📝 Alex's scrapbook

What does it mean to be a professional meta-cognitionist?

Sep 19, 202514 min read

1. Why I’m thinking about this

Professionally confusion-mining at Outcapped

  • Recently I’ve started working at an early-stage startup (Outcapped)
  • Kegan 3 at my last start-up job:
    • When I was last dropped into an early-stage startup (Refract), I was still essentially Kegan 3 (that is, not an adult yet) - just happy to be there, un-opinionated, “I’ll just defer to the founder”, etc1
  • Vs at Outcapped,
    • As a 29 year old and self-described Kegan 4 (lol) having grown in various ways2, I have a more clear sense of “oh, my confusion is useful, and everything hasn’t been figured out yet, so I can help with the figuring out, and I can help future people who are onboarded”, etc
    • Also, ==my resistance is useful== (e.g., to certain tasks or systems), and should be considered as useful signal3
  • So anyway, rather than doing a bunch of concrete deliverables, I’ve mostly been interfacing with my own confusion and lack of understanding
    • E.g., My model of the company has been poor, and knowledge transfer is never straightforward, as it’s an early stage startup where a) documentation is nascent and b) things haven’t been fully figured out anyway

2. Why am I confused?

2a. When you start somewhere, your model of the company is, of course, poor

  • Self-explanatory

2b. And, when it’s an early-stage startup, things by definition aren’t figured out

  • You may not be at product-market fit yet. Perhaps the products haven’t been built yet. The space of possible actions will therefore be maximally, uh... big (it’s late here)
    • You could join a well-defined org with a well-defined role and boom, you know what to do, you do it
    • And that is the case at Outcapped too, if you join and work on the client-facing side. “Oh cool, client x has asked me to do y, I will now do y”

2c. Knowledge transfer isn’t straightforward

  • Even if someone has made an awesome, info-dense thing in say Notion, like a really useful table with an overview of stuff… IMO unless you’re a genius, you can’t just read that and have it click into place in your brain45
  • Re: Michelene Chi’s “ICAP” model - you need to do some kind of active learning, to properly parse/grok the thing
  • Similarly - you can’t just be told about a company strategy and have it click into your brain. You can’t just attend a meeting and have everything click into your brain

3. What can you do about confusion?

”Knowledge management”

  1. Extract pre-existing knowledge (e.g., from founder, from more established employees)
  2. Make knowledge (e.g., combining disparate notes into easy-to-parse maps, red-teaming proposals)
  3. Work to understand things (Hermeneutics (interpretation & understanding)) → work on “grokking” stuff that has already been made

Salient examples of knowledge management

  • (Note: I was once hired as “head of shared knowledge”. Just as a bit of signal that I care about this stuff)
  • Salient examples of knowledge management:
    1. A new person is onboarded: can they get oriented quickly?
    2. Can the company strategy be easily red-teamed?
    3. Is the company strategy well red-teamed and considered?
    4. Are meeting notes turned into genuinely useful artefacts? Are new insights easily folded into key documents and plans?
    5. Is knowledge easy to find? Is the knowledge base self-describing, navigable? Is it overwhelming, is it full of half-finished and defunct documents?
    6. Is the short-term strategy clear? Is it well thought out? How do we know? Is there e.g. a simple prediction market?
    7. Is everyone working on the highest-leverage stuff? (Does each person know what will move the needle?)
    8. Do people know what the current high-level goals are, allowing them to autonomously orient themselves?
      • There’s the Boydian (from John Boyd) idea of “auftragstaktik”, or “mission tactics”, where the higher command gives a clear mission, and the subordinates/officers have essentially total autonomy to achieve the objective. Speeds things up because the officers don’t have to keep pausing to communicate with mission control, there is implicit understanding and high agency
      • “Do all the employees feel agentic and well situated and like they know how to ‘move the needle’“?

