In a session today there was some discussion about agentic skills. A skill is mostly a markdown text file often named, yes, skill.md with a natural language text plus some front matter, describing some specific workflow or bit of knowledge or capability.
Somebody external opined that such skills were a good way of codifying organisational knowledge, which might be useful ‘because of the ageing demographics’ on the work floor and people retiring.

Would agentic skills succeed where process descriptions, intranets, 1990-2010 knowledge systems, quality assurance systems all failed when aimed at documenting the actual knowledge present in an organisation?

I doubt it.

Deep knowledge is relational, chunky, only its underlying facts and heuristics can be codified. And people will resist codification, for different reasons. Such as recognising the futility of it, or recognising its aim of making the people with knowledge superfluous. Or both.

Are skills different enough from earlier systems aimed at codification to gain some willing cooperation from knowledge workers?

They might be.

Unlike the earlier systems, a skill is small and its aim is not explicitly ‘codifying everything’. Smaller than you, an information object in itself. And you get to combine such objects into a chain or bundle as a small workflow. They are a much smaller ask than the others were.
Additionally skills can be deployed by yourself in your own work. For processes, QA, KM it was mostly all done for others, for the hand-offs between actors and mostly seen as external to one’s own work. Why would you write down and maintain what you would not use yourself?
That is where skills can have a different impact. And in terms of ease of use, it helps that such skills are in natural language, as it allows you to handwave the parts where you can’t express your own actual knowledge in any precise detail. Which will be often.

So skills will be likely an easier sell to do some codification.
Still the push for skills is to be small, niche and specific. I’ve seen lots that are actually deterministic, in other words can be automated, except perhaps for the decision when to call a skill and with which inputs. Meaning they codify something different than knowledge.
Such codification will not likely form a coherent whole either. Also because skills are regularly aimed at something that is not within the scope of existing knowledge, a helper function next to current knowledge work.

So, no, I don’t buy the ‘skill codification as knowledge transfer’ suggestion to sell skills with.

Bookmarked How Generative and Agentic AI Shift Concern from Technical Debt to Cognitive Debt by Margaret-Anne Storey

I enjoyed this short posting by Margaret-Anne Storey, a CS professor. The effect of using generative tools can indeed lead to loss of overview, and uncertainty about the project I recognise. It creeps in very quickly, especially if I’ve started from something exploratory, as opposed to planned. A cognitive debt accrues because of wanting to move fast or move at all, at the cost of understanding one’s actions in enough detail. It hinders being able to make changes later.

It also makes me wonder something completely different. Partially because of examples I saw last week in Madrid of how BMW and Airbus had sped up some specific tasks orders of magnitude with AI:

If we see companies as slow AI, i.e. context blind algorithms working towards a narrowly defined singular goal (this is where the notion of AI turning all the material in the world including ourselves into paperclips comes from), what methods have we come up with to deal with cognitive debt in organisations? My intuitive response is reporting chains, KPIs, and middle management. Consultancy too, hiring an external actor to blame if needed. That suggests to me we actually didn’t, as so much of that is management-theater. Does any board of any company above a certain size actually know what is going on in their organisations? Understand what consequences changes may have? There’s a world of hurt out there caused by ‘reorganisations’ that all too often seem ritualistic more than rational when seen from the outside.

It may also be why companies easily embrace AI, despite e.g. warnings about cognitive debt. It looks the same as current practice, just with the promise of higher speed.

I saw this dynamic play out vividly in an entrepreneurship course I taught recently. .. one team hit a wall. They could no longer make even simple changes without breaking something unexpected. … no one on the team could explain why certain design decisions had been made or how different parts of the system were supposed to work together. … issue was that the theory of the system, their shared understanding, had fragmented or disappeared entirely. They had accumulated cognitive debt faster than technical debt, and it paralyzed them.

Margaret-Anne Storey

Imagine you have a recipe for a dish you like. You copied the ingredients and instructions from a magazine once, or your mother wrote it by hand decades ago. You decide to use the recipe, and from its list of ingredients you make a shopping list. Some things you already have at home, other items you need to get. You think of the one or two different shops you’d need to go to, and list the different items in a way that follows the order in which you will walk through the store. Would you describe the recipe as creative output? And the shopping list? Yes, no, neither, or both? Which one would you think qualifies for copyright?

