4mo ago

RFK Jr’s ‘MAHA’ report found to contain citations to nonexistent studies

www.theguardian.com RFK Jr’s ‘Maha’ report found to contain citations to nonexistent studies

Failures uncovered as US health secretary touted ‘gold-standard’ science in health report ordered by Trump team

Failures uncovered as US health secretary touted ‘gold-standard’ science in health report ordered by Trump team

Robert F Kennedy Jr’s flagship health commission report contains citations to studies that do not exist, according to an investigation by the US publication Notus.

The report exposes glaring scientific failures from a health secretary who earlier this week threatened to ban government scientists from publishing in leading medical journals.

The 73-page “Make America healthy again” report – which was commissioned by the Trump administration to examine the causes of chronic illness, and which Kennedy promoted it as “gold-standard” science backed by more than 500 citations – includes references to seven studies that appear to be entirely invented, and others that the researchers say have been mischaracterized.

Hacker News @lemmy.bestiver.se

RFK Jr's 'Maha' report found to contain citations to nonexistent studies

www.theguardian.com /us-news/2025/may/29/rfk-jr-maha-health-report-studies

18 0

AMUSING, INTERESTING, OUTRAGEOUS, or PROFOUND @lemmy.world

RFK Jr’s "Make America Healthy Again" report found to contain citations to nonexistent studies

www.theguardian.com /us-news/2025/may/29/rfk-jr-maha-health-report-studies

133 7

You're viewing a single thread.

50 comments

Ah, written by ChatGPT then.
- That's exactly what I was thinking. A human doesn't accidentally cite multiple made up studies. So these fuckers are writing government policy by saying "Hey ChatCPT, write a scientific paper with citations about how vaccines are bad."
  
  Yeah, humans will cite real studies that don’t say what they claim (either from malice or misunderstanding) or just vaguely say ‘studies say’, not pull an entire citation out of the air.
  I’m leaning towards making it illegal to use AI (generative LLM) in government work. All it can do is mislead and introduce errors.
  
  Good luck with that. Administration not being allowed to cheat on their homework is a pipe dream at this point.
  
  this administration
  
  My friend, you are going to hate a recent memo from the US Office of Management and Budget titled “Advancing Governance, Innovation, and Risk Management for Agency Use of Artificial Intelligence”
  
  Absolutely correct, I do hate it! 🫠
  
  And 70% of people will never check to see if those sources are legit. Kinda like how most people just read the headlines.
  
  More like 99.9%
  
  More like 30%
  Most people don't believe this stupid shit.
  
  And yet, here we are.
  
  Yeah, because we don't live in a democracy.
  
  Right, because in a democracy, everybody reads all the footnotes.
  
  In a democracy the majority is in charge of society.
  And that's not what we have and it never has been. There was a brief experiment with democracy during Reconstruction before white terror destroyed it, but that's it.
  
  What does anything you’re saying have to do with checking citations?
  
  The fact that the majority of people don't believe in this stupid shit should tell you that, actually, the majority of people are informed.
  There's just this persistent and very loud minority of morons who only trust other more charismatic morons who tell them that ivermectin is actually the panacea.
  
  Not checking is one thing, but seeing so many people actually dismissing when someone calls out the factual errors is so infuriating. Like not only you are an ignorant dumbfuck who doesn't bother checking the evidence, but you are stupid too and criticize people who point out the glaring flaws and cry out cancel culture. Or, of course, you are an awful person who doesn't care about the truth as long as you get your way.
  
  ChatCPT
  ChatCCP, ChatUSA, Chat$Trump, TruthChat... What will they come up with next!
  
  a human might not, but worms will.
- I've read three different articles (including the NOTUS one) about this, and none of the alleged "journalists" has the courage to say that these kind of errors are highly symptomatic of LLM generated text. Nobody mentions AI or LLMs even once.
  It's like they're so intent on being unbiased, that they can't bring themselves to connect even the most obvious dots for people.
  What happened?
  Well, the government released this report with lots of weird errors, such as references to scientific research that doesn't exist.
  How could that happen?
  Shhhhh, no no no. We can't talk about that. We're the news media. We just throw puzzle pieces at the busy people trying to keep their underpaid jobs, raise a family, and make ends meet. We don't help them assemble the pieces into a coherent picture. That would be ludicrous.
- Check for the — emdash.
  
  What's wrong with an em-dash? Do I need to stop using it in my daily writing?
  
  Unironically, yes. LLM generated text is chock-full of em-dashes for some reason. I still use regular dashes to separate clauses occasionally - like this - but not full em-dashes.
  
  Ah, glad a single dash is ok - because I use them all the time instead of commas or semi-colons for informal writing.
  
  Guess the option key on my work Mac is gonna get cold and lonely
  
  Oh no, not my emdash!
  
  No! Don't let the robots take our punctuation! What's next, entire words and phrases we can't use because "that's what AI would say"??
  
  You're absolutely right! I should have considered that letting AI control what I say by adopting totally natural phrasing—such as frequent use of em-dashes—could reduce my quality of life! Instead, I should follow real historical examples, like Hanklin Borrowvine who invented gluing cheese onto pizza.
  👠 In conclusion, I should:
  ❎ Not let AI control my punctuation, grammar, and phrasing
  ✅ Use totally natural phrasing, like "I'm a real human" in daily communication
  ✅ Consider the teachings of Tom Morello, who invented machines so that he could fight them as practice and build up enough XP to earn the skill Rage
  
  Be careful about that one, though. In addition to Word, for whatever reason iPhones automatically convert “--“ to “—“ so if you’re dealing with anybody like me who marks mid-sentence breaks with double dashes out of old habit, you’re going to get false positives.

50 comments