4. Why am I currently really into “knowledge management” (implicitly, at least)?

Trying to get to the “why”

  • (Epistemic status: tentative): It feels like I’m no longer good at doing tasks that are just given to me. I need to understand them and why they’re important, why they’re worth me spending my time on, vs other things. I feel less able to do “just trust me bro”-type stuff
    • And when you start at a new startup, you may be given some “hey can you just do these deliverables too please” stuff, and it can feel aversive (if you’re weird)
  • I’d much rather be involved in the meta-layer, figuring things out, than be handed deliverables. I feel like I’m not the guy for one-off deliverables
  • Or, perhaps I’m still good at deliverables, but only once I’ve done my initial orienting
    • Like, “cool, I have a good model of the company, and we all agree that I should be doing x next, because x is clearly the thing that’ll be the most highest-leverage re: y”

Trying to find the most high-leverage things

  • Having a good model = have a good sense of what is most useful

5. How can this be formalised/legitimised into a role? Or an AOR, at least?

  • This stuff all feels kind of… illegible, vague, silly to me
  • I think it’s because it’s (unless I’m missing something!) not a well-established thing.
    • Like, if someone said they work in marketing, you’d be like “oh yes, marketing is a thing”.
    • Whereas I think “I work in knowledge management” is like… wtf is that?
    • So perhaps that’s a sign that I should learn more about it as a field, to see if there are proper answers to my questions here. Like, there may be useful field guides or whatever
      • When I investigated the field of knowledge management briefly when I was at Alvea, I wasn’t particularly impressed with what I found. But that was a few years ago, I imagine I could find much better stuff now, and ask much better questions…

Ridiculous job titles

Head of Understanding

  • This is actually my favourite of the silly job titles
  • “Head of Shared Knowledge” is bad in a kind of multivalent way, as is “Head of Knowledge”. It sounds overblown, and also vague as hell. Definitely not self-describing
  • Also, there’s no clear win condition in the title. “Is our knowledge successfully managed?”, “do we successfully have shared knowledge?”
  • Vs “do we understand?“. “I understand” vs “I don’t understand”, straightforward different. Making sure we all have a shared understanding, vs making sure we all have shared knowledge… the former just tastes better to me

Professional hermeneuticist

  • Hermeneutics is all about "interpretation and understanding"
  • I join a startup, I don’t really understand anything, I iterate and progressively learn
  • I create maps to help my own understanding, and share them with others at the company
  • The maps allow us all to get a better shared understanding, to establish a "shared reality"
  • The maps allow us to spot holes, gaps in the plan, things that don’t make sense, are redundant, etc.
  • They allow us to see “this is the company in schematic form, we all agree that this part is the part we should focus on now, we all agree to neglect these parts”, etc

Professional meta-cognitionist

  • Meta-cogniser? Meta-cognizer?
  • I join a startup, I notice my thinking (“I don’t understand x, my model of y is poor”)
  • I work on improving my models, I share the results
  • Basically the exact same thing as “professional hermeneuticist”
  • Even better, and I guess this comes full circle to what knowledge management is:
    • I join an org, I notice the thinking systems of the org, I improve the thinking systems of the org

“Process improvement” is a good role for a professional meta-cogniser

  • I ended up being a “process improvement specialist” at Alvea, and I think this was a great middle ground (after starting as the “head of shared knowledge” at Alvea and realising that scope was too narrow)
  • Being a “process improvement specialist” basically involved doing basic customer research, where internal employees were the customers. What is their problem space? Get a good understanding of a problem, then solve it
  • This is cool because you create a map of the problem, and then create a solution, implement the solution… you’re doing the full stack

6. In that case → aren’t all ops people meta-cognisers, then?

  • What I just wrote about (“create a map of the problem, create a solution, implement the solution”) sounds like it could describe many roles. Software engineer, product manager, ops generalist, etc.
  • I wonder if there’s a specific role that prioritises one part of this funnel. I think I’d be sad if I was only given one part of the funnel?
    • All you do is map the problem space → sad to not also build stuff?
    • All you do is create solutions → don’t just want to be a e.g. “systems person who makes systems and automations for people”. Sounds dull, divorced of my own autonomy and cognition, “just make this thing to spec, and the spec has already been figured out”. I met someone at EAG who was a systems specialist and I was like “oh cool, what systems have you made” and they were like “oh I’ve made like 8 CRMs recently!” and I was (privately) like “oh god that sounds… so dull”
  • I think maybe what differentiates me is that I want the thinking to be primary. I always want my primary thing to be like “I try to understand things, and then go where I’m needed, and then repeat that process”
    • Understanding as the primary thing?
      • Hence Head of Understanding
  • Knowledge management, learning, and communication/map-making feel like my particular edges, passion
  • And I think, compared to other ops people, I am unusually averse to one-off, ad-hoc tasks. The kinds of things that don’t cumulate/compound, that can be done by anyone
  • There’s a sense that a different ops team member could probably do a better, faster job than me at one-off tasks, and I’d rather be doing the stuff that is unusual to me - the map-making, to clarity-mining, etc

