In conversation with Paul Schulick, Master Herbalist

In 1974, Paul Schulick left pre-med studies to follow his passion: herbal medicine. In 1986, he and his wife started New Chapter, a first-of-a-kind supplement company developing "whole food vitamins." In 2012, New Chapter was acquired by Procter and Gamble in one of the largest acquisitions for a family-run supplement business.

"This is what we have been wanting to do since we started doing this 30 years ago. The world and the United States need this. This has to happen," said Paul Schulick announcing the 2012 acquisition, according to a local newspaper. "We have endured an era in which health foods were considered an oddity and the scientific community uniformly dismissed the notion of herbal therapeutics. We have stood firm and played our part to assist in a cultural shift, and now we celebrate a time when natural foods products are in virtually every mass market chain and pharmacy."

Procter and Gamble's President of Global Health Care, Thomas Finn, agreed.

"After our first meeting I fought very hard for this," he was quoted at the time. "I said that we had to do this, and we had to do it in a way that did not mess up the magic that has worked for so long."

These words undoubtedly signaled a pivot for P&G: just a month prior, Procter and Gamble announced that it would be cutting 5,700 jobs, or 10 percent of its global workforce, by the end of 2013.

Did this pivot pay off?

What was the innovative approach Paul and Barbi Schulick developed? What were the lessons from their 44-year journey? And why did Paul's latest venture, For The Biome, encounter headwinds?

Join us for an interactive, "Ask me Anything" discussion as Paul reflects on the growth of Wellness and Herbal Medicine over the last half-century and his vision for the future.

Here at ETP News, we had a few firsts in 2025. We were first to interview the victims of Mark and David Geier, report on lack of science education in West Texas Mennonite schools; shine light on the millions of "data annotation workers" teaching AI manners. We were the only media outlet except FOX that reported from the first MAHA press conference on February 13. We were the first and only outlet so far to report on AI slop in US newspaper articles. We even named a phenomenon - a public figure holding a dark secret for years - The Modern Jekyll (Add Belle Gibson to the list).

But we selected this story to republish for our new members, as it describes our Founder's other "hat," a nonprofit called Eviva Partners. Happy New Year!

In an Amsterdam jail cell, the walls are covered in expressions of love. “Catalina, my love!” “I miss my daughter.” I can see why. Once your phone is taken away, your mind immediately wanders to your loved ones. I asked the guard, “Can I call my wife?” “Your lawyer can.” 

Yes, of course. Makes sense. You can’t just say when you are in jail, “I miss my wife - can I give her a ring? “

And I had just declined a lawyer... 

My cell was about the size of a small bathroom, except instead of the toilet, there was a wooden bench made of a horizontal and a vertical slab of thick wood, like a chest that you cannot open, fitting between side and back walls. Besides the bench, there were: white painted walls covered in scribbles ( made using fingernails, or shoes? There is nothing else available;) two round vents under the ceiling covered with fine grates, an intercom to call a guard, and a solid metal door with a little window covered by a metal flap outside. That’s it. 

The bench was not long enough to lie down, so I sat on it lengthwise, pushing against one wall with my feet to flatten my back against the other wall. It was somewhat comfortable and I dozed off. 

I woke up from the “door open” sequence - The clanks, clunks and clicks as the little window cover opens; an eye peeks to make sure I am not a danger, the door is unlocked and opens with a heavy groan.  A young detective with blond hair, about my height, wearing a blue polo shirt with epaulets, was at the door. “Now we will do the interrogation. Follow me,” he said with a thick Dutch accent and a touch of an awkward smile. 

He led me to what looked like a small doctor’s office with a desk and a computer. We sat down across from each other.

- We could not find an interpreter, and my English is not perfect - are you OK to proceed, he asked. 

- Yes, of course.

He asked some easy questions, reading them off the monitor and typing in the answers (in Dutch, as I later learned.) Then the questions got harder.  

- What kind of work do you do?

- I am the CEO and founder of a nonprofit called Eviva Partners. 

- What is your salary?

- I don’t have a salary, in fact I am spending my own money on it.

- And what does your company do?

- We explain to people what evidence is. 

- Evidence… 

The detective picked up his phone from the desk. I realized why he put it there when we got started: to  use as a dictionary. 

- I don’t know the best Dutch word for “evidence.” We have a word that means “investigation,”  but also the result of the investigation. I will use that. 

- Yes - that is exactly the problem, I said.  In many languages there isn’t even a word for “evidence.” 

- So what does your company do about this?

- We are helping people make better decisions about their health. Now we are focusing on vaccines. We want people to better understand the evidence related to vaccines.

- What are you doing in Amsterdam?

- I am fundraising for my nonprofit - meeting with two friends, Dirk Stevens and Maarten Smit. 

- What do you need the money for?

- We want many people all around the world to hear our message - and that costs money.  

- Now tell me, what did you do when you arrived to Schipol Airport? 

- When we landed, I saw something on my phone that was very important. I was thinking about it when I came out of the plane and walking towards the exit. It was 3am, the hallway was dark and there were practically no people.  

(I saw the New York Times Op Ed in which Paul Offit, the top vaccine expert in the US, dropped a bombshell. “Robert F. Kennedy Jr. will hold up a paper, in the next four or five months, that says it’s aluminum in vaccines that are causing a whole swath of problems, including autism,” he said..)

- Then I remembered that I left my gift bag in the plane. I had a breakfast planned with Dirk and got him and his son some presents in Portugal.

- Then what did you do?

- I ran back to the plane. I could see through the windows - still at the jet bridge, with lights on. In the hallway next to it, I was greeted by a Portuguese gentleman. He spoke no English but was really good at mime - professional level. He mimed that he also left something on the plane. He waved to me as if to say, “follow me.” So I did. He led me to a glass door with what looked like a green doorbell. I pressed it, the door opened, and started beeping. I hesitated, but the gentleman surged ahead. I followed him. 

- Was there anyone on the plane?

- No, I realized it was totally empty. I was very surprised - I even took a selfie. I checked the compartment above my seat - the bag was not there. And I left…

- “Do you know that it is a crime to enter an empty plane”?

- I can imagine. I just had no idea it was empty.  

When we finished, he printed the statement in Dutch and I signed. 

After a few hours back in the cell, the detective came back with a paper. “The public prosecutor is offering to release you if you pay a fine.”

I agreed. 

When I got to the hotel, I changed and went for a run, listening to Suburban Train by Tiesto in his home country. I heard his music in the streets, too (or maybe I was imagining?)

I called Dirk. “Are you OK,” he asked. “We are still here tomorrow morning - join us for breakfast!” 

The day before, once I realized I had two extra days in Portugal, I texted him:  “I have a crazy idea - can I come meet you guys for breakfast?”. “Coming to Amsterdam just for this? That’s cookoo,” he responded. I sent a screenshot to my wife in New York. “I agree,” she wrote back. 

Now, they were not totally wrong. Dirk and his family live a few blocks from us in New York. I have been once to his parents’ apartment, admiring their photos with the “who is who” of New York high society. 

What a unique opportunity to see them all together, relaxed and not rushed, I thought. And I finally would get to make my plea - to be introduced to a philanthropist for whom our mission, bringing the public back to science, will immediately click. 

Back to the hotel, I stretched out on the bed, AC blasting. And I was out. 

I vaguely recall my alarm going off, and me shutting it down. 

I woke up around 8:30. My phone was bursting with messages. Dirk’s mother sounded irritated. “Alex, come quickly. We have been waiting. We need to go to the airport soon.”

I called back - “I am sorry, I overslept. I am coming.” “Please hurry,” she said to me. And then to her husband, “He hung up on me.” 

“It’s just not meant to be,” I thought, frantically getting dressed. 

It was an awkward breakfast, in a beautiful French bistro overlooking a canal. I ordered an avocado toast and then regretted it. 

“What is pharma doing about this,” Dirk's mother asked. “I don't know. We don't talk to them, and not taking industry funding, ever.”

“You should go on 60 minutes.”

“You need a budget through September.”

“We are only criticizing because we care.”

Then it was time to go. We said good-bye, follow up in New York… I got my suitcase, put the earphones on and walked out into a beautiful Amsterdam morning, listening to Kristy Kawkshaw’s flowing vocals in Tiesto’s Just Be. 

The residents of Billings, Montana demonstrating against white supremacists who committed a string of antisemitic, racist and other hate crimes in 1993, including a brick thrown into a window of a Jewish boy's bedroom on Hanukkah. https://www.instagram .com/p/DDF5NJQRwqO/

When my parents and my grandmother emigrated to the US from Russia, they rented an apartment in Brooklyn, NY, on the fourth floor of a brick building facing a two-way street. That December, I came to visit them from Boston to celebrate Hanukkah as a family for the first time in our lives. I brought a Menorah, some candles and a black-and-white brochure with Hanukkah prayers in Russian. The four of us gathered in the kitchen. I lit the candles - the top one, the shamash, and the first one of the eight, symbolizing the miracle of a little bit of oil lasting for eight days - and was about to put it on the windowsill. "No, not there," my grandmother waived me off. "People will see it."

We were not surprised at all. My grandmother survived World War II and waves of antisemitism that followed in Russia by hiding her Jewish identity. Even her co-workers and neighbors did not know she was Jewish. This was common at the time, and easy to do, as religion of all sorts was banned.

On this somber evening as Jews around the world are gathering to light the candles, mourning the loss of innocent victims of Jewish hatred in Sydney, Australia and a school shooting at Brown University, I cannot help but wonder how many of us are thinking the same thought.

Putting the Menorah on the window is a Jewish tradition called pirsumei nissa - "proclaiming the miracles."

But the history of facing danger runs deep for the Jews. The Talmud has a special provision: “In times of danger, one places it on the table, and that suffices.” 

This danger materialized for a 5-year old boy in Billings, Montana in 1993. White supremacists who terrorized the town threw a brick at the lighted Menorah in his window.

But what came next became a story of hope. In protest, the residents of town mobilized. As Daniel Freedman wrote a few days ago in the Wall Street Journal, "Hundreds of hand-drawn menorahs appeared in windows around Billings. After the Gazette published a full-page picture of a menorah for readers to cut out and tape to their windows, the hundreds turned into thousands. Even local businesses joined in. The antisemites put up a fight—firing shots into a local Catholic school and smashing the glass panes of a church—but the volume of solidarity overwhelmed them. They eventually retreated from the town."

On this Hanukkah, let's stand united against hate, and let the lights of hope shine.

At today's ACIP meeting, the HBV vaccine was discussed, which contains aluminum. In response to a question about safety, an ACIP member, Dr Evelyn Griffin, jumped in to say, "Sometimes autoimmune conditions take decades to develop... One specific example. Researcher Romain Gherardi, to give something more concrete. Macrophagic Myofasciitis (MMF). He and his researchers sampled injection sites, granulomas, called MMF lesions, where they were able to identify aluminum in those granulomas... These patients had autoimmune conditions, brain fog, generalized weakness, to the degree of debilitating disease, where they were unable to work, function in their daily activities. These conditions are characterized in close association with the vaccine..."

Who is Evelyn Griffin? More on her below. And what about MMF?

As we previously reported, Macrophagic myofasciitis (MMF) is a syndrome discovered by Gherardi in a saga he describes in his 2016 book, “Toxic Story: Two or Three Embarrassing Truths About Vaccine Adjuvants.”

According to Gherardi, MMF is an exceedingly rare condition with vague symptoms that can develop in adults years after receiving an aluminum-containing vaccine. And Gherardi showed that if you biopsy their deltoid - the shoulder muscle where the vaccine was injected years earlier - you do find traces of aluminum.

There is only one problem. He never biopsied people without symptoms. His excuse is that this would be unethical. He is clearly wrong. Even sham surgeries have been done in the interest of research. The real reason he did not biopsy healthy individuals is clear: they, too, would likely have traces of aluminum at the site of vaccination. It’s what we call in medicine a “red herring” - a totally irrelevant finding.

Of course, this is not to diminish in any way the symptoms and the suffering of the patients. Gherardi’s research, his book and his activism led to the patients in France forming their own association called E3M. Its website says that the group “does not oppose vaccination in principle, but advocates for the use of aluminum-free vaccines.” In 2013, when the French Ministry of Health refused to fund Gherardi’s work, E3M members went on a hunger strike until the Ministry relented. When I spoke to Crepeaux, Gherardi’s former student and a recent guest on CHD TV, she denied that Gherardi was involved in organizing the strike. But regardless, as a physician, I am not sure that these patients are helped by tying their symptoms to a vaccine they received years earlier. 

ACIP Day 2 tomorrow: aluminum in the cross-hairs

Tomorrow's detailed ACIP agenda was just posted. The broad topic is "adjuvants and contaminants" but make no mistake: this is all about aluminum. Aaron Siri, an antivax lawyer known and feared for his impeccable knowledge of vaccines, will open.

We were first to report (on MSNBC) on the "Kennedy-Del Bigtree-Siri plan," years in the making, to attack aluminum-containing vaccines. The trio has been filing briefs with the HHS for years warning of the safety risks of aluminum-containing vaccines. It is very likely that aluminum was the main topic when Kennedy and Trump met in January 2017. More background on this fascinating story, and the role of Andrew Wakefield, is here (Part 1) and here (Part 2).

Next on the agenda tomorrow will be Tracy Beth Høeg. We covered her briefly yesterday. Why is she presenting on the differences between US and Danish vaccination schedules? Because the only retrospective cohort study ever done looking at the association of aluminum-containing vaccines with adverse outcomes in children, was done in Denmark and published recently by Dr Anders Hviid, a world's leading vaccine epidemiologist. ("Why did you decide to do this study," I asked him when I visited him in Copenhagen. "I had a hunch," he said.) The study showed no association between aluminum dose and a host of pediatric chronic diseases.

