I have a long overdue (several years, half-drafted) piece called “Good Fences” that goes right back to basic “classification & identification, definition & description” of stuff we deal with in the real world: How we subdivide the world into different kinds of stuff and different things and give them various names. It occurs everywhere from the fundamental ontologies of existence to the woke machinations of identity politics – where anti-woke = the new woke, etc. From clarity and useful value to divisiveness and oppression of freedoms. (Still not that essay, but …)
Sam tweeted this today – it really is useful and necessary to be clear about putting things in good and bad pigeon-holes, even whilst rejecting absolute good and evil.
…and to say this is not to say that Ukraine is entirely good, nor Russia (Putin) entirely evil; it is simply to say that on the matter specifically at stake it is not possible to be a person of good will and support the invading army. No, they must be routed and driven out. https://t.co/m6cXLmrGMB
— Sam Charles Norton (@Elizaphanian) May 23, 2022
It struck me as worth recording, because also today there were two “day job” LinkedIn posts – one by Ben Taylor on fitting things into neat categories and another by ex-colleague Keith Williams on the new buzzword “Data Lakehouse”. The former emphasising the possibility of clarity without exclusivity – one point of the Good Fences meme. The latter a compromise – or rather useful integration – between the well defined (well structured, homogeneous) data “warehouse” databases and the more loosely defined collections of heterogeneous objects in data “lakes”. A passion of mine I’ve for many years called “just-in-time vs just-in-case” mark-up tagging – to combine views of well identified and versioned but loosely modelled resources with tighter fine-grained semantically defined resources. I tend to think in terms of the resource in human digestible “blob” form with a keeper or guardian-angel bracelet or lanyard with sufficient linkage to additional (and additive) semantic definition as fine-grained as needed – as and when necessary.
And, because my current read, Kevin Mitchell’s Innate is – like Solms’ Hidden Spring before it – using the same raw, categorical “affect” of good or bad compressed short-hand as the root of all conscious experience and hence all intellectual (scientific) conceptualisation and categorisation.
In fact the whole “systems thinking” / “active inference” approach encouraging seeing different strongly emergent levels / contexts for different levels of granularity and precision in the “definition” of subject-matter content.
(And “computation as compression” is as old as this blog.)
Jeez. Just write something!