Research: AI chatbots could make the way humans think increasingly similar.

A new study published in the scientific journal Trends in Cognitive Sciences suggests that the increasing use of AI chatbots may be reducing the diversity in how humans think and express ideas.

 

The study, titled 'The homogenizing effect of large language models on human expression and thought,' suggests that increasing reliance on large language models – such as ChatGPT – may cause the writing, reasoning, and communication styles of many people to become more similar.

According to Zhivar Sourati, a computer scientist at the University of Southern California and the lead author of the study, diversity in language and reasoning not only reflects cultural richness but also plays a crucial role in fostering creativity and innovation.

He argued that differences in perspective and expression give communities more approaches when addressing complex issues.

AI-generated content is generally less variable.

According to the research team, texts generated by large language models tend to have lower variability compared to content written by humans. This means that if users frequently rely on AI for writing or reasoning, their thinking and expression could gradually become 'normalized'.

 

Over time, this phenomenon can lead many users to think in similar patterns , instead of developing diverse approaches.

AI not only searches for information, but also generates arguments.

Sourati argues that in the past, many technologies have also influenced how people access information and perceive the world around them.

For example, the Internet helps information spread faster, and GPS systems reduce the need for people to rely on spatial orientation as much as before. However, these tools primarily serve to store or retrieve information .

Meanwhile, large language models operate differently. They not only search for data but also construct arguments and generate answers on behalf of the user . This can profoundly influence the human thought process.

The risk of diminishing diversity of perspectives and ways of thinking.

The study's authors warn that if people's thinking becomes increasingly homogenized, society could lose some of its pluralism , that is, the parallel existence of many different viewpoints and approaches.

Diversity of thought has long been considered a crucial factor in shaping collective intelligence and societal adaptability. When many people approach a problem from different perspectives, the chances of finding new solutions are also higher.

Conversely, if ways of thinking become increasingly similar, the community's ability to adapt to new challenges may be diminished.

Sourati also argues that the impact of AI extends beyond simply changing how people write or speak. He suggests that large-scale language models could subtly shape what is considered logical reasoning or credible opinion.

The impact could spread to people who don't use AI.

Another noteworthy point mentioned by the research team is that the impact of AI can extend even to those who don't directly use chatbots .

Sourati argues that if the majority of people around us speak and think in a similar pattern – influenced by AI – then others may feel pressured to adjust their own expression to fit that communication environment.

This view is also shared by psychologist Owen Muir. According to him, a form of 'averaged language' is gradually emerging in everyday communication, even when technology is not directly involved in the conversation.