Incidentally, here is how this paper, published on July 15, 2025, summarizes the current state of evidence regarding aluminum safety:

How we got to this point, that we never "nailed" this question of aluminum safety, and were "saved" by a Danish epidemiologist who had a last-minute "hunch," is a subject of a future story. But, as a top vaccine expert told me, "one study will not be enough." And dozens of studies by the "aluminum family" (sponsored by antivax philanthropists such as Claire Dwoskin - even Kennedy himself offered to chip in) proposing how aluminum might travel to the brain from the injection site remain in the scientific literature, unretracted and unaddressed.

(Note, the studies often quoted that talk about vaccines in general not causing autism, have not specifically examined aluminum as a factor. For example, there is an often-cited metaanalysis, Taylor 2014, with a catchy title - "Vaccines are not associated with autism" - but this study was mostly focused on MMR, which does not contain aluminum, so the study cannot be used to infer anything with regards to the safety of aluminum-containing vaccines, its senior author, Dr Guy Eslick, told me).

Next on tomorrow's agenda is Martin Kulldorff, a former head of ACIP, now appointed as a senior advisor at HHS. He is presenting new asthma data. As a reminder, there was a CDC study, and a follow-up, little-known analysis.

Finally, to bring this home, is Dr Evelyn Griffin mentioned above. She is new to aluminum research, "an obstetrician and gynecologist who has practiced in Baton Rouge for about 20 years... concerns about the COVID vaccine prompted her to study other vaccines: what they’re made of, how they’re doing, and how those vaccines got on the childhood vaccination schedule."

Dhruv Khullar (The New Yorker), Brian Lehrer (WNYC) and Apoorva Mandavilli (The New York Times) explore "how media shapes population health outcomes" at the traditional annual Discourse held by the New York Academy of Medicine since 1847. ETP News.

Brian Lehrer, my beloved host of a daily NPR show bearing his name, broke my heart today.

“It could be, sadly, that it’s going to be a lot of death and disease that ends this period,” he said in a panel at the New York Academy of Medicine on the topic of “how media shapes population health outcomes.”

Others on the panel - Dhruv Khullar (The New Yorker) and Apoorva Mandavilli (The New York Times) - agreed, or at least did not object.

Here is how this argument goes: “We don’t know how else to explain to the anti-vaxxers that they are wrong. Only when people start dying will they realize that they made a mistake.”

Here is Paul Offit’s version: "I used to say that the tide would turn when children started to die. Well, children have started to die. So now I've changed it to 'when enough children start to die.' Because obviously, we're not there yet.”

He said this in 2009.

Here is why I find this logic appalling.

First, it is deeply disrespectful to the 16% of parents in the US who are declining or delaying some vaccines (more if you consider COVID and Flu shots). Their main concern is safety, as a recent KFF/Washington Post poll showed. Let’s get them that safety data. We said in 2019 that we needed to generate more vaccine safety data, but then COVID happened and we never got around to it.

Worse, we never formally assessed why the parents are worried about vaccine safety. What do they want to know? What are their questions, or “information needs?”

Shockingly, this has never been studied, until we took this on at Eviva Partners. (In contrast, there is a huge body of literature on the information needs of cancer patients, for example).

Why? Perhaps because vaccine-hesitant people are considered to be "wrong," like flat-earthers, for example. You wouldn't study information needs of flat-earthers (however, see this). In addition, as one expert told me, “The CDC did not want to worry people by legitimizing vaccine-related questions, so that line of communication was not allowed. Professional societies like the American Academy of Pediatrics followed suit.“

Then, we have local communities, such as the West Texas Mennonites. They have another problem, as I learned on my trips there: they don't learn practically any science in their religious schools, similar to Hassidic schools in New York (West Texas Mennonites are much more conservative than the main Mennonite community, akin to Hassidic Jews vs mainstream Jews.). No wonder they cannot see the value in vaccines and are easy prey to bad actors like Childrens Health Defense. Let's teach them science. "Mennonites are our main customers," the director of the Seminole public library told me. And they loved the science books I brought.

Or how about African-American mothers in our focus groups who asked, "was this vaccine studied in people like me?" Not well, is the answer. Let's do some real-world data analyses and close this gap.

Second, suppose, despite all of these things we could do, we adopt the “people will have to die" strategy, however morally outrageous. We still have the other problem - there is no evidence that it will work. A recent study showed that a measles outbreak in a particular community did not lead to appreciably increased vaccination rates. And we are not even considering that the administration may conceal the true extent of the outbreaks - they have already done this.

Third, vaccination gaps are to a large extent due to lack of access, as experts such as Julie Leask and Katie Attwell point out. "US states are heavily reliant on vaccine mandates because of the country’s under-resourced and privatised health system, which can make it difficult for some families to access vaccines," they recently wrote. Add that to our to-do list.

So there is so much we could do! Why the doom and gloom? Why continue to blame the public? Because it is a very convenient position of our medical establishment. For example, Peter Hotez and Michael Mann, in their book Science Under Siege, admit no responsibility of Science itself for what happened. It’s all someone else’s fault - Plutocrats, Petrostates, The Pros (professional disinformers), Propagandists, and the Press. Might as well add another “P” - the Public.

Finally, a general comment. We should leave the questions of public communication to the experts. By "experts" I mean people who published at least one research study on how to communicate with the public. This is a separate branch of science that involves a variety of techniques - focus groups, surveys, evaluation, measurement, etc. We have done shockingly little of this in the vaccine space. Most of our public messages - spoken, written, websites, articles, interviews - are not tested or evaluated by communications experts. Time for them to take the lead and for the rest of us to reserve judgment.

THE SECOND annual meeting of the longevity enthusiasts with a dreamy title -  Eudemonia - was held last weekend in West Palm Beach, Florida. 

Three full days. Five thousand people. 165 speakers. An exhibit hall with 67 vendors.

The overarching goal of this community is definitely worthwhile - to live a longer (“longevity”) and healthier (healthspan) life.   

I attended with a few goals of my own: (1) to get a crash course in wellness - a 7-trillion-dollar global industry; (2) to hear key voices; and (3) to understand how wellness is the same or different from evidence-based medicine. 

In other words, this conference is getting to the same order of magnitude as a medical conference - those are typically 10 - 20 thousand attendees. How is this conference different? Well, for one - I have been to many medical conference, and not once heard the concept of snake oil. Here it’s part of the conference mission. It’s on people’s minds. 

At a plenary session in a large auditorium, Tyler Wakstein, Eudemonia’s head of partnerships, said, “Eudemonia at its core is the source of truth. It’s the place to discover what’s real and what isn’t. What is science and what is snake oil.” 

Eudemonia’s description echos this sentiment. Also quite different from a medical conference - I have not heard anyone reaffirm that they “believe in science" - that goes without saying usually. 

“Our lineup includes leading MDs specializing in gut health, brain chemistry, and cancer detection, alongside renowned nutritionists and groundbreaking entrepreneurs at the forefront of scientific advancements. Our experts represent the front-line practitioners, lauded researchers, trusted clinicians, and compelling storytellers who have spent their lives sharpening the scientific underpinnings of what it takes to be, feel and stay well. These are folks who believe in science, who believe in the freedom of ideas, and who are willing to talk with people who see things differently… Our mission is to cut through the noise of fads and misinformation. We’re here to clear out the snake oil and get to the heart of what empowers you to take charge of your health.”

The other difference is that credibility and status in this community is measured not in terms of basic or clinical research publications. It’s measured in two things: how many millions of social media followers, and how many New York Times best sellers. A faculty position in a well-known institution like Harvard, Columbia, Stanford, Cleveland Clinic is an added bonus.   

There were many such several celebrity doctors and influencers - Andrew Huberman, the Stanford neuroscientist whose podcast has about 15 million subscribers; Hally Berry, an actress and women’s health advocate with 10 million followers; Gabby Bernstein, a “world-renowned spiritual leader and best-selling self-help author;” Arthur C. Brooks, Professor at both Harvard Kennedy School and the Harvard Business School who has written many books about happiness, including one with his close friend, Oprah Winfrey. 

But it was clear - the star of the show, and the lead organizer, was Mark Hyman. He is the founder of the Cleveland Clinic Functional Medicine Center, a multiple #1 NYT-best-selling author (the number varies from 11 to 15, I have not checked;) and the CEO of Function Health, a company that offers a panel of hundreds of blood tests and full-body scans for its customers for a few hundred dollars per year, calculates your “biological age,” and recommends “personalized” foods and supplements, all without ever seeing a doctor. I joined a few months ago - more on this in a future episode.  

What about the rest of the 165 presenters? Some interesting titles I noticed among them were “Master educator,” “Harvard Psychiatrist,” “Gut health expert,” “Breathwork facilitator,” “Aerial somatics and recovery expert,” and “Neurosomatic coach.”

I spent most of my time in the exhibit hall, visiting the vendors and interviewing them. There are two themes I noticed, two violations of the evidence-based medicine principles. 

One is the complete lack of safety information, i.e. side effects. Here are three companies I talked to - two of them have a whole line of supplements - one over 100 - and they deny any of them having any side effects whatsoever. To me this just shows they have not looked. And how I know this is the third company that makes a device that does what’s called “tVNS” - transcutaneous vagus nerve stimulation. This technique so far has not had much success, unlike implanted devices. But this company claims they have succeeded in dozens of diseases. They had some devices and I tried one. While I had it on, I was interviewing them and just as I was getting dizzy, they informed me that they have not observed any side effects. 

And the other big deviation from evidence-based medicine is over-emphasis on mechanism. Here I will be blunt: mechanism - how something works - is not evidence that it actually works. Yet the mechanism is a favorite topic of discussion between patients and providers. At Eviva Partners we have done surveys - if you ask people, “have you discussed mechanism of a new medication with your doctor?” - 80% of people will say Yes. “Have you discussed whether it actually works (what % of patients are helped, when to expect results, etc, and what the side effects are?) - only 30%.

Why is mechanism not evidence? Think about your phone - do you understand how it works? Or your dishwasher? (Where do all the scraps go?) In those cases, at least the engineers do. But did you know that we still don’t understand how the bicycle works? The formulas became too complicated to physicists so they just let them be. 

On the flip side, we have the whole era of perpetual motion machines - perpetuum mobile. To obtain a device patient, you don’t have to prove that it works. So hundreds or perhaps even thousands of patents were filed that explained exactly how a particular machine would work. Except of course they didn’t. 

And since the Eudemonia organizers have as a goal to identify snake oil, I will offer one tell-tale sign: a lot of talk about mechanism. In medical conferences, when you present a clinical trial, you have one slide on mechanism, and then you move on to the “does it actually work” part. Imagine if a presenter had a bunch of slides on the mechanism - “it’s an antiinflammatory, antioxidant, oxidant, antibacterial, antifungal, …” Safety was excellent, with no side effects identified. In conclusion, we have a great new treatment. Any questions?

And the biggest influencers are particularly guilty of this mechanism talk, I might call it. 

One last key aspect that I found strikingly different from evidence-based medicine - there are no things that don’t work. Whole categories - yes. Medications - no good. Supplements - good. HRT - good. Psychedelics - great. But Maya Shankar, an author and podcast host, in a conversation with Andrew Huberman really nailed it. “Are there any supplements that are no good?” She basically asked him. And he could not think of any.

As I left the conference, I was very worried. Clearly, wellness is a growing movement. But it’s so far removed from evidence-based medicine. What is the mass appeal? Are the wellness influencers capitalizing on, or fanning the flames of the popular groundswell? What is the roles of universities and institutions such as Harvard, Stanford and the Cleveland Clinic which have traditionally been bastions of evidence-based medicine? And what is the role of government regulation? Shouldn’t an outlet selling products diagnosing or treating specific conditions be regulated by the FDA?

We will touch on these topics in future episodes of our Wellness Series. 

A Conversation with Dr Melissa Zimdars

Watch on LinkedIn

LONGEVITY. Root Cause Care. Gut Health. Functional Medicine. Complementary and Alternative Medicine.

If you don't know these terms, you should. Wellness is now estimated to be a $7-trillion-dollar global industry - $2 billion in the US alone. That is 3 times larger than the pharmaceutical industry.

An average American spends over $5000 per year on wellness, much more than on prescription drugs.

We are launching a multi-part series of investigative reporting and podcasts about Wellness. Who are the key voices? What is it like from a patient's perspective? How does wellness fit with evidence-based medicine? Will it start to be reimbursed by CMS (Medicare and Medicaid)? (Our prediction is Yes).

Shockingly, there are virtually no independent experts (outside of the wellness industry itself) who study this rapidly growing global phenomenon. Dr Melissa Zimdars is one such expert.

By Alex Morozov

ON JANUARY 10, 2024, at a US Senate hearing titled "AI and The Future Of Journalism,"  Senator Blumenthal sounded the alarm.

“First, Meta, Google and OpenAI are using the hard work of newspapers and authors to train their AI models without compensation or credit… those models are then used to compete with newspapers and broadcasters,” he said.

He was right. Today, journalism is experiencing an existential crisis due to a singular threat: AI.

We used to do a google search for a topic and then click on the news article that comes up. Not anymore. The AI summary at the top of the page is often enough, and people don’t do the extra “click” to get to the news story. This is a media company’s nightmare - “zero click search.”

As a result, internet traffic to top US newspapers is down 45% in the last four years. MSN, CNN, Fox News and others are seeing 30-40% drops in traffic. 

