Retrolanguage - blog.blb.works

Abstract landscape illustration in midnight blue, muted sage, ivory, and copper. Fine luminous threads weave across the image, then gradually unravel and fade among scattered geometric fragments, suggesting language and cognition losing coherence. Subtle cyan and amber points form a restrained network in the background, evoking AI-assisted systems. The composition is contemplative and editorial, with no people or text.

Unspooling of Human Linguistic Competency Under Sustained, Unconstrained LLM Use: A Multi-Spectrum Demonstration Test Across Education and Professional Domains

BLB 09, October 2026

This paper proposes and operationalizes a testable theory—unspooling—defined as measurable, time-linked loss in language competence when people use large language models (LLMs) without LLM support. The theory makes specific claims about deterioration in both language production and language comprehension and encoding, with effects expected to be dose-dependent and accelerated by deeper patterns of use.

Abstract editorial illustration of rough data analysis awaiting refinement. A dense, chaotic cloud of tiny geometric fragments, tangled lines, translucent chart-like shapes, and irregular clusters gathers on the left. The elements pass through a subtle central sorting field and become more orderly—but remain unfinished—on the right. Muted charcoal, ivory, dusty blue, and amber tones create a layered, thoughtful technology aesthetic. No people or text.

Two Retrolanguage Articles from 2025

BLB 18, July 2025

(This is a two article set of data I intended to parse/edit into a single article, but I never got to it, so you get it as it sits.) The Hidden Crisis of Human Language Bias & Psychological Risk in LLM/AI The proliferation of large language models (LLMs) has sparked…