cross-posted from: https://lemmy.world/post/30459966

WASHINGTON (AP) — The White House will fix errors in a much-anticipated federal government report spearheaded by U.S. Health and Human Services Secretary Robert F. Kennedy Jr., which decried America’s food supply, pesticides and prescription drugs.

Kennedy’s wide-ranging “Make America Healthy Again” report, released last week, cited hundreds of studies, but a closer look by the news organization NOTUS found that some of those studies did not actually exist.

Asked about the report’s problems, White House press secretary Karoline Leavitt said the report will be updated.

  • Null User Object@lemmy.world
    cake
    OP
    link
    fedilink
    arrow-up
    48
    arrow-down
    1
    ·
    9 days ago

    NOTUS reported on Thursday morning that seven of the more than 500 studies cited in the report did not appear to have ever been published. An author of one study confirmed that while she conducted research on the topics of anxiety in children, she never authored the report listed. Some studies were also misinterpreted in the MAHA report.

    And no mention anywhere in the article that these kind of errors are symptomatic of LLM generated text.

    When is the media going to stop dancing around shit in order to appear unbiased and start doing their job as the Fourth Estate by calling out obvious horseshit?

    ETA: Same goes for the original article that the above linked article references. Listing lots of errors that are so ridiculous, that they’re either AI/LLM generated, or somebody intentionally inserting errors to make it look like the report was LLM generated.

    And yet,even that article dances around the obvious issue without apparently having the simple courage to just say, “These kinds of errors are highly symptomatic of AI/LLM generated text.”

    • LupusBlackfur@lemmy.world
      link
      fedilink
      arrow-up
      14
      arrow-down
      2
      ·
      edit-2
      9 days ago

      When is the media going to stop dancing around shit in order to appear unbiased and start doing their job as the Fourth Estate by calling out obvious horseshit?

      Yeah… That horse has left the barn, turned around, struck a match, tossed it inside, and shut the door.

      So-called “Fourth Estate” no longer exists; it has been cowed into subservience and/or just willingly spouts pure propaganda.

      They bear a preponderance of the blame for why we are where we are in the US today. 🤡 🖕

    • Zaktor@sopuli.xyz
      link
      fedilink
      English
      arrow-up
      3
      ·
      9 days ago

      And this shitty fucking title. “Acknowledges problems”. JFC, they can’t even say what the problems were.

    • d00phy@lemmy.world
      link
      fedilink
      English
      arrow-up
      2
      arrow-down
      4
      ·
      9 days ago

      I would say making statements that the mistakes are suggestive of AI/LLM authorship would be speculative without hard evidence. That said, they should call it what it is, a half-assed report that’s a result of either carelessness or a willful desire to misinform.

      • Null User Object@lemmy.world
        cake
        OP
        link
        fedilink
        arrow-up
        6
        ·
        9 days ago

        I disagree. Saying that the mistakes are LLM generated without evidence to support that claim would certainly be inappropriate.

        However, simply stating something like…

        These types of mistakes are frequently seen in text generated by LLMs, such as when [reference to story about lawyer getting caught submitting LLM generated documents in court referencing non-existent cases].

        is completely factual and highly relevant.

  • jballs@sh.itjust.works
    link
    fedilink
    English
    arrow-up
    9
    ·
    9 days ago

    “I understand there was some formatting issues with the MAHA report that are being addressed and the report will be updated.” Leavitt told reporters during her briefing. “But it does not negate the substance of the report, which, as you know, is one of the most transformative health reports that has ever been released by the federal government.

    Mother fuckers used ChatGPT to write a report for the White House to shape national health policy, and they have the balls to call it “some formatting issues”?

  • tvbusy@lemmy.dbzer0.com
    link
    fedilink
    English
    arrow-up
    5
    ·
    9 days ago

    These losers will use LLM again, and my prediction is it will make even worse errors. Mark my words.