Today’s irony: the author of a study on the impact of artificial intelligence on the human brain intentionally inserted small “traps” into their article that prevent AI from correctly summarizing it.
Recently, heated debates erupted online around an MIT article where scientists examined how AI affects the functioning of the human brain.
What was done? A group of students was divided into three parts and asked to write essays. Some used ChatGPT, others used Google, and the third wrote independently.
The results were quite predictable: those who used AI engaged fewer neural connections, struggled to express their thoughts, and in the final versions, they stopped mentally exerting themselves and copied the text with minimal changes.
In contrast, those who wrote without external help or actively searched for information via Google showed significantly higher brain activity and felt satisfied with their work.
The authors of the study do not make any sweeping conclusions. They honestly admit that the sample size is small and that experiments need to be continued. The main idea was to draw attention to the fact that students increasingly use AI for mechanical copying — which makes the brain lazy as a result. However, they add: “In some cases, AI can even accelerate learning and facilitate skill development.”
Work is work — but online, it was quickly blown up to the level of “AI kills the brain.” Moreover, many didn’t even read the article or familiarize themselves fully with its content — they were simply fed through automatic summarization systems. The author anticipated this and inserted several “backdoors” into the text that caused models to ignore important phrases or semantic blocks. As a result, the short summaries often came out like: “ChatGPT destroys your neural connections,” and all of this was actively reposted and discussed on social media.
Such is the amusing story. The moral remains with each person — decide for yourself.