The recipe (both the directions and the ingredient list) is seen as a mere statement of fact. Copyright is not applicable. That’s why cookbooks usually have a clear curated selection stated in their title (‘The 50 most cherished Indian recipes from around the UK diaspora’), photos (the dish with a sprig of herbs positioned just so), and anecdotal flourishes (‘Upon entering the village I saw an old grandmother make this in front of her house, and she shared the family secret of this incredible sauce with me’). Because those elements do carry copyright. Just not the recipe as such. Take 500 grams of chopped tomatoes and cook for 12 minutes on a low fire. Add half a tablespoon of powdered cayenne peppers and stir. Serve cold.

Your shopping list is a unique thing in comparison. It contains a curated list of ingredients (excluding the things you already have), and you ordered them to align with your actual physical path through one or more shops that you mentally selected to go to. Maybe you crossed something out, and added it in a different place after first making the list. Maybe you added one or two other things that you need, as you are going to the store anyway. This does clear the, deliberately very low, hurdle for creative effort recognised by copyright law. It’s just that you as author perhaps think of it as ephemeral, trivial, and something you may well leave at the bottom of the shopping cart as you exit the store.
Some lists are of course solely statements of fact, e.g. a list of all the heads of state and the years of their reign, the planets in our solar system in whichever order, a bibliography of an author. When you add a little bit of purpose to a list, moving it to an unique expression of an idea you had, then it quickly becomes something else.

A creative artefact.
A list that is the result of some internal process of yours with some internal logic, even if it eludes another person encountering the list, is more than the sum of its items. Curation, selection and exclusion, ordering, at one or more levels of hierarchy, are determined by and express the intention and purpose of the list maker.

List maker.
It’s something E calls me every now and then, list maker. Because quite often my first response to anything that requires planning, thinking, or writing is ‘I’ll make a little list’. Because sitting down and making a first list is beginning the work. Writing a list is not the result of thinking, but part of the process of thinking itself.
An Outliner tool is a key digital list making aid (even if they all have their limitations). A good Outliner allows you to put making lists ‘on rails’ as Dave Winer put it.
Moving an item up or down, to the top or the bottom. Nesting a thing under another, or deeper still. Moving a nested item up a level of hierarchy. Hide the subitems under a thing, or revealing them. Make a connection with an element elsewhere in an outline or with/in a different outline. Turn lines into bullets into numberings and back. Switch between different types of visualisation, one of which is the outline. All made seamless with keyboard shortcuts.

In the 1968 Mother of All Demoes Doug Engelbart, showing his vision of what computers should and can do, impresses the audience when he moves things in a list around in an outliner and switches between visualisations of the list, before using another outline for a presentation. The list he makes is of course a shopping list.

Many of the lists I use emerged over time from my notes and work, a type of emergent and earned structure. Some have both an outline structure and a more visualised networked one (a tool like Tinderbox allows you to switch between views, so does Obsidian with Excalidraw). Some have a bit more complicated inner structure or are partly dynamic or help with decision making, making them small knowledge machines.

There are many types of lists I regularly make and use.

  • Checklists for various processes and events (like travel), and periodic reviews
  • Dashboards, which are (check)lists in two dimensions, that ensure I take into account all aspects of something.
  • Daylogs with links to appointment notes, listing events, links to things I found and interstitial journaling.
  • Maps of Content (or elephant paths as I call them)
  • Memory palace overviews (listing loci) of places I might use as mnemonic device
  • Card decks for spaced repetition.
  • Outlines of texts I’m writing, outlines of presentations, both with links to underlying notes and references.
  • Interests I currently have (questions, examples) and how I might see them as elements of practice, knowledge fundament etc.
  • Tasks lists, selected on context (train, home office, company office etc), effort, energy level and time needed.
  • A spreadsheet that provides, yes, my shopping list for larger parties, based on the list of participants, their dietary requests etc., based on previous parties and amounts consumed.
  • Book lists

Book lists are like any list in that they involve selection and ordering too. They are also a bit more than a creative artefact.

Book lists are libraries.
Any list of books you create is a library, even if it’s a library of the mind that you make tangible in a list.
The list of books that are in my home office book case for instance is a simple example describing that part of the actual collection of physical books in our home. But there are more book lists I work with.
All the books, whether I (still) own them or not, that I have read in a given year.
All the physical and electronic non-fiction books I own and have not read yet, by topic. Or don’t have yet but thought interesting enough to note. They form a reservoir of preselected books I thought might be of interest at some point, that can serve as a research tool. That list I call my anti-library.
Or a similar list for unread fiction books I have, to use when selecting a new book to start in.
A list of books I may want to acquire at some point (generated from what I come across online and in shops that looks interesting or fun, without buying it), a library of wishes of sorts.
A list of books I definitely do not want to own or read, which contains books I have regularly come across thinking they looked interesting, and then realising I had rejected them a few times in the past already, and also contains authors I want to avoid.

