AI Is Quietly Rewriting How We Think and Talk
Your words may soon be shaped by machines—here's why that matters.
AI-generated formulations aren’t just outputs—they can re-enter the communicative and institutional practices their training material came from. The article develops a sociotechnical account of this re-entry and introduces “recursive semantic normalization”: once machine-generated formulations are recurrently taken up, evaluated, reused, and institutionally retained, they can participate in reproducing and transforming semantic regularities.
Under “technohybridism,” the relevant unit of analysis is the hybrid semantic system—a recursively coupled configuration of human practices, computational generation, and institutional uptake. The article separates generation, selection, uptake, and persistence, and distinguishes machine-attributable change from change that would have occurred anyway.
It analyzes uptake as routinized, institutionally diffused, and socially stratified, and states conditions under which machine-derived regularities become norms. It identifies social-epistemic consequences: provenance laundering that weakens epistemic vigilance, erosion of the cognitive diversity on which collective inquiry depends, and narrowing of shared hermeneutical resources. The account attributes no normative authority to machine outputs; what changes is the field within which human and institutional uptake operates.
- AI-generated text is being reused in media and institutions, subtly changing how we all write and speak.
- This feedback loop can make language more uniform and less diverse, reducing our ability to spot errors and think creatively.
- We need to be aware of where text comes from and actively preserve variety in how we communicate.
Why It Matters
AI's influence on language could quietly homogenize how we think, making us less creative and more vulnerable to misinformation.