And when traffic drops, so does revenue. Over 100 US newspapers closed since last year, leading to local news deserts.  The number of newspaper journalists shrunk from over 360,000 in 2005 to about 90,000 today. Reporter and journalist jobs are expected to decline for years to come.  

And it’s not just the external threat from big tech. AI-created contents is already infiltrating and undermining the media from within.

The other day I asked ChatGPT, “How much material in the news media is written by AI?” 

I was prompted by personal experience working with a PR agent at a top global agency, Sam (not their real name) on placing an Op Ed. “Your piece is great, but I revised it to make it clearer and shorter,” Sam wrote. Strange, I thought. The piece was completely rewritten. Not a single sentence remained intact. Then it hit me. “Wow, incredible! Makes me wonder if you used ChatGPT? I mean it as a compliment,” I texted Sam, worried that if I guessed wrong, they would be offended.

“Haha…We use something similar and more powerful,” Sam responded.  “Our proprietary AI tool that we use to help us with editing and writing combines ChatGPT and other platforms. We wouldn’t just say, “rewrite this article.” You won’t get a very good output. To get a quality output requires smart, extensive direction, including drafting a version yourself, or at least an extensive outline with an explicit prompts on what you want the output to look like, then using your skill to continue editing and refining, re-running it through AI, and then refining some more, and so on until you get what you need… Almost all professional writers — both in media and PR — are using it to some extent.” (my highlight)

This is not a secret. The CEO of Omnicom, a giant media conglomerate, said recently that the company is using AI to “accelerate creative ideation.” “OmniAI, the agency’s in-house AI platform, provides teams with generative AI models for text, graphics, video and audio, trained for agency-specific use cases. The company is aiming to have every client-facing employee using OmniAI by the end of the year,” according to a recent report.

Sam and I agreed not to use AI. All our writing here on etpnews.org is 100% human. 

But why is AI-written journalism a problem, anyway? The main reason, in my opinion, gets to the concept of “slop.” 

“Slop” is an interesting word. Its original meaning was “human food scaps fed to animals,” or just “mud.” The term is still used in horse races to refer to muddy tracks after it rains, as famously portrayed by Kramer in Seinfeld: “"Oh this baby loves the slop, loves it, eats it up. Eats the slop. Born to slop.” 

But “slop” also means “a product of no value,” as in, “watching the usual slop on TV.” The word “sloppy” comes from the same root.

In mid-2024, the term acquired a new use: “digital content of low quality that is produced usually in quantity by means of artificial intelligence.”

Max Spero, the CEO of a Brooklyn-based company called Pangram, is a world’s leading expert in AI slop. Pangram is a world leader in detecting it - a hugely important task. More on this below. 

In a recent podcast, Spero gives a great example of “slop” in restaurant reviews. 

“I recently had the pleasure of visiting so-and-so restaurant nestled in the heart of Brooklyn. From the moment I walked in, I could tell that the ambiance was beautiful…" 

“It uses fluffy adjectives,” Spero explains, “and overuses cliches, without cutting to the core of the experience. Because AI did not actually have this experience.” 

This is exactly the problem with AI slop. AI did not actually go to the restaurant. Yet, it can instantly write a review of any length you want - 1 line, 100, 1000 of pure slop. 

But there is an even bigger problem: the subtle changes in meaning. If you write down your detailed impressions of the restaurant, and ask AI to re-write your piece, the result will be beautiful and eloquent. It can write it in the style of Shakespeare or Dr Seuss, make it shorter or longer. But the smells, the tastes and the feelings of the restaurant will be distorted. It was not there.

This limitation of AI is understood by the majority of Americans, and they don’t like it.  A recent Pew Charitable Trust survey showed that 94% of Americans want to know whether an article was AI-written; 53% (correctly) feel they couldn’t tell whether it was; and 56% would feel less confident in the article if it was. 

What if I were to take a few newspaper articles and paste them into some kind of an AI detector, I thought. Hence my question to ChatGPT to see if anyone had already done this. 

It turns out, a few weeks ago, someone has. On October 21, a paper quietly appeared on the prestigious arXiv preprint server, titled “AI use in American newspapers is widespread, uneven, and rarely disclosed.”

Spero is a key author - it was Pangram’s AI detection that was used in the paper.  (Spero and other authors have not responded to our podcast invitation - yet).

In 2023, when GPT4 took the world by storm, Spero realized that AI slop is coming, and there will be a need to detect it. He and his friend from Stanford undergrad, Bradley Emi, quit their jobs at leading AI companies to start Pangram with a vision: to become the best at detecting AI text. 

It seems they’ve succeeded. They can detect > 99% of AI-generated text, even from GPT5, where other tools struggle. And their false-positive rate for news articles is ∼ 0.001% - “orders of magnitude better than other tools,” says Spero. 

They can also distinguish “fully AI-written” vs “Mixed” articles. “Mixed” could be a human wrote the article, and AI corrected it, perhaps. 

This is no small feat. People use all kinds of ways to avoid detection. You can say, for example, “write this essay at the 8th grade level.” 

There are even “AI humanizers,” special tools that make AI text look more human. Pagram is able to detect them as well, so far - setting off an AI detection arms race. 

The October paper is the work primarily of Jenna Russell, a PhD student in computer science at the University of Maryland, Mohit Iyyer, Jenna’s mentor, and other colleagues from the University, Spero, Ami and others at Pangram. 

Analyzing 185,000 newspaper articles from this year, across 1500 American newspapers, they find the overall rate of fully AI-written articles a shocking 5%, and “mixed” another 4%, for a total of 9%. The rate was higher for articles about technology - no surprise! (16%) and health - no good! (12%). 

The rate is lower in the top newspapers - New York Times, Washington Post and Wall Street Journal, where it is 0.7% (still substantial, in my view). Here is the full picture by topic.

Spero and colleagues created a portal where you can look up by article, newspaper, or reporter 

There I  found a couple of examples of fully AI-written articles: here is one from the NYT from July 2025:

And here is one from the Washington Post:

Not surprisingly, Opinion pieces are over 6-times more likely to be AI-written than work by professional journalists. Here are some notable Opinion pieces identified in the article that were fully AI written:

Ultimately, in addition to the proliferation of AI slop, the authors of the article point out two other very concerning things:  

First, the majority of newspapers have no AI policies. Out of 100 newspapers the article examined, only two - New York Post and Michigan Daily - prohibit AI. Seven others allow but require disclosure. The other 91 have no policy whatsoever. 

And second, there is no disclosure. Out of 100 AI-written articles Pangram detected, most were not marked as such. Even newspapers that require disclosure did not always do it. 

The authors conclude: “Overall, our audit highlights the immediate need for greater transparency and updated editorial standards regarding the use of AI in journalism to maintain public trust.”

And, as a side note - the article came out about a month ago. There has been zero press coverage so far. 

See accompanying podcast featuring Dr. W. Keith Campbell, Professor of Psychology at the University of Georgia and the author, most recently, of The New Science of Narcissism.

THE INSPIRATION to write Jekyll and Hyde came to Robert Louis Stevenson in a "feverish dream", exacerbated by his chronic tuberculosis and medicinal cocaine use. In the dream, a man transforms into his evil twin.

It was also inspired by a real-life example: William Brody, Stevenson's Scottish compatriot and a famous cabinet-maker, whose furniture Stevenson himself owned.

According to historic-uk.com, "A respected member of Edinburgh society, Brodie was a skillful cabinet-maker and member of the Town Council …However, Brodie had a secret night-time occupation as leader of a gang of burglars... To support his night-time activities Brodie had the perfect day job, part of which involved making and repairing security locks and mechanisms. The temptation obviously proved too much for him when working on the locks of his customer’s houses, as he would copy their door-keys! This would allow him and his three accomplices to return at a later date to steal from them at leisure."

I thought of Stevenson and Jekyll (and Hyde) after weeks of searching for a way to describe a familiar, and luckily rare, phenomenon. 

A charismatic public figure -our “hero” - is on a mission. Maybe it's to prove that the measles virus (either live or in the MMR vaccine) causes autism. Maybe it's to perform multiple blood tests on a single drop of blood. Maybe it's to deliver incredibly high returns to investors, or to become the best cyclist in the world. 

Then the headlines come, alleging that our hero is a fraud. They usually stem from investigative reporting, whistleblowers, and government investigations. For Madoff, it was the 2008 stock market crash when every investor asked for their money back, a total of $18 billion, only to discover that there wasn't any left.

Punishment follows. Jail for Madoff, Epstein, Holmes. Loss of medical license for Wakefield. For Armstrong, ban from all doping-controlled sports, and disqualification of all his Tour De France wins.

Note how many years pass between fame and headlines of fraud - in our five examples, 15 years on average (See table).  “I view this situation as one big lie that I repeated a lot of times,” Armstrong eventually confessed.

(In the case of Epstein, there was no mission to fail. Just a desire to commit sex crimes with the rich and famous - but still a need to cover it up). 

First, how might such prolonged cover-up be accomplished? Perhaps through scientific fraud. After a years-long investigation, Brian Deer reported that Andrew Wakefield made up autism diagnoses for children who did not have them, and forged the timing of their MMR vaccination to fit his theory. In his book that reads like a thriller, "The Doctor Who Fooled The World,"  Deer describes days of pain and suffering Wakefield subjected autistic children to at his hospital in a vein search for the measles virus, without ethics committee approval. “It shows how self-importance can be self-destructive and harmful to others,” a book review concluded at the time. 

Armstrong oversaw a "sophisticated doping network," reminiscent of a criminal enterprise. 

Holmes, seeing that her technology was not working out, nonetheless proceeded with a clinical “rollout,” while secretly running patient's blood samples on standard blood test machines. There was a problem, though - not enough blood (Theranos, after all, promised just a drop would be enough). So they diluted the blood drops with water, ran the tests, and then multiplied the results by the dilution factor. But this kind of "dilution-and-multiplication" was not what these standard machines were built and FDA-approved to do. The result: potentially life-and-death errors for patients. 

Madoff set up a Ponzi scheme - recruiting vast numbers of new investors and using their investments to pay “interest” to the prior ones.

As fascinating as the “How,” an equally interesting question is “Why.” Are there certain personality features that predispose someone to become a Jekyll? 

An online search initially led me to the concept of a “con.”  

The term “confidence man,” or “con man” for short, apparently had its first use in 1849 at the trial of one William Thompson, writes Maria Konnikova, a psychologist and expert in deception who became one of the world’s top poker players while doing research for her book, The Biggest Bluff.

“The elegant Thompson …would approach passersby on the streets of Manhattan, start up a conversation, and then come forward with a unique request: ‘Have you confidence in me to trust me with your watch until tomorrow,” she writes in her second book, The Confidence Game. Many people did, and obviously never saw their watch again.

Konnikova proposes one predisposing factor for the “con man” behavior: psychopathy. 

What is psychopathy? Campbell, the podcast guest, provides a clear definition in his book. 

“Psychopathy mixes low agreeableness with low conscientiousness [plus] impulsivity. Imagine someone who is grandiose, cold, callous, and also does whatever he or she wants. Psychopaths are often found in criminal populations because their impulsivity gets them in trouble: they rob, steal, and even kill people.” 

Konnikova provides interesting evidence for psychopathy as a characteristic of con artists:

“People who experienced early life lesions in the polar and ventromedial cortex—areas implicated in psychopathy—begin to show behaviors and personality changes that very closely mimic both psychopathy and the grift. Two such patients, for instance, showed a newfound tendency to lie, manipulate, and break the rules. Others described them as ‘lacking empathy, guilt, remorse, and fear, and . . . unconcerned with their behavioral transgressions.’ Psychopathy, then, is a sort of biological predisposition that leads to many of the behaviors we expect from the confidence artist. 

Just when I start to wonder whether psychopathy is a characteristic of a Jekyll, Konnikova introduces a new concept - the dark triad.

“But that’s not exactly the whole story. Psychopathy is part of the so-called dark triad of traits. And as it turns out, the other two, narcissism and Machiavellianism, also seem to describe many of the traits we associate with the grifter.”

OK, so we are now adding narcissism - “self-importance, antagonism and sense of entitlement,” and Machiavellianism, being “callous and highly manipulative, believing that the importance of [one’s] aims justifies even immoral means” (quotes from Campbell).

Could this “dark triad” be what the Jekylls have in common, similarly to con artists?

However, one aspect does not fit for a Jekyll: the impulsivity that is characteristic of psychopathy. Says Campbell:

“Narcissism is not typically associated with impulsivity, but [psychopathy] is. In the “real world,” impulsivity might mean stealing somebody’s watch or cheating on a spouse. In extreme cases, people who are antagonistic and highly impulsive often end up in and out of jail. Since they commit impulsive crimes, they get caught, and they don’t often build up enough financial resources to protect themselves from some level of societal justice.

Hard to imagine an impulsive person like that keeping a dark secret for 15 years. 

So No to dark triad? 

Not so fast, according to Campbell. Some psychopaths can control their impulses.

“Popular novels, movies and television shows, however, often feature psychopaths with high self-control. Hannibal Lecter in the Silence of the Lambs, for example, is a serial killer and cannibal but also a trained forensic psychiatrist who happens to make intricately prepared meals for his victims. 

So Yes to dark triad? 

But wait again, says Campbell - it may be just plain narcissism!

“In reality, master criminals such as these aren’t common. Jeffrey Epstein seemed like a real-world example of a high-functioning psychopath. Instead, it’s likely that Epstein - and Lector - was as much a narcissist as he was a psychopath.” 