These lists overlap, interlink and morph. Most of my daily note making is in the shape of lists, where items may get extended into paragraphs. Using outliner functionality I move them around, extend, link and change them. Parts get shunted into their own notes, some becoming a note in my core personal knowledge notes, others ending up in more mundane notes. Most will remain where I wrote them. Some will become lists I use more frequently or have a structure that is a piece of personal knowledge in itself, such as the ones listed above.

Making lists is not a chore or something predefined, but key to the work of eliciting meaning from all the disparate things I encounter in a day. It allows manipulation of all the small bits of information, from which meaning and structure may emerge. It’s a way to locally reduce entropy in my notes where useful.

Make lists. As your creative artefacts, as your libraries.

left part of screen shows an outlined list, right part shows Doug Engelbart during the 1968 Mother of All Demoes.
State of the art list making, 1968. (Screenshot of the 1968 demo by Doug Engelbart of an outlining tool)

Obsidian has added a functionality called ‘Bases’. An interface driven way to make overviews like you already could make with Dataview queries. Where Dataview is a plugin, Bases is a core functionality. Bases use properties for filters, and I don’t really use properties, at least not as frontmatter like Bases assumes. Such frontmatter properties suggest the need for structure and consistency that can only feel like a burden more than an aid. I do have inline data fields (inline works better for me, because I can turn any part of a note into a piece of data that way).

Perhaps it helps make overviews of ongoing projects somewhat easier to make. Let’s see.
I’m hesitant as it seems a step towards Obsidian trying to be everything to everyone, bloating functions, whereas for me having plain text notes that I can approach with any other tool besides Obsidian is the key thing to ensure works.

The past year I’ve had more conversations about personal knowledge management than for a long time before. Regularly questions come up along the lines of what structures should one use, what should one keep in their notes, what are the best tools? Most of my contributions to such conversations boil down to ‘there’s nothing you should do other than what fits you, and I could show you what I do and why that works for me’.
Ultimately I keep thinking people don’t take the P in PKM far enough. I think PKM is personal not just in one way, but in at least three ways.

  1. It is personal knowledge management in the sense that your system is yours only. You do it, build and extend it as you see fit. It can exist independently of whatever other working environments you have, it is fully personally yours. This enables and secures your professional and learning autonomy, allows you to be pro-active. You bring professionalism to its upkeep and tools. When you go someplace else, you take it along with you.
  2. It is management of your personal knowledge. Personal knowledge in the sense of your own networks of meaning, your insights, the expressed connections you made between pieces of it, your associations and verbiage that tie into your internal world. It exists to feed your agency, your combinatory creativity, and is aimed at your own mix of evolving personal interests. It’s there for just you to interact with. This allows you to steer the direction of your own learning intentionally, and enables an activist and artisanal stance towards your interests, professional or otherwise. Your personal knowledge doesn’t need to be understandable or be useful to others. Sharing in context is an act in itself, an output rather than a function.
  3. It is personal management of knowledge. You have your own methods and structures, that are geared to how you work, think, learn and create best, and which have emerged over time and you then reinforced because of their utility to you. You create your own mechanisms and algorithms working with your material, and reflect on them. Your system does not need to be understandable or workable for others. Your structures fit your internal world, have their logic and starting point in the fact that your behaviour is pretty predictable to yourself at least. Some of the structures in your system may well be just in your head rather than made tangible in your external system. Some structures may be intentionally omitted to ensure you can surprise yourself with your system, feeding discovery and wonder.

I can’t tell you how to do PKM, I can show you what does and doesn’t work for me and tell you why I think that is. I will tell you to take PKM very personally, and then some more.

Bookmarked Tools for Thought Library in Zotero (by Chris Aldrich)

This large and interesting collection of books, writings, quotes and other artefacts on note making and personal knowledge management, tools for thought in short, through the ages is something to explore at leisure. It seems to have been put together by Chris Aldrich some two years ago. Great stuff. Came across it in Chris’s Hypothes.is stream, and didn’t immediately realise it was a reference to his own collection.