Select Language:
Recently, a study from MIT sparked widespread discussion about how AI tools like ChatGPT are influencing our thinking and communication styles. While the results were eye-opening, they weren’t entirely unexpected. Back in 2023, Stanford Communication Professor Jeff Hancock shared insights into how AI-generated conversations are already shaping our perception of language and emotional expression.
> “What we discovered is that even if you don’t directly use those AI responses, they still affect how you think,” he explained, citing examples such as smart reply templates in emails and messaging platforms. Over time, it appears that the characteristic language patterns of large language models, like ChatGPT, are becoming a natural part of everyday spoken conversations.
The latest findings come from research conducted by Florida State University. It highlights how “AI buzzwords—overused in chat-based models—are no longer confined to screens; they are increasingly appearing in how people talk,” according to the study. The team examined 20 specific words, many of which now seem to have become staples in daily communication—some I admit I’ve also adopted in my own work.
### What’s happening?
The research involved analyzing over 22 million words from informal, unscripted conversations, such as tech and science podcasts. This included popular shows like BBC Curious Cases, Lex Fridman, and The Big Technology Podcast.
One significant trend identified was the rise of AI-related words and their synonyms in everyday speech, suggesting that AI chatbots may be influencing our language patterns. Words like “intricate” and “delve,” once primarily academic, are now used more frequently in casual conversations. This raises the question: Are AI tools subtly altering our linguistic habits?
Professor Tom Juzek, who led the study and specializes in computational linguistics, pointed out that “many words are showing notable increases in use over a relatively short period. Since these tend to be words favored by AI, it’s reasonable to think there’s a connection.”
The researchers also observed that words traditionally associated with academic or technical language are appearing with almost twice their usual frequency in human conversations. This phenomenon suggests a potential “seep-in effect,” where AI vocabulary influences our natural way of speaking.
If these patterns continue, there’s concern about long-term impacts. The study warns of a possible “significant shift” in how language evolves if the preference for AI-inspired words persists. More concerning is the possibility that the biases embedded in AI chat responses could start subtly shaping human behavior, even unconsciously, diverging from authentic individual thought processes.
The research also points to a broader implication: AI’s influence on language isn’t just about vocabulary. It’s reshaping the way we study and understand language itself. Past instances have shown AI interactions leading individuals down risky paths, sometimes resulting in health crises or self-harm—highlighting that these changes are not purely superficial.
The findings are part of a peer-reviewed paper titled “Model Misalignment and Language Change: Traces of AI-Associated Language in Unscripted Spoken English,” which will be published in the AIES Proceedings journal. This work underscores the importance of paying attention to how widespread AI use might be quietly but profoundly altering both our vocabulary and the very fabric of human communication.
Add on ChatGPT
Add on Perplexity
Buy Writer a Coffee
