"The computer scientists' view of textual content as "unstructured", be it in a webpage or the OCR scanned pages of a book, reflect a negligence to the processes and labor of writing, editing, design, layout, typesetting, and eventually publishing, collecting and cataloging."

from the piece "A bag but is language nothing of words" by Michael Murtaugh from page 79 of Mondothèque: A Radiated Book

@wakest This perspective on "unstructured data" traces most prominantly back through The Unreasonable Effectiveness Of Data (2009, Google), the ideological predecessor of all big data neural networks (I'll mention LLMs specifically in the off-chance this gets boosted to a context where that isn't immediately understood), and was criticized basically immediately on this axis by people doing ACTUAL work adapting structured-data-assembled-for-human-processing into algorithmic use.
I'll try and find the contemporary link to a few responses, that I'm thinking of.

@wakest Handful of links: https://thenoisychannel.com/2009/03/31/the-unreasonable-effectiveness-of-data/
(assembled by Daniel Tunkelang)

The very-brief but evocative criticism I was thinking of: http://datamining.typepad.com/data_mining/2009/03/strings-are-not-meanings.html
(Strings Are Not Meanings, by Matthew Hurst)

@wakest yes very true, typography is a whole thing, and that's what html should be a standin for, like much of this semantic web shit is a bit silly because html is already semantic silly, and texts themselves are semantically rich, inherently lmao. computer scientists need the humanities