Dem
47
GOP
53
Two weeks ago, the Dept. of HHS released a 71-page report on children's health, backed up by 522 citations to the scientific literature. It concludes that childhood health is negatively impacted by ultra-processed foods, environmental toxins, technology, medications and vaccines. Gee, exactly the things HHS Secretary Robert F. Kennedy opposes. Kennedy right, scientists wrong. How convenient. Case closed.
Until two young women interns at NOTUS actually looked at the report carefully and discovered that some of the papers cited don't exist. Fake citations to back up the conclusions. And some of the "authors" of papers cited in the study insist they never wrote any paper with the title given.
After the NOTUS article was published, four reporters at The Washington Post smelled a red herring and began digging on their own. They found substantial evidence that at least part of the report was written by an AI bot, not actual researchers with experience in children's health. AI bots are trained by having them scour the Internet. This means that the batsh**-crazy theories of all kinds found there go into the training and into the bot's knowledge base. Their article gives an annotated list of some of the fake references. For example, the report cites a paper by R.L. Findling entitled: "Direct-to-Consumer Advertising of Psychotropic Medications for Youth: A Growing Concern." Findling is a professor of psychiatry at Virginia Commonwealth University. He has published in the area of psychiatric medicine but when contacted, said he never wrote any article with that title. There are many more such imaginary references. Dr. Ivan Oransky, who teaches medical journalism at NYU, said that the errors are typical of work done by AI. Similar errors crop up all the time in legal briefs written by AI.
Since the NOTUS and WaPo articles were published, the administration has been scrambling to replace the fake citations with real ones. However, the damage has been done now. Needless to say, if the report was based on nonexistent published papers, its conclusions can't be taken very seriously, much as Kennedy would like them to be.
Nevertheless, some parts of the report are probably true, even with fake citations. It isn't a secret that American children eat far too much ultraprocessed food. But the conclusion that vaccines are harmful (Kennedy's hobbyhorse) is based entirely on fake sources, since there are no actual sources showing that. (V)