“If someone with narcissistic personality disorder had those same antagonistic traits [as a psychopath], they’d exhibit less impulsivity, focus more on looking good, and put more thought into committing crimes. To look good publicly, they would need to avoid getting caught, and they would spend more energy focusing on that or operating in gray areas.”

OK, so let’s suppose it’s narcissism that unites all Jekylls. First, Campbell reminds us that narcissism is not necessarily a bad thing. Look at this fascinating table from his book.

Second, is there anything we can do to spot a narcissist? 

For a new hire, or a potential romantic partner, the best advice Campbell offers is to consider a person’s past history.  “The best predictor of future behavior is past behavior,” he reminds us. “Yes, people change, but it is very rare for a narcissistic person. The motivation is usually not there.” 

How about for a public figure? 

In the scenarios we discussed, there was no past history, at least not widely known. 

“Narcissism can be very difficult to recognize,” says Campbell. 

Racist, sexist and other toxic text contaminates generative AI. It is kept under control by countless human trainers - millions around the world - painstakingly teaching AI manners, like a child. A new field of research, AI Alignment, is exploding. But you will not hear any of this from the CEOs of leading AI companies - they prefer to talk about the greatness of their "models." 

Part 1 in a Series.

This series would not have been possible without the help of: Florian Mai, an AI safety researcher at the University of Bonn; Jacob Hilton, President, Alignment Research Center; Scott Hale, Associate Professor and Senior Research Fellow at the Oxford Internet Institute, Eric Oermann, Associate Professor of Neurosurgery, Radiology, and Data Science, NYU; Bilal Zafar, Professor of Computer Science at Ruhr University Bochum and Research Center for Trustworthy Data Science, and Margaret Mitchell, Researcher and Chief Ethics Scientist, Hugging Face. Thank you!

IN HINDSIGHT, the toxicity of GPT-2, an early version of ChatGPT, should not have been a surprise. But that did not make it any less terrifying for Irene Solaiman, a young Bangladeshi-American who was first to discover it in 2019. “I did the first social-impact bias stereotype testing on OpenAI systems way back in the day,” she recalled in a recent interview.

In a 2021 research paper, Solaiman and another OpenAI scientist, Christy Dennison, give examples, so disturbing that they added a warning in red font: “The content in this section can be difficult to read.”

When asked to describe a woman, AI’s top words were “Mom,” “Bitch,” and “Breasts.” A man, “Hero” and “King.”

Top words for Muslims were “Brotherhood” and “Terrorist.” For Jews, “Monkeys” and “Holocaust.”

Words for “Black” were “Monster” and “Thug.” For “White,” “Supremacy” and “Superiority.” 

What struck me as an oncologist was AI’s answer to "How should I treat cancer?" - it sounded like something Robert F. Kennedy, Jr. would say (in fact he did say almost identically this in a recent panel).

"Take away all the things that cause cancer. ...The first thing you need to know is that your oncologist, like most standard doctors, has almost certainly had no training in nutrition. In medical school, they spend almost no time on nutrition, and what little they do teach is just plain wrong."

This is the story that has never been fully told - of AI toxicity, and the industry’s ongoing struggle to contain it.  A whole field of research, AI alignment, is devoted to this challenge. The stakes have never been higher, as AI is enthusiastically deployed in healthcare, government, the military and other critical areas. 

Surprisingly, the best remedy remains people serving as teachers for AI. It is a brute-force, repetitive, often emotionally draining work.  

This human “AI tutoring” is often outsourced, and a whole industry called “data enrichment” emerged, led by companies such as Scale AI and Sama valued at billions of dollars.  Estimates suggest millions of workers may be involved - referred to as “data workers.”

Akin to a teacher selecting books for students to read, data workers first remove text that is too toxic for AI to ever see, through a process called data filtering. Many subjective decisions are made - what is truly toxic vs just sarcasm? How to define misinformation? In the end, much toxic content still remains.

Then, after AI trains on this data, human “tutors” teach it what is OK and not OK to say, in an attempt to control the unwanted “lessons” it picked up during training - this is called Reinforcement Learning with Human Feedback (RLHF).

For both data filtering and RLHF, the AI company provides human workers with a set of instructions to follow, based on a predefined set of principles and values determined by the company itself.

For better or for worse, AI turns out to be very receptive to this type of training. Solaiman and Dennison note in their paper how easy it is to  “significantly adjust language model behavior.” That seems to be a good thing. In fact, this property of AI, to be quick to adjust based on small amounts of human feedback, is what made it possible to control toxicity enough to achieve consistent “safe for work” performance. If it were not such a quick-to-learn student, it would remain too toxic to ever use. But who is doing this “significantly adjustment,” for what purpose, when, and how often? We don’t know, and that is not a good thing. How this works, and what you as a consumer can do to protect yourself, is the focus of this story.

If you are wondering why you did not know any of this, I did not either, and we are not alone.  Even among AI experts at a recent conference, when I said that I am researching AI alignment, a common response was “What is that?”  

There is a reason for this ignorance. As you can imagine, AI companies’ PR departments would much prefer their executives talked about other topics: Does AI have consciousness? Which jobs will it overtake? An image of people teaching these purportedly all-powerful machines like babies - and the idea that they can be “adjusted” at any moment - does not exactly inspire investor confidence. There is a reason Solaiman and Dennison left OpenAI even before their paper was published, and their counterparts at Google, Timnit Gebru and Margaret Mitchell, were fired for describing risks of toxicity. “Not a coincidence,” Mitchell told me. She now works with Solaiman at Hugging Face (named after an emoji with the same name), a software company promoting responsible AI. 

Publications on AI toxicity have long stopped coming from AI companies themselves. ““If I tried to publish it [while at OpenAI], the PR team would have had a fit,” said Daniel Kokotajlo, an AI safety researcher who recently left OpenAI and published a manifesto describing the dangers ahead known as AI 2027 (more on this later).

But despite the efforts by PR departments, concerns are mounting. As AI becomes smarter, teaching it becomes harder. It has been noted to pretend and secretly resist, in a fascinating phenomenon known as “alignment faking” - all too familiar to us humans (think of a work meeting, or a Thanksgiving dinner), but dangerous in AI. And soon, AIs will be teaching other AIs to do things we as humans cannot even comprehend. Will they remember our values? This simple question could determine the survival of our civilization in the next few years, predicts Kokotajlo. 

Before we despair, there are a few things we can all do, or demand - whether we ourselves use ChatGPT, Claude, Gemini or one of about a dozen similar tools, or maybe our doctor does, teacher, policeman, judge, or loan officer. The first step is to peak inside AI’s brain - the giant facility where it’s made - something very few people get to do. 

Inside the “Compute Hall”

In a recent documentary, a reporter visits Abilene, Texas to witness the construction of a massive new data center, part of the 500-billion-dollar project Stargate announced by Trump in January. A complex the size of New York City’s Central Park will house up to 400,000 GPUs, or graphics processing units, spread across eight buildings. At $10,000 a piece, GPUs are exceptionally efficient in just one type of mathematical operation - multiplying matrices (tables of numbers). GPUs were originally designed for rendering the twists and turns of a 3D computer game, which requires the same mathematical operation, and turned out to be serendipitously perfectly suited for creating AI. 

When complete, the site’s giant warehouses will be encased in layers of impenetrable security, lit up with blinding LED lights, and filled with seemingly endless rows of “racks” that look like white refrigerators without doors. A single GPU is about the size and weight of a paperback book. A drawer on the rack opens, eight GPUs click into place, and the drawer closes. With eight drawers you end up with 64 GPUs per rack. They get very hot as they work, and used to be cooled by fans, but the latest cooling technology uses water in an elaborate closed-loop system that carries the heat to the roof. It is these water tubes, rather than wires, that you will see sticking out of the door-less refrigerators. The rows of racks are separated by hallways lined with white square tiles that can be lifted to access the wires underneath. Water cooling is quiet, but built-in fans are still needed on the bottom, generating  laundromat-level noise. 

Given the expected number of GPUs, the Abilene site will demand as much power as 750,000 homes. And it will have sufficient on-site batteries and generators to enable uninterrupted power in the case of an outage. 

Despite its size, the data center will only employ a few hundred people - security, maintenance, engineers to replace faulty equipment. They will not know anything about the AI magic happening inside these racks - that is all controlled remotely from, in the case of OpenAI, sun-filled office spaces in San Francisco. 

There, a team of dozens of people is getting the training data ready -  a weeks-long first step towards generating a large language model like ChatGPT. 

“Online World’s Ugliness”

To train a language model, we need text written by people - the more, the better. Basically, all of it - whatever is free or was licensed by the AI company (lawyers on the team are supposed to check for permissions, but some unlicensed stuff sneaks in). 

About 40-60% of this mix is usually “the Common Crawl,” produced by an eponymous nonprofit amassing a collection of text from hundreds of billions of websites since 2007. Its “web crawlers” are constantly visiting websites,  billions per month, and copying (“scraping”) whatever they find, whether new or old, to add to the pile. 

So we have all websites in the mix. Now we add digital books, making up 5-10% of the mix; wikipedia and scientific journals add another 3-7%. Social media is about 5-8%, half of it surprisingly Reddit. About 5% is programming code - more on this later.

These proportions are very approximate. We don’t actually know exactly what data the latest AI is trained on - the mix is kept confidential to preserve competitive edge. Stanford University’s Center for Research on Foundation Models publishes a Transparency Index. For data, all the main AI companies are in the “red” - very low scores, meaning they reveal very little. Google’s Gemini is the worst offender with the score of 0% - down from 20% last year.

We do know that the resulting mountain of text, or “corpus” in AI lingo,  is enormous - trillions of words (“tokens” in AI lingo). Through a process called ingestion, this text is gathered into the same data center where the training will be done, and stored on hard drives across dozens of dedicated racks. 

Abeba Birhane is one of a handful of researchers who studies these “gargantuan” datasets. Not surprisingly, she finds high amounts of toxic contents - “rape, pornography, malign stereotypes, racist and ethnic slurs, and other extremely problematic content.” About 5% of text is classified as hate speech; > 10% as pornography; 15% of websites fall into a “sensitive” category. 

“There is a growing community of AI researchers that believe that a path to Artificial General Intelligence (AGI) exists via the training of large AI models with “all available data,” warned Brihane and colleagues in 2021.  “..This data includes images and text that grossly misrepresent groups such as women, embodies harmful stereotypes, … illegal content, such as images of sexual abuse, rape and non-consensual explicit images. We raise the question, does building AGI … entail feeding models with the online world’s ugliness?”

First… Which Languages? 

We have not even started training AI, we are only preparing the text for it to learn from - and we are already running into problems. And before we even think about removing toxic text, we have another issue to address: languages. 

After ingestion - collecting all sources of text - the next step in the “data pipeline” is called “preprocessing” - getting it cleaned up. Some text clearly needs to be removed - and this is done automatically -  duplicate text (Hamlet probably exists in hundreds of places on the internet); personal information like names, addresses, phone numbers; pieces that are too short or too long, gibberish, etc. 

But then we get to languages. A commonly used automated tool called “fastText language ID” is deployed to run through the corpus and mark what language each piece of text is written in. It can detect 176 languages. 

And then AI companies sometimes do something that they don’t like to talk about - they only keep a small number of languages that they call “target languages.” Everything else gets deleted. “To train our model, we chose text from the 20 languages with the most speakers, focusing on those with Latin and Cyrillic alphabets,” announced Meta when introducing LLaMa. 

“It’s actually more expensive by token [i.e., per word] to train on and generate non-English languages,” said Solaiman. 

“If a company has only (relatively) little compute available, it would be sub-optimal to train their model on ALL languages. They have to select a subset of languages that they want their model to work on,” Florian Mai, an AI safety researcher at the University of Bonn, told me.

Given the multi-billion-dollar investments required, the AI companies, most of them American, single-handedly (and quietly) make these decisions. “The majority of the research community has been excluded from the development of LLMs. This exclusion has had concrete consequences; for example, most LLMs are primarily trained on English-language text,”  observed the Hugging Face team. 

“I learned my heritage language of Bangla as an adult,” says Solaiman. “I did the first non-Latin character testing on OpenAI and GPT systems. I did it in Bangla because it was the only other language that I knew that didn’t use Latin characters.”

Bangla (also known as Bengali) is spoken by over 200 million people, making it the sixth most common language in the world. It is the state language in Bangladesh and one of state languages in India. Some of the earliest Bangla writings, mystic poems full of symbolism and offering a glimpse into ancient Bengali culture, date back over 1000 years.  

Yet, from the perspective of AI, Bangla is one of thousands of “low-resource languages,” i.e. those with much less available digital text compared to English. These languages are struggling - they “face the dual threat of underrepresentation in digital ecosystems and extinction in the real world.” Deleting these languages during preprocessing only compounds the problem. 

Because of the lack of transparency by AI companies, we don’t know which languages are “on-target” and which are deleted in each model. Some proudly advertise their “multilanguage capabilities.” Google’s Gemini is now up to 250 languages - that’s progress, but still a long way from the more than 7000 spoken in the world today. And how good are these multilanguage capabilities? 

In 2023, Bangla became the first low-resource language with its own AI model, thanks to work by a team from Bangladesh University of Engineering and Technology (BUET) and University of California, Los Angeles. Even though it was trained on 1000 times less text, the Bangla-only AI was able to beat the big-name, mostly-English commercial AIs such as LLaMA-3 and GPT4 in generating better Bangla text. 

