Self-Organizing Ideas
The goal of Subconscious is to generate self-organizing ideas.
Organizing ideas through sheer force of will is difficult, and I am lazy. Luckily there may be another way. Given the right mechanisms, the right conditions, and enough energy and time, a system can self-organize.
What if ideas could self-organize? Imagine putting your scrap notes in a box. You shake that box, and after a while you start to notice something interesting begin to emerge. Related ideas begin to clump together. Unrelated ideas split apart. Sometimes ideas die off. Old ideas give birth to new ones. The concepts refactor themselves, evolving into thoughts, theses, outlines, articles, even books, all organically grown from the bottom-up.
How might such a system be constructed? Some threads to pull on…
Self-organization requires a mechanism of composition
Two important prerequisites for self-organization:
Self-organization happens through composition. Composition is combining things with other things to create new things.
You can think of the units and the mechanisms of composition together as an alphabet. Alphabets tend to follow Gall’s Law: simple alphabets produce complex behaviors, complex alphabets produce stupid behaviors. Consider: DNA has only four basepairs, but it was expressive enough to generate you!
To provoke emergence, you want to prefer a small alphabet.
Universal mechanisms of composition maximize expression
When everything can be combined with everything else, you expand the number of possible forms a system can take. Consider: Lego has a single mechanism of composition, the dot. This means any lego piece can be composed with any other lego piece.
Another example: files and copy-paste are like the lego dot of computing. We might think of an app as an alphabet of features that can be used to construct a workflow. As a user, I have a goal, and I use the features of the app in a particular order to accomplish that goal. But an app can only contain a finite set of features, the features the product team can anticipate and imagine. This is a closed set. What happens if my goals are outside the expressive range of the product’s alphabet? What if the product doesn’t have the right features?
Files and copy-paste separate data from application, to enable composition. Composition of what? Apps. Any app can be composed with any other app, as long as they speak the same file type. Any text can be combined with any other text, in any other context, as long as they support copy-paste.
Files and copy-paste enable composition by default. They make the set of composable features an open set. You can combine multiple apps to create workflows more expressive than any single product team in Cupertino might anticipate or allow. That's the difference between a device and a computer. It is what makes a computer an open-ended tool.
Evolution is composition with memory
Evolution emerges in any system with:
Mutation
Heredity
Selection
This describes more systems than you might think!
When we write, we flatten the cloud of associated ideas in our head into a linearized subset (lossy). The reader then unflattens this linearized subset into their own cloud of associated ideas (lossy). Each lossy step is an opportunity for mutations in understanding to emerge. Useful mutations are remembered (selection) and shared (heredity). And so... mutation, heredity, selection. Ideas evolve!
Subtext turns text into composable lego blocks
Prose is hard to decompose. A passage is a careful tangle of interconnected paragraphs, a paragraph a careful tangle of interconnected sentences. A single word can be load-bearing.
Prose typically comes in books, which are a linear bundle of pages. Sentences often break midway, flowing from one page to the next.
If self-organization requires a unit, and a mechanism of composition, this is a difficult unit to work with. It is unclear how to work with it, except as a whole artifact, which is what it is.
If we want to generate self-organizing ideas, perhaps what we want is less like prose, and more like an index card, where 1 card = 1 idea.
This is the motivation behind Subtext markup. Not formatting, but the smallest amount of structure to unbundle text into sub-atomic units, so it can be composed and recomposed.
Now we have units, and a mechanism of composition. We can put all that Subtext in a box, and begin to shake the box. With the right evolutionary mechanisms, perhaps we might see our notes come apart into component ideas. Related ideas begin to clump together. Unrelated ideas split apart. Sometimes ideas die off. Old ideas give birth to new ones. The concepts refactor themselves, evolving into thoughts, theses, outlines, articles, even books, all organically grown from the bottom-up.
Links
How Self-Organization Works, Resnick
Self-Organization and Artificial Life, (Gershenson, et al. ALife Journal)