top of page

🧟 MIT proves AI is rotting your brain. 🧟‍♀️

  • Writer: candyandgrim
    candyandgrim
  • 2 days ago
  • 2 min read

"83% recall failure. 47% drop in brain connectivity."


Shared confidently. 


Cited in boardrooms. 


Used as a stick to beat AI adoption.



So let's actually stop and think this through. 🤔 


The study is real. 


The researchers are credible. 


But here's what the viral version quietly dropped:



It's a preprint. Not peer reviewed. 54 participants. All from Boston. All writing SAT essays. The lead researcher herself asked people not to use words like "terrifying" and "brain rot" when describing it. That request was almost universally ignored.



The recall finding (the one everyone cites) has a hole in it. By session three, participants weren't writing anymore. They were copying and pasting. You can't measure memory failure for content you never cognitively processed. That's not cognitive decline. That's not reading something you submitted.


Here's the one that stops me completely.



Before mobile phones I could recall around 30 numbers from memory. Family, friends, area codes. Now I know two. 


But I didn't lose that capacity...I reallocated it. The same brain now manages hundreds of passwords, pins, and security credentials that didn't exist before. That's not decline. That's adaptation. The MIT study doesn't account for what the freed cognitive load is actually doing.



This matters because decisions are being made on the back of it.



AI tool bans. Policy proposals. Strategy rewrites. All citing a 54-person preprint as settled neuroscience.



This isn't the first time a single study hardened into unquestioned truth before the science caught up. The alpha wolf theory: dominance hierarchies in wolf packs, was taught as gospel for decades until the researcher who proposed it spent years publicly trying to retract it. 


Nobody listened! 


Power poses failed replication. The 10,000 hours mastery rule is a misreading of the original research. 


The pattern is consistent: clean numbers, confirms existing fears, travels faster than the caveats, researcher's own nuance stripped out in transmission.



The MIT study might be directionally right. The signal might be real. But it hasn't earned the certainty being placed on it yet.



And here's the uncomfortable irony: uncritically sharing a study about uncritical thinking is a live demonstration of the very problem it claims to have found.



Show me the replication. Widen the sample. Then let's talk.

 
 
 

Comments


bottom of page