some of) where people post links to some of their favorite Inkhaven posts.
I could imagine setting up Inkhaven with the intention of having the residents do linkposts. Maybe each Sunday is linkpost day when residents are encouraged to make their daily post a linkpost (with no word requirement) where they link to 1-3 of their best posts from the past week, 3-10 other Inkhaven posts from the past week that they liked, and optionally a few things from elsewhere. Then on Monday there could be a centralized roundup post which links to all of those linkpost and all the posts which got multiple recommendations in those linkposts.
Yeah, making a subreddit with links to Inkhaven posts is relatively low effort, and it would filter the high-quality content, creating a good reference for Inkhaven.
Do you have data on how valuable writers find the feedback they get from peers or mentors? Do people who get a little tend to want more?
The average ratings out of 10 for how helpful are:
Fellow residents: 7.4
Your assigned coach: 7.4
Contributing writers: 7.0
My impression is 7 is neutral, so this isn’t very good. I’m surprised and view this as a huge lost opportunity.
I don’t think 7 is neutral. But also I predict it will go up by the end of the cohort; we were still ramping up interactions with contributing writers when the form was sent out. Plus Scott Aaronson arrives in half an hour to do a gauntlet of 1-1s, Zvi Mowshowitz on Monday for similar, and Dynomight arrived after the form, so I think more value will be gained.
I agree that one post a day is unsustainable, unless you are either a miracle writer, or satisfied producing large quantities of low-quality text (which may or may not be a good business strategy) for the rest of your life.
From my perspective, it is more like an exercise to show you what’s possible. Once you spent a month writing one post per day, it will no longer feel unrealistic to commit to publishing one post per week (or two posts per week, if they are a significant source of your income, i.e. you don’t do it after a 40-hour job week).
It is also an exploration into how much your writing quality decreases when you increase the quantity. Maybe a lot, or maybe just a little. Perhaps it helps you get rid of some extra steps, which will make writing easier after you return to a more relaxed schedule.
I don’t remember the source, but I heard that when you train some ability that can be decomposed into multiple skills, it makes sense to train each skill separately, even if maximizing that specific skill temporarily makes the overall result worse. For example, if you want to learn to throw a ball into the basketball hoop, you can separately practice “throwing far enough” and “throwing at the correct angle”, and stop worrying for the moment that the balls that fly far enough are flying at a wrong angle, and the balls at the correct angle don’t fly far enough. Practice these two things separately, try to maximize each one; and later try to join them. From this perspective, you are currently training the ability to write a lot. You will integrate it with quality writing later.
Or to put it differently, your current failure mode is “writing a lot of mediocre stuff”, but your usual failure mode is “procrastinating on actually writing and publishing the post”. If you can’t overcome both failures at the same time, at least learn to overcome each one of them separately. Maybe one day it will click together.
Actually, I think it would be a very useful exercise to try to mirror each popular blogger’s style for one day. Like, spend one day trying to write an article in a way that will make everyone think that Scott Alexander wrote it. The next day, try to write like Zvi. Etc. Plus you have the advantage that if you get stuck with a question such as “okay, here is the place where they would insert some X or Y, but how would they find it?”, you can simply ask them.
People are confused about the original voice. It does not develop by carefully avoiding contamination. Actually, many “uncontaminated” writers sound quite similar to each other. Instead, the voice probably develops by trying many different things, and then keeping the subset that works for you. It is important to try more than one thing. If you only try to copy Scott Alexander, you will at best become a weaker copy of Scott. But if you copy Scott Alexander one day, and then maybe Ernest Hemingway the next day, and someone else yet another day… and you copy dozen authors like this… then afterwards, whatever mix remains, will probably seem quite original to most readers. If you want to push it further, try crazy exercises, such as avoiding adjectives, or trying to write the article backwards, etc.; again, the goal is not to keep writing like this, but to have it tried.
I think you should give yourself an artificial deadline of e.g. 5 PM, and spend the rest of the day watching others. (And maybe keep a notebook all the time, and note some ideas.)
Maybe, try each of that once? Or do it once, collect feedback, then do it the second time trying to incorporate all the feedback. (But don’t e.g. spend the entire month doing interviews, that would be a waste of time.)
You could try to achieve sINKularity—a blog with paying subscribers that covers the costs of staying at Inkhaven.
This could either be done by individual participants (each participant starting their own blog), or collectively (the Inkhaven managers starting a blog, where the participants can contribute). The advantage of the individual blog is that the participants can take their source of income with them. The advantage of the collective blog would be more content; you could even split it to multiple blogs, by the type of content, e.g. one for fiction, etc.
Maybe the organizers could help with this by printing all the posts, giving everyone a copy of each, and a red pencil. The participants then wouldn’t have to approach everyone asking for feedback individually. They could also take the annotated papers home, and read them afterwards as a reminder.
Do you have data on the distribution of lengths of the posts people are publishing?
Yep, we ask people to submit wordcount with every post that they submit.
Here’s the frequency of posts at each length.
Here it is in three simple buckets
It looks like a lot of people are easily exceeding the 500-word minimum. Maybe that was insufficiently ambitious (given that no one has dropped out yet?), and Inkhaven 2 should bump to 750 or 1000 words. (Note that LLMs are already good at giving feedback and asking questions on essays and suggesting parts to expand on, so adding another 500 words is often quite easy—as long as you are willing to do the work!)
I think most of the value is in the publishing, rather than the amount of content. I think probably the word minimum should be less. If I look through world spirit sock puppet (IMO a great blog), the majority of the posts seem to be <500 words. It’s possible that it’s just too hard to police quality, or at least effort, with shorter posts. In that case, maybe it’s worth increasing the proof of work.
I would rather post 1000 words voluntarily, but stop at 800 words if it feels like I have concluded the topic for the day.
We might recommend the participants to aim at 1000 wordy, but not put it as a hard limit.
or have a weekly minimum that’s greater than the daily minimum x7
Would maybe work if you have to write >500 new words every day + publish >3500 words every week, as two separate requirements.
Or maybe add explicit requirements for polishing things?
Working more on polishing is valuable, but I think most XP points are gained just from writing.
I think forcing people to publish is good. I think forcing them to publish 2 blog posts on a private inkhaven blog for the first two weeks would help solve some of your problems. That self selecting aspect would make me actually want to seek out what some of your folks are writing.
As it stands I don’t really want to read something someone wrote in a day who isn’t sleeping well and is hyperfocused on how they compare to the people around them.