Even AI can experience brain rot

New research reveals that large language models (LLMs) are susceptible to 'brain rot,' demonstrating significant declines in reasoning, long-context understanding, and safety when exposed to popular short-form, low-quality online content, mirroring effects seen in humans. This finding, which shows a dose-response relationship, suggests that the content itself, rather than just human susceptibility, can degrade cognitive functions in both humans and advanced AI, posing critical implications for the reliability, safety, and long-term development of AI systems and highlighting the broader societal impact of digital consumption.

Analysis

The article highlights a concerning phenomenon termed "brain rot," which new research indicates affects not only human cognitive functions but also large language models (LLMs). A preprint by Xing et al. on arXiv demonstrated that LLMs exposed to popular short-form content experienced "non-trivial declines" in reasoning, long-context understanding, and safety, alongside an inflation of "dark traits," with a clear dose-response relationship. This suggests the content itself, rather than just human susceptibility, is a significant factor in cognitive degradation.

Beyond AI, the article underscores the detrimental effects of short-form content consumption on humans, linking doomscrolling to emotional desensitization, reduced attention spans, and poor self-esteem. Furthermore, the extensive use of AI tools like OpenAI's ChatGPT (800 million weekly active users as of Oct 2025) and Google's AI overviews (2 billion monthly users) is implicated in reducing critical thinking, as evidenced by an MIT study showing lower brain activity in students.

This pervasive issue carries significant implications for major technology firms, including Meta Platforms (META) and Alphabet (GOOGL, GOOG), which operate leading social media platforms and develop advanced AI tools. The strongly negative sentiment (-0.7) and pessimistic tone reflect growing concerns over the long-term societal and cognitive impacts, prompting calls for policymakers and industry to establish evidence-based standards for platform design and algorithmic redesign to prioritize mental well-being.

Key Decisions for Investors

• Monitor regulatory developments closely, as calls for stricter guidelines on social media algorithms and AI content could impact revenue models and operational costs for companies like META and GOOGL.
• Assess the long-term reputational and user engagement risks for platforms heavily reliant on short-form content, as sustained negative cognitive impacts on users may lead to shifts in user behavior or advertiser sentiment.
• Evaluate the ethical AI development strategies of companies like GOOGL and OpenAI, as the susceptibility of LLMs to "brain rot" necessitates investment in robust content filtering and model integrity to ensure reliability and avoid future liabilities.