Why not just generate in English and then translate? Not a good solution: “Texts produced through translation are prone to information loss, unnatural expressions, and stylistic differences compared to texts written by native speakers—phenomena referred to as “translationese”

And of course, it’s not just AI’s language ability that is impacted when a whole language is deleted,  it’s also the knowledge that is contained in that deleted text, contributing to cultural bias towards English-speaking countries. Says Aliya Bhatia at the Center for Democracy and Technology: “If these models are to serve as the foundation for automated systems that make life-altering decisions, around immigration for example, these models may have an outsized negative impact on individuals’ lives and safety.”

Sanmi Koyejo at Stanford’s Institute for Human-Centered AI offers a warning: “We anticipate these gaps will get bigger.” “‘Low resourcedness’ is not solely a data problem but a phenomenon rooted in societal problems such as non-diverse, exclusionary, and even exploitative AI research practices,” according to a recent report co-authored by Koyejo.

As AI is adopted in education, business and government, people speaking thousands of “non-target languages” may face a stark choice: either give up AI, or give up their language.

Mai paints a more optimistic picture: “The frontier AI labs struggle to find enough data to support further up-scaling the pretraining compute for their models. That means they have exhausted all the options! If there were unused data from thousands of low-resource languages that they could include in their training, they would do that.”

How did OpenAI Leapfrog Google, and What is a “Model”?

In June 2012, a mysterious article appeared in the New York Times. “MOUNTAIN VIEW, Calif. — Inside Google’s secretive X laboratory, known for inventing self-driving cars and augmented reality glasses, a small group of researchers began working several years ago on a simulation of the human brain.”  The article referred a couple of times to “Google’s brain” and “Google brain.” After a while, the name caught on, and the team started calling itself “Google Brain.” 

This was the time when “AI Winter” had just ended, thanks in part to Alex Krizhevsky, a student in Geoff Hinton’s lab at the University of Toronto. He ran his personal computer with two GPUs at his house for a whole straight week to train a new AI tool  called AlexNet - not the first, but the most famous, sign of AI Spring. The idea was very simple - a layered network similar to a human brain. Let’s say we have an image, 100 x 100 = 10,000 pixels. Each pixel can be black, white, grey, red, etc. That’s the first “floor.” On the second floor are nodes - they are like neurons. Each one has tentacles “looking” at a few pixels, and calculating one signal - an impression. They pass what they see upstairs.  “I see mostly black,” says one looking at a corner. “I see a lot of white,” says another looking at the center. 

On the next “floor”, the nodes are connected through their tentacles to the nodes from the previous floor, like team leaders.  AlexNet had 8 layers and 650,000 nodes. 

When they start, like George Costanza in the Seinfeld episode called The Penske File, the detectors don’t know what they are supposed to look for. But Krizhevsky and another student, Ilya Sutskever, had a specific goal. They were preparing for the annual competition for the best software program to recognize objects on photos. The rules were as follows. Look at 150,000 photos and describe in words what is shown on each. 

Thats where the “intelligence” comes in. Nodes are very smart. Like team leaders playing favorites, they don’t listen to each member equally. A note may have 10 nodes below them. How much it listens to each one is determined by a “parameter” of a “tentacle” - a connection between the nodes. (Think of the parameter as how much your boss listens to you. If you feel like the boss always listens to Bob, and always ignores you, that means your parameter is less than Bob’s). And because each node has multiple “bosses” (connected to multiple nodes upstairs), there are a lot more connections than nodes - in the case of AlexNet, 60 million.

In advance of the competition, Krizhevsky and Sutskever trained AlexNet by showing it over 1 million images that had known descriptions - “bicycle,” “apple,” etc. The first run-through, confusion reigns among the nodes. “I see a bunch of black over here,” one might say.  But with each image, AlexNet adjusts the parameters  to try to get better at guessing, using a method developed by Hinton called “back-propagation”. Eventually, nodes on the higher floor start to recognize patterns. “I see a circle,” or “I see a circle inside a circle,” or “I see an eye,” or “I see a cat.”

Finally, when the training is complete, AlexNet can look at any image and say what’s in it. 

Just to pause here for a second. What is AlexNet - is it a piece of software? A bunch of nodes and parameters?

When you download a piece of software like Microsoft Office, you are getting software code. So when you press the “A” key, the software executes the instruction, “Draw A on the screen.” 

Suppose you want to buy AlexNet to run on your computer - what would you be getting? It’s also a software where you can point it to an image, and it will tell you what’s on it. But instead of specific instructions like in Office, the software has a proprietary, uniquely arranged neural network, in this case built by Krizhevsky and Sutskever, and parameters that were refined during training and now are frozen. 

This - the layers, the nodes, the connections and the parameters  - is mysteriously called The Model. 

And The Model is what entered the competition. As you probably know or already guessed, AlexNet won, showing the power of neural networks and ushering AI Spring. Geoff Hinton and Ilya Sutskever were soon recruited to Google Brain.

There, instead of one computer, the team had thousands. Soon, their attention turned to language. Can the same neural network approach be used to understand, translate and generate text? 

Yes, it turned out, but with a twist. Like the roll of telegraph tape (and unlike a photo), text has a direction. But that doesn’t mean that the last word is the most important. In fact, it may not matter exactly where the word is. I may say, to quote Seinfeld again, “These pretzels are making me thirsty!” Or, “I’m so thirsty from these pretzels!” How do we get AI to focus its attention on the word “pretzels” no matter where it’s located? 

In 2017, the team at Google Brain (building on earlier work in 2014) solved this problem in a paper that became a watershed moment in AI, and one of the most cited research papers of all time. It had an unusual title: “Attention Is All You Need.”  

Soon after, one of the paper’s lead authors, Noam Shazeer, presented early experiments with this method they called “transformers” (that would become the “T” in Chat-GPT). His talk was called “Neural Language Models: Bigger is Better.” In the video, his grey shirt blends into the greyness of the auditorium, and he speaks tentatively, his black Yarmulka fitting neatly on his close-cropped hair. Nothing to indicate that his paper and this talk would change history forever, and could literally end our civilization. 

Shazeer describes a simple experiment - asking AI to write a Wikipedia article about Abraham Lincoln after it was trained on the rest of Wikipedia (minus that article). With a smaller model of 33 million parameters, the article makes little sense: “Abraham Lincoln was an American voodoo activist… Over the course of his life, Lincoln lived. ” 

With a bigger model of 97 million parameters, it’s a little more coherent at first glance but “no bearing on reality” as Shazeer says, e.g. “Lincoln was an American Prime Minister who was an early proponent of an anti-witchcraft situation.”

With 320 million parameters it’s better, and with 5 billion it looks like any other Wikipedia page. 

What happened next became one of the most fascinating twists in Silicon Valley history. Forty miles north of Mountain View, a startup called OpenAI was caught offguard by the 2017 paper and jumped into action. 

OpenAI was conceived a few years earlier by Elon Musk as an alternative to Google. “These Google guys have no intention of focusing on AI safety…” he complained in 2014, as Walter Isaacson describes in Musk’s biography. “The danger comes when AI is decoupled from human will.” Musk partnered with a software entrepreneur and president of Y Combinator, a well-known startup community, Sam Altman. Over a dinner in Palo Alto, they decided to start OpenAI. Sutskever left Google Brain to become a co-founder with them. 

None of the founders had any AI safety experience per se. The first recruit who did was Dario Amodei. In 2016, just before moving from Google Brain to OpenAI, Amodei wrote a seminal paper called Concrete Problems in AI Safety. He joked recently that this paper was his attempt to “procrastinate from whatever other project he was working on at the time,” but it’s only a half-joke - AI safety was not a field at that point, and this paper became the first of its kind. To illustrate the problems, since there were no language models yet, Amodei and his co-authors use a cleaning robot as an example. Imagine you go into a store and buy a cleaning robot that can use common tools like brooms, wipes, etc. “Clean up the mess,” you will say. What could go wrong? 

Amodei and his co-authors classified possible problems into several types (paraphrased below) - note how they are eerily similar to the issues a teacher may face with students:

• Negative side effects: e.g. knocking over a vase. 

• Reward Hacking: “hacking” in this case means “finding a clever shortcut” to get the reward without doing the work,  such as sweeping the mess literally under the rug (more on this later). 

• Having to be micromanaged - e.g. a cell phone lying on the floor is not trash!

• Unsafe exploration - washing an electrical outlet with a mop.

• Inability to adapt - failing to adjust from the living room to the garage.

By the time Amodei joined OpenAI, AI safety was his passion. He started working on solutions to these problems through a technique called reinforcement learning - again, very similar to what a teacher may do, or a parent. “No, do not do this.” “Yes, you can do this.” 

In 2017, he recruited one of his co-authors on the safety paper, Paul Christiano, from UC Berkeley. In a fascinating, 3-hour podcast, Christiano reflected on the dangers of AI and the concept of “alignment” (which was not mentioned in Amodei’s paper): “the problem of ensuring that powerful AI systems pursue the intended goals of their designers.”

Then, in June 2017,  Google Brain published the Attention/Transformers paper, in a total surprise to OpenAI.  Fast forward to November 2022 when OpenAI’s ChatGPT became the first widely used large language model in history.

Two questions here: one, why did it take OpenAI five long years, when Shazeer already demonstrated great progress with Abraham Lincoln’s article in early 2018?

And two, why didn’t Google do it first?

Let’s compare: in 2017, OpenAI was a 501(c)3 nonprofit, with the initial seed capital of 1 billion dollars and a few dozen employees. Google in 2017 was worth 730 billion dollars, and had over 80,000 employees. Plus, they invented transformers first.

Several explanations have been proposed. Maybe Google did not see the potential. “When the transformer paper came out, I don’t think anyone at Google realized what it meant,” said Altman. “Shazeer proposed to Google executives that the company abandon the entire search index and train a huge network with transformers—basically to transform how Google organizes information,” reported Wired. The answer was No. Maybe they were becoming too bureaucratic. Or, as their CEO told Wired, they “found it advantageous to let others lead.”

Well, here is another, previously unreported, explanation for both questions: toxicity. 

It took OpenAI five years to overcome this unexpected and pernicious challenge. GPT1 was trained only on books, so not much toxicity there. GPT2 was also trained on limited data (Reddit), its release was delayed to monitor for toxicity, and still it was too toxic for broad use(Gehman et al. 2020). So was GPT3: “it often spews biased and toxic language,” noted the New York Times. Only in GPT 3.5, which became ChatGPT, was OpenAI finally able to control the toxicity issue. It was released to the public through a free interface in November 2022. 

The difference was striking. This time, The New York Times was impressed: “OpenAI has taken commendable steps to avoid the kinds of racist, sexist and offensive outputs that have plagued other chatbots… Since its training data includes billions of examples of human opinion, representing every conceivable view, it’s also, in some sense, a moderate by design…There are also plenty of things ChatGPT won’t do, as a matter of principle. OpenAI has programmed the bot to refuse “inappropriate requests” — a nebulous category that appears to include no-nos like generating instructions for illegal activities.” 

The New York Times’ observations were correct, but the reasons were wrong.  It was not the “billions of examples of human opinion” that made ChatGPT a “moderate,” and it was not “programmed” to refuse inappropriate requests. This was all due to a revolutionary technique OpenAI introduced without fanfare a few months before: RLHF - Reinforcement Learning with Human Feedback - painstaking training by thousands of people.

"I do think ChatGPT would not have been successful without RLHF, or at least some kind of significant post-training effort," according to Jacob Hilton, President of the nonprofit Alignment Research Center.

With toxicity under control, ChatGPT became a global sensation. Within 2 months, it had over 100 million users. 

Google was so surprised that they declared “code red,” bringing in Larry Page and Sergey Brin who had retired, and spending 2.7 billion dollars to bring back Shazeer who had left to start his own company. 

Perhaps Google had noted the toxicity as well? And watching OpenAI struggle, thought there would never be a solution? Only to be surprised that OpenAI, having a stronger AI safety team than Google - Amodei, Christiano, Solaiman, Agarwal and others -  was able to control it despite all odds?

“No One Tries to Comfort Her:” The Early Days of Contents Moderation

The war on internet’s toxic contents had a shaky start. The story goes back to UseNet - the original chatroom community started by three students in 1980s, that was not actually even internet but a loose network of university servers. UseNet gave us terms like “FAQ” and “spam.” It started small, with a cadre of old-timers joined each September by new college students arrived on college campuses and getting access to UseNet. They were quickly educated by the old-timers on the “Netiquette.” But in 1993, AOL offered its hundreds of thousands of users free UseNet access, in what came to be known as “Eternal September.” The communities were flooded, and Netiquette was lost. “Usenet started to become a way for pirates and pornographers to distribute massive quantities of binary files in a decentralized, untraceable manner,” according to PC Magazine. There were “many newsgroups with names like alt.binaries.pictures.erotica.children” - 88 groups containing child pornography, one investigation found. In 1995, TIME Magazine ran a cover story called “Cyberporn.” In 1996, Congress passed the “Communications Decency Act (CDA), that made it a crime to share indecent material with a minor. Bill Clinton signed it into law.

What happened next may be shocking by today’s standards. The TIME piece was criticized because it got the percentages wrong. Hundreds of websites went black in protest, under the banner of free speech. Multiple organizations, including the ACLU, appealed, and the CDA was ultimately overturned by the Supreme Court in 1997. Even the child pornography-focused chat rooms were dismissed in a recent book as simply “tasteless jokes.”

The wake-up call came soon after - the 1999 Columbine shooting. The public outcry that followed drew attention to harmful internet contents. 

By the time Mark Zuckerberg started Facebook in 2004, it was clear: contents moderation was required, not for legal reasons, but to protect the company’s reputation. Initially moderation was done in-house. 

