Artificial intelligence is homogenizing human language, draining writing of its unpredictability and pushing speakers toward a narrow band of standardized expression. Research shows that since ChatGPT's public launch, the diversity of how people write has measurably declined, with AI-favored vocabulary trickling into everyday conversation and reshaping what counts as polished communication.
A University of Southern California study that examined scientific journals, local news articles, and social media found sharp drops in writing style variation after the chatbot arrived. Simultaneously, researchers at the Max-Planck Institute for Human Development analyzed hundreds of thousands of hours of content and detected a flood of AI's signature words: "delve," "meticulous," "boast," "comprehend." These terms are appearing with growing frequency in ordinary speech and writing, even among people who never touch an AI tool themselves.
The mechanism is subtle but powerful. Users encounter AI-generated text, absorb its polished cadence, and begin mimicking it to achieve what feels like influential, authoritative writing. "People get used to this idealized, very predictable form of language, and even people who are not using it, in order to have that sense of powerful, influential writing, they start writing more like LLMs," says Morteza Dehghani, a USC professor who led the research.
What emerges is technically correct but emotionally hollow. Alex Mahadevan, chief AI instructor at the Poynter Institute for Media Studies, describes AI writing as noticeably "soulless" and "mediocre." He calls it artless, despite its grammatical precision.
The scale of AI adoption makes avoiding the technology difficult. A 2025 Brookings survey found that 32 percent of small businesses deploy AI for customer service and customer outreach, while 16 percent of individuals use language models for communication or social media. Emily Bender, a linguist at the University of Washington, said she does her best not to read synthetic text at all. But often, she receives messages from others without knowing they came from a machine.
Linguists and writing experts worry that standardization erodes authenticity. Bender warns that chasing what she calls "ChatGPT level polish" silences genuine voices and encourages the "LinkedIn average," that bland corporate vernacular dominating the professional networking platform. Mahadevan expressed nostalgia for what he terms "good bad writing," the kind so poorly executed that it radiates humanity and originality. He now finds himself second-guessing his own prose, wondering whether readers will assume an AI wrote something he crafted by hand.
The deeper cost, Bender argues, lies in abandoned struggle. "There is value in the struggle of writing, because we learn to express ourselves, and we learn to do the thinking that happens as we're writing," she said. "Each time we choose not to do that, we are losing out, both individually and societally."
Author James Rodriguez: "We're witnessing a slow collapse of linguistic personality, and it's happening because AI doesn't make mistakes the way humans do,and mistakes are where character lives."
Comments