7. How much do I actually need to understand?

As much as feels necessary

  • I personally want to understand as much as feels necessary, in a very like, Gendlin’s Focusing, felt sense way
  • I think I have a decent sense of when I’m confused and it would genuinely benefit everyone to have my catch up, vs a gratuitous like “going deep just to procrastinate”

When a deep understanding isn’t necessary

  • A business might just be like “look, we don’t need you to properly understand x, we just need you to do y thing”, and I might totally agree with them
  • E.g., at Alvea, no one was like “we need you to understand the strategy”, and I was actually content not knowing the strategy, because it was already well defined by a leadership team
  • I didn’t need to learn about biosecurity in order to be useful at Alvea. I didn’t need to have opinions on the strategy - it was taken care of
  • So, I think my current interest is downstream of wanting to be more of a help. Like, I think if I have a proper, high level, meta understanding of the business and plan, I’ll be able to have better contributions.
    • Vs a) I don’t want to be “flying blind”, just shipping deliverables without an understanding of what they’re connected to
    • And b) just shipping the deliverables that I’m told to ship, rather than zooming out and double checking the meta-level model and strategy
      • It feels to me at Outcapped that there is room for manoeuvring re: “hey I know the plan is x, but what about x-v2”

Appendix

Feedback I’ve received re: this

  • Feedback from Ethan, co-CEO of Alvea, re: “Lucidity (mental clarity, ability to make things clear and structured)”:
    • “Impressive ability to crystallise, structure and organise information”
    • “Historically you haven’t celebrated this enough”
    • “Sounds like you think it’s easy to do this and therefore replace-able, not valuable”
    • “Counter-example: the surprise I felt after we had a few conversations re: [topic] and you turned it into something very legible, crystallised it”
    • Even has a guy currently working on this with 25 years experience and doesn’t seem to be able to do the job like I could, I would have grok-d more and put it in a more legible format
    • Therefore this isn't trivial and should be done more!
    • “You blew my mind a few times with our early collaborations, this was really valuable."

"Head of Shared Knowledge” wasn’t actually needed at Alvea

  • I think a few things happened at Alvea

Lack of skill

  • I really knew nothing about org-wide knowledge management. I was hired because it’s something I care a lot about, but suddenly it was hoped that I’d have good interventions for ~25 people in an org. I’d also never done e.g. customer research, never heard of Paul Graham, etc. Basically, I made lots of mistakes, early on

Lack of need (?)

  • This seems like it could be an illusion, downstream of my lack of skill, but still…
  • The org chart was essentially a few small teams - science team A and B, leadership, manufacturing, etc
  • It felt like each subteam member was already an expert, could easily interface with other team members, the shared knowledge could exist in their heads, etc
  • They didn’t need a solution, really, IMO. “Help with knowledge management” actually really a node in their problem-space

Footnotes

  1. In retrospect, I think imposter syndrome is a key blocker to being a good professional meta-cognitist. You need to be able to say (to yourself, and to other members of the org) “I’m confused! This is confusing!“. If you have some imposter syndrome, you think “this confusion is a sign that I’m too dumb to be here”, whereas if you are more of a neutral observer, you may think “aha, so, thing x has confused me, perhaps it will confuse or is confusing other people”. ↩

  2. Post-rationalist arcs of e.g. fixing alexithymia, learning to trust/notice your feelings, the “notice dart 1 vs dart 2” arc of Buddhism, the “have enough self-esteem to know that you’re not an imposter and are therefore safe to speak up re: your confusion”, etc ↩

  3. Instead of “oh god, I just really can’t get myself to start this task, eeee this is bad”, being like “huh, ok, let’s investigate my feelings about this task and see where the blocker is”. Purposefully facing ugh fields ↩

  4. I imagine that if you’re there to do straightforward deliverable stuff, then you don’t need to grok the table fully. But I have a sense that, my whole kind of way of being at the moment is oriented around properly learning stuff. So, I don’t want to be someone who just does the deliverables, without properly understanding the company ↩

  5. Tbf, I think it depends on how much background info you have. Like, if you worked in marketing for 10 years, and then joined a new marketing job, and the head of marketing explained the strategy to you, you’d probably retain most of it, because it’d make a lot of sense to you, and you’d have a lot of stuff in your head that you could attach the new info to. Vs for me, I’ve never interacted with this kind of work before, seeing on-the-ground operational support stuff, NDAs and invoices and etc etc, so a huge amount of stuff is brand new to me ↩

Graph View

Backlinks