Peter SIMONFeb 26
Bruce Davie

I really like this viewpoint on AI performing "semantic ablation". It hits home as I work diligently to write a textbook that contains plenty of nuance and (I hope) a unique perspective on the topic (computer networking).

"By running a text through successive AI "refinement" loops, the vocabulary diversity (type-token ratio) collapses. The process performs a systematic lobotomy across three distinct stages:
- Stage 1: Metaphoric cleansing. The AI identifies unconventional metaphors or visceral imagery as "noise" because they deviate from the training set's mean. It replaces them with dead, safe clichés, stripping the text of its emotional and sensory "friction."
Stage 2: Lexical flattening. Domain-specific jargon and high-precision technical terms are sacrificed for "accessibility." The model performs a statistical substitution, replacing a 1-of-10,000 token with a 1-of-100 synonym, effectively diluting the semantic density and specific gravity of the argument.
Stage 3: Structural collapse. The logical flow – originally built on complex, non-linear reasoning – is forced into a predictable, low-perplexity template. Subtext and nuance are ablated to ensure the output satisfies a "standardized" readability score, leaving behind a syntactically perfect but intellectually void shell.
#AIslop #writing

https://www.theregister.com/2026/02/16/semantic_ablation_ai_writing/

Why AI writing is so generic, boring, and dangerous: Semantic ablation

opinion: The subtractive bias we're ignoring

The Register
Feb 25 at 11:23pmWeb
Show threadChristian HuitemaFeb 26
@Drbruced That's exactly my experience with Microsoft's "copilot" when writing documentation. You start a sentence, and it "helpfully" propose a completion, which is its best guess of what I would like to write. Except it would also be the readers' best guess, and it would not bring them any new information. And thus the copilot guess would only be right if I was lazy and predictable and did not care for the reader.
Show threadBruce DavieFeb 26
@huitema I despair when I see competent colleagues putting their trust in ChatGPT et al. for their technical writing. I've already had one experience of a colleague attributing a hallucinated quote to me, but this article on ablation really gets to the more pernicious downsides of outsourcing our work to a statistical averaging machine.