MJD 59,326

This entry is part 15 of 16 in the series Captain's Log

I am considering adopting two rules for projects that I think are very promising for 40+ lifestyles.

  1. No new top-level projects (TLPs) (twitter thread)
  2. Ten-year commitments to projects or no deal (twitter thread)

I don’t mean practically necessary projects like doing something to earn money. I mean non-necessary life projects like writing a blog, or a maker project.

Shitposting and idle dabbling are still allowed so long as they don’t grow into new TLPs. They can grow into subprojects of existing TLPs, but even then I need to make a ten-year commitment or not do them at all. There’s a bunch of ambiguity here, around what’s a project versus the contents of one (is a new blogchain a subproject or just a thread of content for the overall blogging project?), but set that aside for now.

By top-level I mean it doesn’t fit into the logic of any other project as a sub-project. In the case of public-facing projects, it must pass what I call the domain-name test. It’s a top-level project if you buy a distinct domain name for it. No new top-level projects is a clear enough idea I think. It doesn’t require much elaboration. This means I can drop top-level projects, but not add them, which means each time one bites the dust, there’s more attention to go around for the rest. It also means I’m not allowed to buy any more domain names unless I am rebranding an existing TLP.

The second rule is the tricky one.

By 10-year commitment, I mean you must be comfortable with the idea of committing steady, continuous attention to the project for the next 10 years, but it doesn’t have to be a commitment to a specific goal or plan or structure. Or even to specific themes, activities, habits, or collaborators. Indeed, 10 years is too far out for any of those things to be stable enough to commit to.

For example, I’m comfortable committing to ribbonfarm for the next 10 years.

Will it still be a WordPress site in 2031? I don’t even know if WordPress will be around.

Will it still be primarily a longform blog, or might it be a neural-implant VR video game or something? I don’t know.

Will I still be writing/publishing primarily non-fiction here or doing something like comics? I don’t know.

So what does continuous attention mean if there is no stable vehicle of continuity like medium or theme? Is it just a vacuous commitment to a name? That I will be doing something with something named “ribbonfarm” in 2031?

No, it’s more than that. It’s a Ship-of-Theseus sort of deal. I’ll be doing something with the memory stream named “ribbonfarm.”

Continuity of attention means continuity of memory. There may be twists and turns and pivots. There may be disruptive resets of goals and plans. But it will be the same project in terms of identity-as-memory. The memory may be ported to a new medium, but it will be the same memory.

Ten years is a good period for forcing memory to be the only defining element of a commitment.

Almost nothing is likely to be a stable constant-of-evolution through ten years. Software, planning models, collaborators, attitudes towards goals: everything shifts around given 10-year horizons. The only thing you can rely on is continuity of memory. Which means laying down accurate memories is the most important thing. You don’t know what the future of a project will be like, but you know it will draw on the memories you’re laying down now.

Continuity of memory is in fact a necessary and sufficient anchor of commitment to a project. There are ways to hold many other things (plans, goals, habits, platforms) constant but break continuity of memory in a way that destroys identity. So new commitments have to be made.

People are like this. No matter what happens to you, you have a single stream of memory. Even the most radical change in your life won’t randomly graft somebody else’s memories into your consciousness, or get you thinking of your own memories as someone else’s. At least not with today’s technology. Memory tests are a good way of smoking out impersonators too, when you don’t have photographs or fingerprints to go by.

That idea, but for projects. Rule 2 is a way of consciously anthropomorphizing projects and breathing artificial life into them.

That is why the domain name test is a good one. Things have unique names that persist when they have continuity of identity, defined in terms of continuity of memory. And you can tell when projects are taken over by impersonators. They start acting like entities with different memories, maintaining only the facade of familiarity (brands often go through such amnesiac break-points).

So saying “no new top-level projects, where projects are 10-year attention commitments” is almost like saying “no more kids, but I’ll be a committed parent to the ones I do have, and grandparent to any any kids they have.” As with kids, you’d be making such a commitment knowing very little about what they might turn out like when they grow up, what their interests and adult personalities will be like, who they might marry, or what sorts of grandkids they may inflict on you.

You’re making a commitment to the being that will treat today’s experiences as memories 10, 20, 30 years from now, come what may. At least I assume that’s the sort of commitment actual parents make, since I’m not one and have no plans to be one.

A project can be like a child only if it has the capacity to develop such a memory if itself. To borrow a phrase from the philosophy of mind, there is something it is like to be that project.

Once you think in those terms, you look at projects with an interesting memory lens.

For example, one of the more troublesome projects on my candidate list is the ribbonfarm school. It passes the first test (it’s not a separate, named top-level named thing, and I have no desire to spin it off as an independent brand), but the second test gives me trouble. Yes it has a name, but it doesn’t seem to have a structural capacity for memory. There’s nothing it is like to be the ribbonfarm school so far.

While schools in general of course have memories and histories, I don’t know what it would mean for this school to have a memory, or if I’m capable of building one that does. My mental model of the school is ahistorical. It’s a portfolio of courses that I mostly haven’t updated at all. It’s a collection of random acts of teaching.

