Quote Tweet
One of the challenges of modern life:
Going deep while also going wide on a variety of different topics.
Conversation
okay so *EX*tension is among other things a modality for finding information that involves buckets of things labeled with some kind of concept and whatever is in the bucket is a kind of instance of that concept…
1
2
the complement of extension is *IN*tension (not intention) which is the actual conceptual entity that identifies the things that go in the bucket.
back in 2017 i argued that information technology dating from sumerian clay tablets only did extension:
2
3
and what computers do is make intension cost the same as extension, which is to say asymptotically zero.
and unsurprisingly, this has consequences, so how do they manifest?
1
2
4
well, since extensions are sets, the sets always (at least for practical purposes) have to be labeled somehow.
this results in familiar artifacts like typologies/taxonomies: structures of categories. you can think of "extension" as being essentially synonymous with "category".
1
5
if you are in the business of creating taxonomies then you will be familiar with this problem: some of the categories will be excruciatingly hard to name.
2
12
and the reason it's hard to name categories sometimes is because the space of possible categories is infinitely large while there are only 600,000 or so words in the english language (and english is a particularly big language)
3
1
7
i know people will accuse me of fanboying here but this was one of the key insights of christopher alexander's phd thesis (amazon.com/dp/0674627512?)
read image description
ALT
1
8
this is the insight more than anything else that has driven my design philosophy:
since there are so many more possible structures than labels, overprescribing categories up front will result in poor structures.
1
6
21
the argument i made in that talk (youtu.be/eV84dXJUvY8) was that categories were for several thousand years the only game in town, because of the materiality (the physics/economics/math/linguistics) of analogue information technology, but computers totally disrupt that
1
1
3
i cover this in the talk: computers basically have variations on a single verb, which is pronounced roughly like "dereference"
Quote Tweet
Replying to @doriantaylor
What other methods exist to identify or address a category besides naming it?
1
6
some of you may be old enough to remember going to a record store and seeing the cds laid out in sections of the store (rock, classical, jazz, whatever)
those genres still exist but their coverage is spotty; now we organize music by what other music it's like—this is intension
2
4
13
imagine you couldn't write a novel or a song or make a movie without first selecting from a menu of genres
this is the prison we have been living in for centuries
2
4
18
this is consequential for things like projects: if you overprescribe the structure of the project, you will incur a nasty surprise when what actually needs to be done (and the order it needs to be done in) does not match your model
2
3
11
so one of the things i have been thinking about a lot and working toward over the last decade is how to organize projects "computationally"—that is, the structure of the project and the sequence of operations are determined on the fly.
1
1
9
that, by the way, is what attracts me to christopher alexander's work, and why i think it is important, because that is precisely how he operated.
(not to mention why i am writing buttondown.email/natureofsoftwa)
1
7
this is also why i have put considerable effort into modeling tools to facilitate organizing projects this way (and just got my first customer; you could be my second)
2
1
5
(don't worry; it will look pretty different by the time you get your hands on it)
1
2
actually the "tools for thinking" people ought to be pretty familiar with intension as a modality for finding information, "organizing information the way your brain works" etc, because yeah that's what intension is: associative linking.
2
4
13
as hofstadter put it, "analogy is the core of cognition"
1
11
i like that talk because instead of buying and reading another massive hofstadter tome (amazon.com/dp/0465018475?) you can just watch that
(i bought the tome anyway)
3
7
but anyway, back to this remark: if you gain knowledge by moving analogically (intensionally) from thing to thing, the structure of your knowledge will be extremely dense but it will also almost certainly be un-labelable
Quote Tweet
neither specialist nor generalist but a secret, third thing
Show this thread
3
9
31
formal education/training/certification &c is organized by EXtension, ie categories, which is why you get the "jack of all trades/master of none" (false) dichotomy
ie if you don't have a degree (necessarily in some topic X) then you mustn't have any expertise at all
1
2
8
universities have faculties, each faculty has to be the faculty of *something*; same goes for companies have to have departments. this is necessary because that's how the money is organized. (see portfolio.du.edu/downloadItem/1)
1
6
organizations are centrally managed which imputes (no matter how much they resist) a hierarchical structure on them because if your departments overlap you can't control the resources allocated to them
(the alternative, intensional modality, is a market.)
2
1
13
i actually run into this problem a lot because every time i have settled on a job title, people assume certain things about my process, my work product, and my role in an organization
2
12
• if i call myself a programmer people assume i'm "not a big picture person"
• if i call myself an information architect people assume i make website navs
• if i call myself a designer people are like "interior?"
2
10
i call this kinda phenomenon the yellow pages effect, because (if you're old enough to remember) the yellow pages is organized by category of commercial offering; if your business isn't in a clear category then you can't advertise in the yellow pages
1
13
(not to be confused by the other yellow pages effect which is businesses naming themselves something that begins with A so they show up at the top of the list)
1
4
this is also why in that talk i took a few potshots at LATCH (location/alphabet/time/category/hierarchy), because that metacategorization scheme only matters under extension
1
2
4
Show replies