A recent study from the Max Planck Institute for Human Development (Germany) has made an interesting discovery that humans are currently communicating more like ChatGPT. And the use of this AI tool in daily life not only affects the way people write but also gradually shapes the way people talk.
Within 18 months since ChatGPT became popular globally, researchers have discovered that users are increasingly using phrases that are considered ChatGPT's characteristics, such as "delve", "realm" (field), "meticulous" (meaning)... These words rarely appear in common language but are used by AI regularly.
To verify, the research team asked ChatGPT to edit millions of pages of text including emails, articles, news and academic documents.
From there, they filtered out a list of the most commonly used words by AI, called "from GPT".
They then analyzed more than 360,000 YouTube videos and 771,000 podcast sessions to see how these words appear in English before and after ChatGPT's launch.
The results showed that "GPT"s increased significantly in online conversations. According to researcher Levin Brinkmann from the Max Planck Institute for Human Development, this shows that AI is not only learning language from humans, but on the contrary, humans are also affected by the language of AI.
Mr. Brinkmann emphasized that: "We tend to imitate people or entities that we consider knowledgeable and powerful".
Some other "GPT terms" that are gradually becoming popular in daily communication include: "underline", "insight", "advocate", "boast", "swiftly", " Investigate" and "groundbreaking".
This study not only reflects the profound influence of AI on modern society but also raises the question: when language becomes a hybrid between humans and machines, will our way of thinking and perception also change?
