Social Media Users Notice They Are Mimicking Writing Style Of AI Models: 'Am I Going Insane?'

In what can only be described as a life-coming-full-circle moment, a social media user has gone viral for pointing out that they were unconsciously mimicking the structured, precise, and often repetitive writing and speaking style of artificial intelligence (AI) models. Originally trained on vast amounts of human-generated data, the large language models (LLMs) have come under scrutiny in recent times for their 'It's not X, it's Y,' style of communication that is evident across the mainstream AI chatbots.

The user highlighted that they were not a native English speaker, so most of their knowledge about the language came from movies, online articles and social media. However, the excess usage of AI in recent times may have brought about a change in their vocabulary.

"Lately, I am interacting with AI more than online articles for knowledge and news. Today, I suddenly realised that I have started mimicking LLMs' style for a while. I have started using the patterns like "It is not x, it is y" or so," the user wrote in a Reddit post.

The user said they could not explain, but there was "clearly" a pattern of being influenced by AI when wrting or saying something.

"It is quite reasonable, as I am getting most information through AI these days, but I have a weird feeling. AI was supposed to learn from humans how to talk and how to make sentences effectively. Now, it has started to go in reverse," they wrote.

"I just want to know if I am going insane or it is happening in general, especially for non native speakers."

Also Read | Foreigner Praises India's Healthcare After Receiving Free Treatment At Government Hospital: 'I Was Surprised'

Social Media Reactions

As the post gained traction, social media users said it was simply the completion of a conversational loop as humans were now learning from AI bots.

"LLM's write this way because they learned it from all the texts they inhaled. Now you're learning it from them," said one user, while another added: "I've started saying "that's a good question" before I answer anything."

A third commented: "It's natural to learn from the things you encounter regularly. If you want to not sound so much like an AI, read a range of different styles of books. Broaden the types of text you are reading."

A fourth said: "Not just you. I've observed the same thing. It's hard not to get sucked in when you're constantly interacting with AI style of writing. If everyone else is sending you messages and emails written this way, you just naturally start responding in a very similar way."