Tiago Forte has found a memory-like angle to online courses: cohorts. If you do cohort-based teaching, then the successive “classes” form a memory stream, and each offering is a memory-creating event.

Then, even if you change platforms, styles of teaching, and the syllabus, but keep the name and curate a cohort-based alumni community, the course has a memory. It’s something you can meaningfully commit to for 10 years, because there is “something it is like” to be that course for ten years, defined in terms of a memory-like unfolding process. The same logic applies at school level too. The entire curriculum and course catalog could change without disrupting the memory stream, so long as the alumni community remains a continuously evolving body.

This is a good memory angle, but one that doesn’t really appeal to me. If repeated offerings and cohorts are the only way to do schools with memories, then I don’t really want to do a school, and should shut this one down.

I do see one promising memory angle that’s not cohorts — the Oxford-Cambridge style tutorial system. I’m considering rethinking the ribbonfarm school along lines inspired by that.

Memory in this sense is like compound interest looked at from the other end. What is it that has been compounding? How does it shape the identity of the thing-that-is.

I haven’t yet committed to either rule, or properly parsed my portfolio of projects through this lens. It would be a pretty serious move if I did. But so far, I don’t see any fatal flaws in either one, and really liking the idea.

Series Navigation<< MJD 59,256MJD 59,354 >>

Get Ribbonfarm in your inbox

Get new post updates by email

New post updates are sent out once a week

About Venkatesh Rao

Venkat is the founder and editor-in-chief of ribbonfarm. Follow him on Twitter


  1. What do you call a fox that’s tied himself to a tree? Or actually this reminds me more of the fox like little prince making himself at home on his tiny planet 😃

  2. Project versus the contents of one … domains versus the subdomain of one?

  3. Sounds like a manifesto, though not born from youthful energy which shouldn’t be wasted but funneled into great success but anxiety of aging: one has to care about ones savings as one hardly gets another big chance. Ageism is a thing and it is fine to rationalize it with “neurodegeneration” as an unavoidable fatality. Follow the science or so. Even High Postmodernists may be instantly convinced as they don’t care about contradictions as long as they can quote the right people.

  4. Max Tanner says

    > Then, even if you change platforms, styles of teaching, and the syllabus, but keep the name and curate a cohort-based alumni community, the course has a memory.

    It’s not clear that an unchanged name and an alumni community are the prerequisites for continuous memory. The Recurse Center is perhaps a good test case for both:

    1. The Recurse Center was originally called Hacker School. There weren’t any significant changes to the experience of attending Hacker School / Recurse Center (i.e. change in self through forgetting) after the name change.
    2. The existence of an alumni community (situated in an internal chatroom where alumni and current participants interact daily) was not enough to persist the Recurse Center’s “self-understanding”. They moved to overlapping cohorts (“batches”) exactly to remedy memory loss occurring at the end of each batch: https://www.recurse.com/blog/36-overlapping-batches

    The overlap of cohorts changes the way that memory is persisted. When batches do not overlap, memory is transmitted from alumni to participant through recollection: “this is what the Recurse Center was when I was there.” In this system, the Recurse Center is not an ongoing phenomenon. It ceases to exist in the present each time a batch stops. And even if batches start and stop in quick succession, so that alumni were very recently participants, the continuance of the experience depends on new participants to restart something they know only second-hand. In the gap between stopping and restarting, something (apparently a lot of something) was lost.

    What was lost isn’t information about what happened or why it happened. It’s customary for Recurse Center participants to keep a log of what they do each day with statements about motivation and reflectons on outcome. There’s also an internal shared calendar which records the history of group activities. And the internal chatroom documents bits of cultural history in a high level of detail. So even without direct intervention from the alumni, there are persistent records of what has happened at the Recurse Center and why.

    What was lost at the end of each batch wasn’t knowledge of the past but a feeling for new participants that the past of the Recurse Center is *their* past, not a past belonging to someone else. The continuity of the Recurse Center’s memory depends on individuals owning memories of things they didn’t experience. Thus, continuous institutional memory is actually an occurrence of this:

    > Even the most radical change in your life won’t randomly graft somebody else’s memories into your consciousness, or get you thinking of your own memories as someone else’s.

    The past of the institution is grafted onto the memory of the present participants. That grafting occurs during the overlap of batches. The new batch flows into the experience stream of the old batch and they begin to share this experience, which is based on and formed by conditions which preceded them. Members of the new batch come to understand the roots of their experience through the memories of members of the old batch. That’s the graft – “I’m having this experience because of something that happened to someone else. I retain their memory of what happened to them so I can understand what is happening to me.”

    Perhaps Theseus’s ship continues to be felt to be Theseus’s ship not because it continues to be called that or because there continues to be a group of people who remember it that way, but because there is an unbroken chain of custody, custody of experience, leading back to Theseus.

Leave a Comment


This site uses Akismet to reduce spam. Learn how your comment data is processed.