AI-Powered Brain Rot: Chatbot Users Have Zero Memory, Social Media Children Have Worst Vocabulary
An AI-powered brain rot study found that chatbot users are experiencing memory loss.
In the age of "brain rot," a slang term describing the deterioration of mental health caused by engaging with low-quality internet content, Oxford University Press, publisher of the Oxford English Dictionary, selected "brain rot" as its 2024 word of the year after scouring its vast database.
Oxford University's definition of "brain rot," which won over a shortlist that included "knowledge," "calm," "romanticism," "dynamic pricing," and "slop," cites how social media apps like TikTok and Instagram have "damaged" people's brains by consuming short videos.
This year's most notable study on the impact of AI on the brain, conducted by the Massachusetts Institute of Technology (MIT), found that students were asked to write essays of between 500 and 1,000 words. One minute after writing, 83% of the ChatGPT users couldn't remember a single sentence.
The New York Times reported on the study, "Students wore sensors that measured electrical brain activity, and ChatGPT users showed the lowest brain activity." They added, "Students using Google's search engine were able to quote portions of the essay, while those not relying on technology could recite many lines, and some even quoted nearly the entire essay verbatim."
The MIT study divided the students into three groups: one group used chatbots, a second group could only retrieve information through traditional Google searches, and a third group relied solely on their brains to complete the essay.
The study, led by M.I.T. Dr. Nataliya Kosmyna, a research scientist at the Media Lab, told the New York Times about ChatGPT users: "A minute has passed and you can't really say anything? If you can't remember what you wrote, you don't feel ownership. Do you even care?"
"I'm concerned about the impact on people using AI chatbots in high-maintenance fields like pilots studying for their licenses," Dr. Kosmyna said. "More research is urgently needed on how AI affects people's ability to retain information."
Professor Shiri Melumad of the Wharton School at the University of Pennsylvania conducted a comparative study in which she gave a group of 250 people a simple writing task. Some were able to use traditional Google searches to find tips, while others relied solely on Google's AI-generated summaries of information.
The results showed that the group using the AI-generated summaries wrote the most common, obvious, and largely unhelpful advice: "Eat healthy food, stay hydrated, get enough sleep." In contrast, those who found information through traditional Google web searches "focused on the various pillars of well-being," including physical, mental, and emotional health, and shared more nuanced advice on these.
The New York Times reported, "The tech industry says chatbots and new AI search tools will enhance how we learn and thrive, and that those who ignore them risk being left behind."
However, Dr. Melumad's experiment, like other academic studies published so far on the impact of AI on the brain, found that people who rely heavily on chatbots and AI search tools for tasks like essay writing and research generally perform worse than those who don't use them.
"It's quite scary, to be honest," Dr. Melumad told the Times. "I worry that young people don't know how to do traditional Google searches."
In the United States, reading scores for children, including eighth graders and high school students, hit an all-time low this year. The New York Times reported, “These results, collected from the National Assessment of Educational Progress, long considered the most reliable gold standard test in the United States, were the first to be released since the COVID-19 pandemic disrupted schooling and increased screen time for young people.” They added, “Researchers noted a growing body of evidence showing a strong link between lower cognitive ability and AI and social media.”
In addition to recent studies that found a correlation between AI tool use and cognitive decline, a new study led by pediatricians found that social media use was associated with lower performance in children taking reading, memory, and language tests.
Students’ social media use may be linked to lower reading scores.
In September, the medical journal JAMA published data from the ABCD for Adolescent Brain and Cognitive Development, a research project at the University of California, Berkeley, that followed more than 6,500 adolescents ages 9 to 13 from 2016 to 2018. Dr. Jason Nagata, the pediatrician who led the study, surveyed all children once a year about how much time they spent on social media.
Then, every other year, the children were given several tests. For example, a visual vocabulary test involved matching pictures with words they heard.
The data showed that children who reported little (one hour per day) and a lot (at least three hours per day) social media use scored significantly lower on reading, memory, and vocabulary tests than children who reported no social media use.
Dr. Nagata told the New York Times, "The only safe conclusion about why social media apps like TikTok and Instagram harm test scores is that every hour a child spends scrolling through them takes time away from more enriching activities like reading and sleep."
What are healthier ways to use social media and AI? Dr. Niigata stated, "Despite research showing a correlation between social media use and cognitive decline, it's difficult to recommend ideal screen time for adolescents, as many children spend time in front of their screens doing things unrelated to social media, such as watching TV programs."
Instead, he suggested that parents implement 'screen-free zones,' such as prohibiting phone use in areas like bedrooms or dining tables, to help children focus on studying, sleeping, and mealtimes.
An MIT study that explored how people might best use AI chatbots to learn and write offers a new perspective.
The study reversed roles in the group and revealed the following results:
Those who relied solely on their brains to write could use ChatGPT, while those who relied solely on ChatGPT could use only their brains. All students were tasked with writing essays on the same topic they had previously chosen. Dr. Kosmyna said, “The results showed that students who had originally relied solely on their brains recorded the highest brain activity after being allowed to use ChatGPT. In contrast, students who had initially used ChatGPT did not reach the same level as the previous group when restricted to brain use.”
The New York Times reported, “This suggests that those who want to use chatbots for writing and learning should consider starting the process on their own before turning to AI tools later in the revision process, similar to how math students learn formulas and equations using pencil and paper and then use calculators to solve problems.”
Professor Melumad of Wharton School, who led the early research on AI search tools, said the problem with these tools is that they automate what was once an active process in the brain—perusing links and clicking to find reliable sources—and turn it into a passive process.” Dr. Melumad said the key to using AI in a healthier way is to be more mindful of how it's used. Instead of asking chatbots to do all the research on a broad topic, use them as "part of the research process," answering small questions like looking up past dates.
However, to learn more deeply about a topic, reading books is recommended.
The earliest known mention of "brain rot" was in 1854 in Henry David Thoreau's classic story, "Walden," about a man who moved alone to a cabin in the woods.
Thoreau lamented, "While England was working to cure potato rot, why weren't they working to cure a far more widespread and deadly brain disease?" Oxford University announced the selection, stating that "young people today often cite the term 'brain rot' to describe 'a deterioration of a person's mental or intellectual state' due to excessive consumption of online content, particularly trivial content, on social media."
Its usage has surged by about 230% over the past year.
Casper Grathwohl, CEO of Oxford Languages, said the term's rise reflects the "tremendous pace of social media-driven language change." He added that "'brain rot' is a phenomenon in which young people are skewering language trends on TikTok."
The New York Times stated that "Oxford's Word of the Year is based on usage evidence derived from a continuously updated corpus of approximately 26 billion words from news sources across the English-speaking world. The idea is to reflect the "mood and conversation that will shape 2024," based on data from last December. The report said on the 1st.
As in previous years, Oxford invited the public to vote on a shortlist of candidates, and the winner was selected by the publisher's expert team based on votes (approximately 37,000 people) and additional analysis.
The "Word of the Year" contest began 20 years ago with the selection of "chav" (a British slang term for working-class people), and over the years has consistently produced new words like "podcast," "selfie," and "post-truth."
The Times reported that “the competition itself has influenced the language, with a flood of news reports last year saying that after Oxford chose “rizz” (Gen Z or Gen Alpha slang for “style, charm, or allure,” possibly derived from “charisma”), usage spiked by more than 1,500%,” adding that “Oxford’s data shows that current usage is twice as high as it was just before the announcement last fall.”