AI's drive for token efficiency may make human language more precise

paulnovosad · x · 2026-08-25

The AI's preference for token efficiency might gradually make human spoken language more efficient as well. Words with redundancies or ambiguous meanings could become less common, while precise jargon becomes more prevalent. This could be seen as a feature rather than a bug, allowing us to spend more time scaffolding and less time bikeshedding over trivial details.

Related event: Economist Paul Novosad: AI-generated jargon is seeping into economics papers and everyday speech(5 posts)→

Original post →

More from AGI Musings

AGI Musings channel →