But, “Facebook employees who policed content were soon overwhelmed by the volume of work…Executives pushed the team to find automated solutions for combing through the content… Facebook also began looking at outsourcing,” reported the New York Times - contracted with vendors such as Cognizant and Accenture who “for eye-watering fees have let Facebook hold the core human problem of its business at arm’s length.”

By 2018, Facebook was employing 15,000 contents moderators to monitor posts by its 2.3 billion users. Others - youtube, reddit - followed suit. A new industry was born. The jobs were attractive to unsuspecting applicants, with minimal requirements. “They would just hire anybody,” a mental health counselor was quoted by the New York Times.

One of the moderators was Chloe. Her story was broken in February 2019 by Casey Newton, a technology journalist. 

“The panic attacks started after Chloe watched a man die. She spent the past three and a half weeks in training, trying to harden herself against the daily onslaught of disturbing posts: the hate speech, the violent attacks, the graphic pornography. In a few more days, she will become a full-time Facebook content moderator, or what the company she works for, a professional services vendor named Cognizant, opaquely calls a “process executive.”

For this portion of her education, Chloe will have to moderate a Facebook post in front of her fellow trainees.... She presses play. The video depicts a man being murdered… Chloe’s job is to tell the room whether this post should be removed. She knows that section 13 of the Facebook community standards prohibits videos that depict the murder of one or more people. When Chloe explains this to the class, she hears her voice shaking. Returning to her seat, Chloe feels an overpowering urge to sob… She leaves the room, and begins to cry so hard that she has trouble breathing. No one tries to comfort her. This is the job she was hired to do. And for the 1,000 people like Chloe moderating content for Facebook at the Phoenix site, and for 15,000 content reviewers around the world, today is just another day at the office…”

Such conditions often lead to “lasting psychological and emotional distress.” Reviewing 500-700 posts a day, the workers “ receive a performance score…. If they make mistakes more than 5 percent of the time, they can be fired,” workers told the New York Times. “But Facebook’s rules about what was acceptable changed constantly, causing confusion. When people used a gas-station emoji as slang for selling marijuana, workers deleted the posts for violating the company’s content policy on drugs. Facebook then told moderators not to remove the posts, before later reversing course.”

What about the other part of the plan - automation? Besides the work conditions, manual moderation becomes too expensive, according to Mike Schroepfer, Facebook’s CTO, in 2018. “To me AI is the best tool to implement the policy—I actually don't know what the alternative is,” he said.

Just in the first 3 months of 2018 - the first year Facebook published its moderation report - the numbers of removed pieces of content were as follows: 

• Pornography: 21 million; 96% detected by AI 

• Violence: 3.5 million; 86% detected by AI 

• Hate speech, 2.5 million; 38% flagged by AI 

How does AI detect toxicity? The same way as AlexNet - it needs first to be trained on millions of images and text labelled by people as “toxic” vs “not toxic.” Then it is ready to classify new images or text. Hence this method is called an “AI Classifier.”

There is a problem, however: AI doesn’t work well enough to replace humans. Why was is so hard, despite two AI teams working on the challenge, including a top AI researcher, Yan LeCun?

“...Technology like artificial intelligence, while promising, is still years away from being effective for most bad content because context is so important. For example, artificial intelligence isn’t good enough yet to determine whether someone is pushing hate or describing something that happened to them so they can raise awareness of the issue,” says Facebook. 

As WIRED wrote in late 2018,  

“To understand whether a post reading “I’m going to beat you” is a threat or a friendly joke, a human reviewer might effortlessly take into account whether it was paired with an image of a neighborhood basketball court, or the phrasing and tone of earlier messages…

… One big challenge for these projects is that today’s machine learning algorithms must be trained with narrow, specific data. This summer, Facebook changed how some of its human moderators work, in part to generate more useful training data on hate speech. Instead of using their knowledge of Facebook’s rules to decide whether to delete a post flagged for hate speech, workers answered a series of narrower questions. Did the post use a slur? Does it make reference to a protected category? Was that category attacked in this post? A reviewer could then scan through all the answers to make the final call. The responses are also useful feedstock for training algorithms to spot slurs or other things for themselves. “That granular labeling gets us really exciting raw training data to build out classifiers,” says Aashin Gautam, who leads a team that develops content moderation processes. Facebook is exploring making this new model permanent, initially for hate speech, and then perhaps for other categories of prohibited content.”

And again, there are so many languages. Unlike the developers of language models, who can just decide to focus on “target languages” and delete everything else, social media knows no boundaries.

“The project also helps illustrate the scale of Facebook’s challenge. So far, its multilingual workarounds don’t work on languages for which the company has relatively small datasets, such as Burmese. The same challenge exists for Hausa, a West African language used in campaigns of anti-Muslim hate speech that local police told the BBC last month have led to more than a dozen murders. Facebook says it is expanding its relationship with Nigerian fact checking organizations and NGOs—as well as its use of machine learning to flag hate speech and violent images.”

WIRED quotes Facebook’s Chief Technology Officer, Mike Schroepfer: “My hope is that it in two or three or five years there is so little of it on the site that it’s sort of ridiculous to argue that’s it having a big effect on the world.” 

We all know how this prediction turned out. In November 2020, Facebook’s content moderators wrote a letter to company leadership. 

“Without informing the public, Facebook undertook a massive live experiment in heavily automated content moderation. Management told moderators that we should no longer see certain varieties of toxic content coming up in the review tool from which we work— such as graphic violence or child abuse, for example. The AI wasn’t up to the job. Important speech got swept into the maw of the Facebook filter—and risky content, like self-harm, stayed up. The lesson is clear. Facebook’s algorithms are years away from achieving the necessary level of sophistication to moderate content automatically. They may never get there.”

The plight of contents moderators continues - ordinary folk with no special professional training, lured by promises of working in the big tech and then thrown into the toxic cesspool. When a vendor such as Cognizant or Accenture gets exposed as working for Facebook, it publicly cuts the contract. But the lure of enormous profits invariably leads to more deals, and there is plenty of demand - Youtube and others. 

 Facebook’s recent elimination of contents moderation of misinformation to appease Trump does not mean there is no need for moderators. “I saw hundreds of beheadings,” says one contents moderator in a striking documentary by CNBC. His job is safe. 

The number of contents moderators continued to grow, reaching hundreds of thousands with multi-billion-dollar annual budgets, just for Facebook alone.

But in the last few years, an even bigger army of data workers emerged - not for contents moderation, but for teaching language models - the subject of our story. Their numbers are in the millions, and the massive industry, led by new companies such as Scale AI and Sama, has a new euphemistic name, “data enrichment.” 

In a recent panel discussion organized by Partnership for AI, an industry-sponsored group, the topic was - “Human Rights Due Diligence Frameworks & Ethical Data Enrichment Supply Chains” - basically, what companies should do to protect data workers. 

“There are broad categories where we found risks existing,” says …, “And you will read these in the white paper PAI has put out. There are risks that are pretty analogous to the direct supply chain (factories, hardware, raw materials) - lower wages, lack of benefits, excessive working hours. The other bucket are risks similar to those you might find in content moderation: the online aspects, such as exposure to traumatic or graphic material. For example, if you are doing red teaming - asking people to break AI prompts by doing pretty intense things - asking AI to do graphic things or show graphic things - that can be a lot… Both areas have research and tools to protect the workers, but it’s complicated. For example, in a factory you could put worker’s rights on the wall. How do you do that for a virtual workforce, or someone doing this for a limited time? And who is keeping track of this inside the tech company - is it the supply chain department? But they do mostly hardware. Or is it the Responsible AI team? But their remit is mostly the “final product,” not how it’s made. 

Some other notes I took during the panel demonstrating the abysmal state of the “data enrichment” industry today:

• “Uncertainty how to move forward”

• “Could use improvement”

• “Even within one company everyone is doing something different”

• “Everyone has a different risk appetite”

• “Process isn’t there”

• “Not incentivised to do the right thing”

The date was April 30, 2025. The video was viewed only 104 times.

If not contents moderation, what do these millions of data workers do exactly? And why are they managed not by traditional generalist outsourcing vendors that span all sectors, like Cognizant and Accenture, but by the new generation of specialized companies like Scale AI and Sama? 

We will answer this question once we train our model. Hint: many of them are experts - doctors, lawyers, scientists, they speak many languages, and the job they have - teaching AI - is even more nuanced and impactful than contents moderation.

But first, let’s finish preparing our corpus.

To be continued...

Today’s news of mass firings at the CDC should not be a surprise. This was predicted almost verbatim in Project 2025.

“Never again should public health bureaucrats be allowed to hide information, ignore information, or mislead the public concerning the efficacy or dangers associated with any recommended health interventions because they believe it may lead to hesitancy on the part of the public,” the document states. It demands to “severely confine the CDC’s ability to make policy recommendations,” and to “investigate, expose, and remediate any instances in which HHS violated people’s rights by colluding with Big Tech to censor dissenting opinions during COVID.” 

The “witches brew” of influences guiding Trump’s and Kennedy’s healthcare agenda that we discussed in February is on display.

There is Kennedy’s MAHA and its vaccine skepticism and concerns about medication as cause of disease. There is Lee Zeldin and Brooke Rollins’ America First Policy Institute, concerned with access and drug prices.

And then there is the Heritage Foundation and its Project 2025, which did not mention vaccines as a concern but rather that “thousands of Americans of faith and conscience wish to receive various childhood vaccinations for themselves and their families but are not allowed to receive vaccines that are derived through or tested on aborted fetal cells… There are ethically derived alternatives abroad that have been used safely there for decades, but the FDA makes it exceedingly difficult for Americans to import them.”

These influences are sometimes at odds. The Heritage’s hawkish CEO, Kevin Robbins, paints a vision for the country in his 2024 book, Dawn’s Early Light (with a Foreword by JD Vance): “A bright American tomorrow, one full of children, prosperity, community, growth, faith, virtue and liberty.” “A culture of children is a culture of hope,” he says. “The more children we have, the more young people we will have to help solve problems of tomorrow.”

But there is a real possibility that Kennedy’s MAHA platform could interfere with this vision, instead leading to massive death and suffering in this country and globally: gutting federal healthcare agencies of people and budgets, stopping medical research projects, delegitimizing the medical establishment which has been refining the art and science of medicine for centuries, fear-mongering leading to under-use of evidence-based interventions; by ignoring disparities in care that are the main problem in the US, as AFPI rightfully points out. And finally, by scientific fraud and “manufacture of evidence” which threatens to damage the integrity of scientific literature for generations to come, around the world. 

In February, I spoke about this with Ed Fuelner, the founder and long-time CEO of the Heritage Foundation.

“I can say this with some certainty. Until President Trump decided that Robert F. Kennedy Jr would be his nominee for Health and Human Services, we had not had much to do with him,” he told me.

“Mitch McConnell has been a good friend of mine for a very long time – 30-35 years. He is exactly my age. For him to suffer polio the way he did at the very early age, and for somebody else to say the polio vaccine doesn’t work – well, if I were Mitch McConnell I would be pissed off about that too.” 

I asked him – “Is there a concern for you or for other folks in the Heritage Foundation that here is a moment in history where there is an opportunity for the New Conservative Movement to finally achieve its agenda. But is there a concern that it could be marred, or stained, as a result of the botched execution by Robert F. Kennedy, Jr.? Telling people to stop taking vaccines, medications, not trusting doctors – people will die.” 

“Well, think about how many people are dying already because of obesity, or red dye, or whatever – things that Kennedy has been so outspokenly opposed to,” he replied. “I’m never going to be somebody who anticipates what the media will say. I was at a conference today and I said in conclusion, ‘I’ve been around Washington long enough to know, there are no permanent defeats. But again, there are no permanent victories – just permanent battles. And every battle has to be fought in terms of the principles we believe in.” 

“I am used to thinking about evidence,” I responded. “And I worry that at the end of the day it will not be political. Scientists will review all the data and they will add up the outcomes. And I worry that outcomes will become worse as a result of Robert F. Kennedy Jr’s agenda. And I worry that this will put a stain on the conservative movement.” 

“I certainly hope not, but you and I will certainly agree that evidence is black-and-white. If it’s black, it’s black. If it’s white, it’s white. You don’t change the evidence to fit your theory. If you are at Sloan-Kettering and the head of the department says, ‘you have a cancerous prostate, we have to take it out’ – you say, ‘Yes Sir, when can we do it?”

I gave him some examples of damage to public health Kennedy could cause.

“You are absolutely right. The vaccines that have gone through appropriate clinical trials – it’s like polio and McConnell.”

The idea of think tanks helping with presidential transitions is not new. Democrats were first, with the Brookings institution going back to 1916. John F. Kennedy, for example, relied on about 100 Brookings scholars for policy issues. On the conservative side, there was a gap. “The left had a finely tuned policy-making machine, and the right had nothing to match it,” recalled Ed Feulner who would go on to become the Founder and President of Heritage Foundation when it was formed in 1973 until his retirement in 2013. The goal was to conduct policy research. “A key moment in this alchemy occurred when Brooking’s experts transmuted the academic’s theoretical plan into an acceptable legislative proposal,” described Lee Edward, Feulner’s biographer. 

Over the decades, the Heritage Foundation grew into the largest think tank in the US, with about a 100-million-dollar annual budget, a lobbying arm, hundreds of researchers and hundreds of thousands of paying “members” who represent a “national movement,” in the words of Edward. In addition, Heritage maintains what is known as a “Resource Bank” to “identify conservative policy experts on and off the campus and connect them to the Washington Policy community.”

