A collaborative study by researchers from three US universities warns that AI models are showing signs of cognitive decay, popularly known as brain rot, when repeatedly trained on low-quality web and social media data.
Scientists from Texas A&M University, the University of Texas at Austin and Purdue University find that continual exposure to junk social media content leads large language models (LLMs) to witness lasting cognitive decline across reasoning, safety and comprehension benchmarks.
The study’s controlled experiments retrained four open-source models on what it classified as “junk data.”
“We construct junk and control datasets from social media (Twitter/X) via the two junk metrics: M1 (engagement degree) selects short but highly popular posts that often engage users longer online; and M2 (semantic quality) flags content based on content styles that draw users’ attention.”
When compared with AI models given controlled datasets, those fed with junk social media posts suffered significant declines in reasoning, long-text understanding and inflating dark traits such as psychopathy and narcissism.
Researchers attribute the damage to a pattern they call “thought-skipping” — models increasingly truncating their reasoning chains as low-quality data accumulates.
The investigators also warn that large-scale retraining failed to reverse the damage.
“Even if we used up all instruction data, consisting of 4.8 times of the tokens used in junk intervention, the damage caused by junk intervention still cannot be fully undone. The gap implies that the Brain Rot effect has been deeply internalized, and the existing instruction tuning cannot fix the issue.”
The report also draws parallels between AI systems and human cognition.
“Brain rot is defined as the deleterious effect on human cognition that comes from consuming large volumes of trivial and unchallenging online content… Therefore, it is natural to ask if the analogous ‘Brain Rot’ emerges in LLMs.”
Disclaimer: Opinions expressed at CapitalAI Daily are not investment advice. Investors should do their own due diligence before making any decisions involving securities, cryptocurrencies, or digital assets. Your transfers and trades are at your own risk, and any losses you may incur are your responsibility. CapitalAI Daily does not recommend the buying or selling of any assets, nor is CapitalAI Daily an investment advisor. See our Editorial Standards and Terms of Use.