LLMs Can Get “Brain Rot”!

LLMs can Get “Brain Rot”!, Texas A&M University
https://www.arxiv.org/pdf/2510.13928
https://github.com/llm-brain-rot/llm-brain-rot

This might be the most disturbing AI paper of 2025

Scientists just proved that large language models can literally rot their own brains the same way humans get brain rot from scrolling junk content online.

They fed models months of viral Twitter data short, high-engagement posts and watched their cognition collapse:

  • Reasoning fell by 23%
  • Long-context memory dropped 30%
  • Personality tests showed spikes in narcissism & psychopathy

And get this even after retraining on clean, high-quality data, the damage didn’t fully heal.

The representational “rot” persisted.

It’s not just bad data → bad output.

It’s bad data → permanent cognitive drift.

The AI equivalent of doomscrolling is real. And it’s already happening.

This is hilarious! Now you have to protect your AI from stupid people not just your children.

Lastly, visiting social media online isn’t the ONLY way to “rot your brain”. Other ways include:

  1. Not questioning what you are told and believing every word.
  2. Not fact checking what you are told be fore you repeat it to others, and thus operating on hearsay.
  3. Not learning the law so you have to rely on a corrupt lawyer or judge to tell you what the law requires. This makes you a useful idiot.

For fact-checking tools, see:

Media and Intelligence Topic, Section 10: Fact Checking Tools, Family Guardian Fellowship
https://famguardian.org/Subjects/MediaIntell/mediaintell.htm#FACT_CHECKING_TOOLS