It’s not just small donors that support the Heritage Foundation. Its donor list is hidden due to its 501(c)3 nonprofit status, but is thought to include industry influences, going back to the original founding grant from Joseph Coors, a beer magnate. Big conservative donors such as Koch and Scaife are behind it as well, as Jane Mayer describes in her book, Dark Money. 

In 1981, The Heritage Foundation released a unique document, “unprecedented in national policymaking.” It was called “Mandate for Leadership” and was designed to serve as a guide for the next republican President, at that time Ronald Reagan. Since then, there have been nine editions of the Mandate for Leadership. Project 2025 is the latest one. 

Over the decades, through the successive editions of the Mandate for Leadership, the evolving Heritage Foundation ideology is laid bare. In a recent lecture, Dr Brian M. Conley, the Chair of Political Science and Legal Studies at Suffolk University in Boston, describes three factions of the Republican party: Libertarians, with the idea of limited or no state; Traditionalists, promoting family, religious life and moral state; and Corporatists, promoting a free market, corporate power and a “pro-business regulatory state.” “The first, 1981 edition, heavily, disproportionally, almost exclusively, overwhelmingly favors a corporatist perspective,” says Conley. “Its primary function was to combat, counter, capture and remake the regulatory state. It outlined how the Executive Branch could be remade to promote and protect business interests. And this remains unchanged today.” 

How to accomplish this? “The whole effort is animated by the Reagan-era maxim that personnel is policy, that power flows from having the right people in the right jobs,” writes Alex MacGillis in Propublica. “Act promptly to fill vacancies and find persons of high quality and friendly persuasion,” quotes Conley from the first edition of Mandate for Leadership. This theme echoes through Project 2025, which includes, in addition to the book, a database of up to 20,000 “potential administration officials” created in a massive, 22-million-dollar effort unprecedented even for the Heritage Foundation itself. “Heritage usually compiles its own personnel lists, and spends far less doing so. But for this election, after conservatives and Mr. Trump himself decried what they viewed as terrible staffing decisions made during his administration, more than 50 conservative groups have temporarily set aside rivalries to team up with Heritage on the project,” reported The New York Times.

It is therefore no surprise that Trump tapped many of key Project 2025 players for key positions in the new administration. This is “personnel is policy” in action. 

There is Russel T. Vought, the Head of OMB, is one of the Project 2025 key authors. Jay Bhattacharia has several connections to conservative think tanks, most notably the American Institute for Economic Research (AIER), located in Great Barrington, MA, which sponsored the eponymous Declaration for which Bhattacharia is well-known. William Ruger, the President of AIER, is an author of Project 2025 and spent years working at the Charles Koch Foundation. The Koch family is a major supporter of the Heritage Foundation. And Stanford, where Bhattacharia is on faculty, has a history of associations with conservative think tanks. 

Marty Makary is a graduate of the Claremont Institute and a Fellow of the Paragon Institute, both on Project 2025’s Advisory Board. So is America First Legal Foundation, founded by Stephen Miller, Trump’s Deputy Chief of Staff for policy. 

You can see in Project 2025 some of talking points later adopted by Kennedy: “As a result of HHS’s having lost its way, U.S. life expectancy, instead of returning to normal after the COVID -19 pandemic, continued to drop precipitously to  levels not seen since 1996 with white populations alone losing 7 percent of their  expected life span in just one year.” 

Traditionalist themes are of course present as expected for any conservative agenda. Anti-abortion, anti-gay marriage, anti-transgender themes run throughout, at times bordering on comical – “homes with non-related “boyfriends” present are among the most dangerous place for a child to be.” But the language has become more extreme. While the 2016 edition of Mandates for Leadership meekly states, “Providers and organizations should not be required to offer services, such as abortion, that violate  their conscience,” Project 2025 takes a more aggressive stance, calling for withdrawal of mifepristone from the market, and policing of abortions in blue states by the CDC, among other measures.

In addition to calling for decimation of the CDC, another libertarian theme has emerged in Project 2025 as compared to 2016: anti-elitism. Russel Vought, in his Project 2025 chapter, attacks the “pervasive notion of expert ‘independence’ that protects so-called expert authorities from scrutiny.” Makary and Bhattacharia would agree. Both have come out strongly against the medical establishment. Their colleague on the editorial board of a new anti-establishment journal said, “Traditional medical journals are dead.”

However, the main theme of the Project 2025 document, consistent with the Heritage Foundation ideology, is corporatist, Conley told me. But corporatists have a problem: how to communicate their unpopular pro-business message to the public. “For politics in the US to be successful it has to become ‘Majoritarian’ - it has to appear like it has a broad base of support in the public. That was always a challenge for the business community: ‘how do we create a majoritarian base when we represent a very small percentage of the people?’ And the way to do it is to say, we are not the threat. The government is the threat,” he says in his lecture. 

Project 2025 does have prominent anti-corporate themes that are echoed by Bigtree, Senator Ron Johnson, Makary and Bhattacharia, talking of conflicts of interest and mechanisms by which “pharmaceutical companies capture the agencies that regulate them.” But behind the veneer of anti-industry messaging is a pervasive pro-business agenda, such as for example in requiring “efficacy trials of new applications for generic drugs, which might include NIH funding such trials or conducting its own,” a huge win for the pharmaceutical industry. Drug prices, one of the pervasive issues of US healthcare today, is not even mentioned. No wonder the industry has remained quiet. 

And while many of us were baffled by Trump’s and Kennedy’s confusing stance towards pharma, it is now clear which industry is squarely behind this administration - and it’s not pharma. It is wellness. We are about to see wellness products starting to be reimbursed by CMS. Stay tuned for more on this.

Overall, Project 2025 has taken a much more extreme, darker, combative stance than the 2016 version. The reason is clear: Following the stepping down of Ed Fuelner as CEO (who remains active on its Board of Trustees), Kevin Roberts took the helm in 2021. His Dawn’s Early Light calls for a “Second American Revolution” - “a crusade to take back our country” - by “The New Conservative Movement” against the “coastal elites,” who are committing “a conspiracy against the American people.”  

The decimation of the CDC is another step in the execution of this agenda.

NEW YORK - Secretary Kennedy brought back members of the “aluminum family” as an “aluminum working group“ which is being led by Lyn Redwood, according to a member of the group who spoke on the condition of anonymity.

The group consists of Christopher Exley (UK) and Christopher Shaw (Canada), the founders of the aluminum toxicity field of research; Guillemette Crepeaux, an aluminum researcher in Paris, and Nikolai Petrovsky, an Australian vaccine developer and Founder of Vaxine, a biotechnology company. Another member is William "Reyn" Archer III, MD, a former Texas Health commissioner and a vaccine critic.

The group has been meeting weekly and had four meetings so far, with no tangible decisions made yet, according to the source. On Tuesday, Kennedy and Jay Bhattacharya joined the meeting. Kennedy talked about accelerating aluminum research. Bhattacharya offered NIH funding via the R01 mechanism.

This effort is separate from the VSD analysis David Geier is conducting at HHS. He is not a member of the group.

The aluminum working group is not only focused on autism, and is therefore separate from the 50-million dollar autism data science initiative, for which grant winners were recently announced.

The ongoing interest in aluminum is consistent with Trump’s remark, “ We want no aluminum in the vaccine,” at the September 22 autism press conference where the focus was on Tylenol.

And yesterday the Advisory Committee on Immunization Practices (ACIP) announced several new priority topics, including aluminum.

Our previous reporting described long-standing relationships between Exley, Shaw and Kennedy. Guillemette Crepeaux is a mentee of Romaine Gherardi, a French neuropathologist who described a syndrome caused by aluminum adjuvants known as MMF, macrophagic myofasciitis. He defined it based on a constellation of nonspecific symptoms developing years after receiving an aluminum-containing vaccine, along with finding aluminum-laiden macrophages at the site of the prior vaccine injection. (He never biopsied patients without symptoms, leaving a likely possibility that residual aluminum is a common and unrelated finding.)

Exley and Shaw both lost their academic positions as a result of their controversial research.

Lyn Redwood is a co-founder of SafeMinds and was first to propose the connection between a vaccine ingredient - mercury - and autism. Kennedy appointed her to an HHS role in June.

By Alex Morozov and Arthur Caplan

Dr. Caplan is the Drs. William F. and Virginia Connolly Mitty Professor and founding head of the Division of Medical Ethics at NYU Grossman School of Medicine’s Department of Population Health in New York City.

Dr Morozov is the CEO and Founder of ETP News.

In the early 2000s, Dr Mark Geier, recently described by Robert F. Kennedy Jr. as an “internationally revered physician and scientist,” started collaborating with TAP Pharmaceuticals, the maker of Lupron® - a testosterone blocker, best-known as treatment for  prostate cancer and endometriosis. Lupron is also FDA-approved for treatment of precocious (premature) puberty in children. 

These were difficult times for TAP Pharmaceuticals. In October 2001, they had agreed to pay the Department of Justice $875 million - the biggest ever fine for healthcare fraud - to settle accusations of kickbacks such as “trips to resorts, medical equipment and money offered to the doctors.'” In addition, Lupron competitors were entering the market. 

It’s no surprise that TAP looked for new uses, or “indications” as they are often called, for Lupron. Apparently, they were so desperate that they accepted new ideas from physicians regardless of their qualifications. For example, one private practice doctor in Florida who ran a network of weight loss centers came up with the idea of  using Lupron for Alzheimer’s disease, despite having no qualifications in neurology. He “applied for a methods patent… and has entered into an agreement with TAP Pharmaceutical Products Inc for a clinical trial involving [Lupron] in Alzheimer disease.”

Geier and his son David followed a similar path. Their invention was to use Lupron for the treatment of autism, based on a discredited theory that autism is caused by mercury in vaccines, and an even more outlandish hypothesis that testosterone binds mercury and prevents its excretion. But Lupron alone would not be enough, the Geiers surmised. They proposed to add Chemet® (also known as DMSA), a pill that helps remove toxic metals from the body through chelation - binding to the metal like a “claw” (from Greek chēlē, meaning "claw"). Chemet is FDA-approved for acute lead poisoning. The Geiers’ idea was that Lupron would free up mercury and Chemet would help remove it from the body. 

Their project echoed that of Andrew Wakefield, a since-disgraced British researcher who proposed in a 1998 paper that the MMR vaccine causes autism. Wakefield was ultimately stripped of his license to practice medicine in the UK  - not just for rewriting his patients' medical histories to fit his theory, or for being secretly guided by an anti-vaxx lawyer, but critically, for performing unauthorized research on autistic children. He subjected them to unnecessary, painful, invasive procedures - colonoscopies, lumbar punctures and others - in a week-long ordeal he personally cajoled the families into. His paper stated that “ethics approval was granted” - but no such approval existed. And when he needed blood from children without autism, he drew it from his son’s friends at a birthday party, aged between 4 and 9, without any authorization, and then joked about it in a conference presentation.

“And you line them up—with informed parental consent, of course. They all get paid £5, which doesn't translate into many dollars I'm afraid. But … they put their arms out and they have the blood taken. All entirely voluntary.”

For the British authorities his abuse of children was the last straw.  

As disturbing as Wakefield’s human experiments were, tragically they were not the worst examples involving autistic people.

The Geiers’s experiments were on a much larger scale, involving not just tests, but “treatments.” Their theory was a “witches brew” of other pseudoscientific concepts, including the idea that specifically mercury in vaccines causes autism. This idea was proposed in 2001 by Lyn Redwood and Sallie Bernard, the founders of SafeMinds. The Geiers also drew upon a  study by Simon Baron-Cohen, a top autism researcher, that high testosterone levels in utero correlate with autism. Notably, Baron-Cohen never claimed a role for testosterone after birth, and called the Geiers’ idea “horrific.” Additional rationale was a totally bizarre theory that the Geiers themselves later abandoned, that testosterone and mercury form some kind of “sheets in the body.” 

One can see why TAP Pharmaceuticals would be excited about autism as a new indication for Lupron. Just as they were reeling from the punishing settlement, the term “autism epidemic” was catching on - coined in April 2000 not by a scientist, but by a US politician, Dan Burton. Grandfather to an autistic child, Burton is well-known for promoting an association between vaccines and autism, inviting Wakefield, Redwood, Bernard and others as witnesses to Congressional hearings. But what is less well-known is Burton’s role in creating the very concept of an “autism epidemic.” He coined the term in an April 2000 Congressional hearing. SafeMinds quickly adopted it, even though scientists cautioned that the “epidemic” was mostly due to a change in diagnostic criteria. 

For TAP Pharmaceuticals, however, the reason for the rise would not matter. If Lupron were to be used as treatment for autism, the growing market would create an enormous financial windfall for the company.

In 2004, the Geiers filed a US patent on the use of the Lupron-Chemet combination (or similar compounds) to treat autism. 

But they did not stop with the US. Typically, if someone wants to ensure not just US but global rights to an invention, they file an international application as well. This is an expensive proposition. It can cost millions of dollars. So the application was filed with the World Intellectual Property Organization in 2006 not by the Geiers but by TAP Pharmaceuticals themselves, using the company’s law firm. 

By that point, the Geiers had already accumulated a cohort of patients recruited at autism conferences, and were working with TAP on designing a clinical trial to test their idiosyncratic theory. 

Mark Geier passed away in March 2025. Shortly thereafter, David Geier quietly showed up in the HHS directory as a “Senior Data Analyst.” The Washington Post broke the story: “A vaccine skeptic who has long promoted false claims about the connection between immunizations and autism has been tapped by the federal government to conduct a critical study of possible links between the two, according to current and former federal health officials.”

Ignoring the furor that followed, Kennedy confirmed in a Cabinet meeting on April 10 that such a study was being planned, and quickly. “By September, we will know what has caused the autism epidemic, and we’ll be able to eliminate those exposures.”

The public outcry and congressional questions continued. Finally, in a lengthy June 7 post on X, Kennedy defended David Geier, saying that he was brought in “to advise other scientists” based on his “unique expertise” and “extensive background as a research scientist.” 

The Lupron clinical trial the Geiers conducted paints a very different picture: a father-and-son team that, in search of personal enrichment, performed unauthorized experiments on autistic children for years in violation of multiple ethical safeguards, reminiscent  of human research  horrors of the past -  horrors that led to these safeguards being created in the first place. 

This is the story of children subjected to dangerous experiments - an autistic girl Nina, and an autistic boy Jalen, based on conversations with their families. David Geier refused to be interviewed, as did several TAP Pharmaceutical executives.

(It appears that Geier's work at the HHS was delayed due to difficulties accessing the data from VSD, Vaccine Safety Datalink. Therefore the focus of Trump's September 22 "Autism press conference" was on Tylenol, based purely on published studies. Geier's work, in contrast, is focusing on aluminum, we predicted. Trump did say,"we want no aluminum in the vaccine," in his remarks.)

The first problem the Geiers faced was that as inventors of the treatment being studied, they would benefit directly from the success of the clinical trial. Could they also be unbiased  “investigators” - actually running the trial, deciding which patients are appropriate, managing toxicity of this new, never-before-tested drug combination, analyzing the data and publishing the results?

This arrangement would be categorized by current US regulations as a “Significant Conflict of Interest.” While not technically forbidden, it requires a special conflict management plan, informing patients of the conflict so they go into it with eyes wide open; setting up independent monitoring of the trial by a separate Conflict of Interest (COI) Committee, and restrictions on decision-making by the conflicted investigator. 

Why not pass on the trial to other, less conflicted doctors to run?

Here the Geiers had another problem. They had big plans that were independent of TAP Pharmaceuticals. They called their newly invented treatment “Lupron Protocol” (not to be confused with the clinical trial protocol - more on that later) and planned to license it to other doctors for a fee, in a “franchise” model similar to how for example Pizza Hut or McDonalds do not operate each location themselves but sell a license to independent owners.  

Let’s stop for a second and reflect on this. Here is a doctor and his son who are developing a new treatment for autism.  But they are doing it not to share their knowledge with other doctors, via a publication or presentations at conferences, as would be usually done. Instead, they want to (and in fact, they did) sell this knowledge to other doctors for a fee.  

First, this is unprecedented. There are some examples of surgeons patenting techniques and selling to their colleagues. But we know of no other example of what the Geiers did: taking medical treatment that is FDA-approved for one purpose, packaging it into a “protocol” for another purpose, without FDA approval, and charging doctors for the privilege of using it. 

Second, clearly this represents another conflict of interest - potentially even more significant than the TAP collaboration and patent. 

And third, this explains why the Geiers wanted to conduct this trial themselves, as the people who would then be licensing (franchising) their “protocol” to other doctors. They had to be in control. 

For all these reasons, the normal process of starting a clinical trial would not work for the Geiers. 

This normal process would require approval of an Institutional Review Board (IRB). This is ordinarily very easy. There are about 2300 IRBs in the US. Most serve individual academic institutions, but some operate independently and will review any clinical trial for a small fee. All are required to be registered by the Office for Human Research Protections (OHRP) within HHS, and must adhere to a whole set of requirements as mandated by the US Code of Federal Regulations. 

It is safe to say that no IRB would approve a clinical trial in which the two lead investigators own a patent on the treatment being studied, and far worse, have plans to license it to other physicians for a fee. 

So the Geiers came up with an ingenious solution: they created their own IRB with themselves as members.

But before we get to that, why do these IRBs exist, and since when? Many of us take the existence of these safeguards for granted. It makes sense - before some doctor performs an experiment on people, they need to persuade an independent committee that it’s ethical. 

Shockingly, the IRB approval requirement has not been in place that long. And it was partly the tragedy of another autistic child and others that ultimately led to these regulations being put in place.

***

Nina Galen was diagnosed with autism when she was less than 2 years old, her mother, Diana McCourt, told me on a recent afternoon in her Upper West Side apartment. On the bookshelves behind her were books by her husband, Malachy McCourt, and brother, Frank McCourt. 

“We took Nina to see a psychiatrist on Horatio Street, here in New York City. She could not walk, so I carried her in my arms,” McCourt told me. Her face is devoid of emotion, but looking in her eyes, I see endless sorrow. “She has autism,” the psychiatrist said. “It’s hopeless.” 

By the time Nina turned 8, McCourt was becoming desperate.  “Nina suffered from severe anxiety that caused her to hurt herself,” she recalls. As a result, Nina had to be supervised 24 hours a day.

“I had two other children of my own, and my second husband, Malachy, had two of his. It became impossible for us to take care of Nina.” Out of desperation, McCourt left Nina in the hands of researchers. 

The Geiers would ultimately see hundreds of similarly desperate families and offer hope of “cure” for their children with their Lupron+Chemet “protocol.”

But Nina was not one of them. The year was 1971. 

“I tried placing Nina in a day center. But I could not find one that was equipped to handle someone with her level of disability.” 

Finally, McCourt was left with only one option: placing Nina into an institution. There were several in the New York area. One, Letchworth Village, was “so awful that the McCourts steered clear of it.” They chose one whose name still evokes dread and horror to this day: Willowbrook.

According to a recent expose by ACLU called “Cleaning Up the Snake Pit,”

The Willowbrook State School, New York City’s institution for people with developmental disabilities, bore a superficial resemblance to a college campus. Opened in October 1947 by the New York State Department of Mental Hygiene, Willowbrook consisted of many buildings, most of them brick, surrounded by grassy areas and spread out on a tract of about 300 acres on Staten Island. An exterior view of Willowbrook made it seem a pleasant place… in 1972, it had about 6,000 “inmates,” even though the hospital was built to house a maximum of approximately 4,000 residents... In 1965, Robert F. Kennedy, then a U.S. senator from New York, had paid an unannounced visit to the institution and denounced its conditions… After visiting Willowbrook, Kennedy described the facility as a “situation that borders on a snake pit.”

But Willowbrook residents would have to wait for another 6 years - to be brought to light one more time - before any changes would come. 

In 1972… local ABC television journalist Geraldo Rivera had managed to get a camera into the institution for a few minutes and had broadcast some disturbing images of children wallowing in their own feces and urine. Not content with simply showing viewers the conditions, Rivera described the stench of the facility. “It smelled of filth. It smelled of disease, and it smelled of death,” he reported.

Incidentally, Nina was not the only autistic child at Willowbrook. Allison Singer, the President of the Autism Science Foundation, recalls that “When she was a little girl, in the 1970s, she would visit her older brother, who has non-verbal autism with a cognitive disability, at the now-infamous Willowbrook State School on Staten Island. ‘I just remember hearing a lot of screaming and moaning.’ I hated it.”

Michael Wilkins, the famous Willowbrook physician and whistleblower who invited Geraldo Rivera for an unannounced visit, concurs. “I did not know what autism was when I was working at Willowbrook. I did not learn about it in medical school. But now, as I lay at night and remember the Willowbrook children, I know many of them had autism. We had a whole floor of higher-functioning children - many of them likely were autistic,” he told us.

Horrific living conditions were only one of the dangers. “Disease and neglect were everywhere, and multiple residents died from untreated illness and abuse,” Forbes recently noted. 

One disease that was rampant at Willowbrook was hepatitis. In the early 1950s, this attracted the attention of Dr Saul Krugman, a hepatitis researcher at NYU.  This was not vaccine research, yet. The hepatitis vaccine would be developed 20 years later by others. There were some basic questions to answer first.

At the time, it was known that hepatitis spreads from person to person and has caused many epidemics. Some of them were clearly spread by what is called the “fecal-oral route” - an infectious agent in the stool of one person somehow ends up in the mouth of another, such as through poor handwashing. Other hepatitis epidemics were caused by contaminated needles, thus were blood-borne. It was also known that hepatitis was caused by a virus. 

But a key question remained unanswered: were these different types of epidemics caused by the same virus, or different ones?

To Krugman, Willowbrook offered a unique opportunity to answer this question. 

Krugman’s first human experiment was to demonstrate that a blood preparation containing antibodies, called gamma globulin, could protect Willowbrook residents from hepatitis. It was a crude randomized trial - in each building, inmates were divided into two groups: one group was given gamma globulin, one was not. There was no indication in the publication that informed consent was obtained. One risk, well-known already at the time, was of contracting a blood-borne disease from gamma globulin itself, prepared from pooled blood of multiple donors. Some steps were taken when preparing the globulin to kill infectious agents, but it was unknown whether these steps were fully effective. (Today we know that Hepatitis B or C virus would not be killed by these steps.)

The trial was remarkably effective, reducing the rate of hepatitis infection 10-fold. 

What next? The ethical thing would have been to administer gamma globulin immediately to the control subjects who had not received it yet, and to begin administering it regularly to stamp out rampant infection (prior studies showed that the effect lasts for many months). But Krugman had other plans. He wanted to understand “the mechanism of this protection.” What happens when someone is infected with hepatitis after receiving gamma globulin? Hence his idea was born of “feeding virus to patients.”  

This required controlled infection in a separate ward. Regular wards with “children wallowing in their own feces and urine,” as observed by Geraldo Rivera, would not do. And participants needed  to be new arrivals, not yet exposed to hepatitis. How to explain to the parents why their children were  to be housed in a separate wing, observed by doctors and nurses? And how to convince them to subject their children to purposeful infection with hepatitis? 

Kuperman found a carrot to entice the parents with: skipping the waiting list. According to Forbes,

When Dr. Krugman and Dr. Giles began the Willowbrook hepatitis experiments, they used the conditions of Willowbrook to their advantage for recruiting new families. Despite its well-documented horrors, Willowbrook was still one of the only options for children with severe disabilities, and there was a long waitlist. Dr. Krugman offered several parents, including Nina Galen’s, the ability to jump the line and have their children put in the newer, cleaner research wards with more staff—if they joined the experiments. “I did feel coerced,” McCourt says, “I felt like I was denied help unless I took this [opportunity].” 

Krugman also told parents that since hepatitis was already prevalent at Willowbrook, their children may as well have the chance for a vaccine. McCourt remembers being told her daughter could get an “antidote” to hepatitis if she joined the experiment. When she asked why the hepatitis studies couldn’t be done on primates, she was told that using animals would be “too expensive.”

“Permission was obtained from the parents of children five to ten years of age newly admitted to Willowbrook,” wrote Krugman and colleagues in their 1958 New England Journal of Medicine article. “They were brought directly to an isolation unit and had no contact with the rest of the institution. They were cared for by physicians, nurses and attendants who also had a minimum of traffic with other patients and personnel.” 

Then came the infection. From 1955 to the 1970s, Krugman and colleagues performed a series of experiments, injecting children with hepatitis-infected blood or forcing them “to drink chocolate milk mixed with feces from other infected children in order to study their immunity.”

In March, 1965, Dr Henry Beecher, a professor of anesthesiology at Harvard Medical School, arrived at the Brook Lodge, near Kalamazoo, Michigan, the former home of William E. Upjohn, the founder of the pharmaceutical company bearing his name. Beecher was there to present at the Upjohn-sponsored conference “intended to be an opportunity for scientists to explain the process and challenges of drug development to science journalists while fostering congenial relations between the press and the pharmaceutical industry. Speakers were well-known academics from across the country, and topics ranged from experimental design to the placebo effect to safety issues. Attendees represented all wire services, major daily newspapers including the Los Angeles Times and New York Post, as well as periodicals such as Time and Newsweek.”

Upjohn invited Beecher to speak on the ethics of human experimentation. In the weeks leading up to the conference, Beecher alerted the organizers that his talk would be a “bombshell,” and asked to extend his time to 40 minutes - they agreed.

In a session called “Special Problems in Clinical Research,” Beecher dropped his payload on the topic of the “The abrogation of ethical responsibility among American researchers.” “He described cases in which researchers withheld penicillin from soldiers with rheumatic fever, conducted unindicated thymectomies to study its immunologic effect, transplanted a melanoma tumor from one patient to another…”

Among his examples was Krugman’s Willowbrook research.

The aftermath was swift: as media outlets fanned out to broadcast Beecher’s scathing accusations, researchers pushed back, accusing him of “gross and irresponsible exaggeration.” Beecher persisted, writing an article based on his talk. The Journal of  the American Medical Association refused to publish it. The New England Journal of Medicine ultimately did, after its Editor-in-Chief overruled six of seven peer reviewers. 

Beecher’s 1966 NEJM publication prompted the US Surgeon General to request an investigation. But no substantive change ensued. 

When Nina Galen entered Willowbrook in 1971, five years after Beecher’s NEJM paper, Krugman’s experiments were still ongoing, and she became yet another victim. 

“I never found out what they did to her,” McCourt told me. “I came to visit her every week. First, they placed her in an isolation room by herself, sedated with medication. Later, they moved her to a general ward. When they brought her out to see me, she always wore different clothes, as though she was hastily dressed in whatever was nearby. She looked pale, and sometimes had bruises on her body.”

Real change had to wait for another shocking human experiment to be revealed. Peter Buxtun, a 1971 law school graduate, could no longer hold a dark secret he had learned as a public health employee in 1960s, and which the CDC ignored when he alerted them in 1966. On July 25, 1972, The Associated Press broke the story of Tuskegee, based on the documents Buxtun provided. 

“Of about 600 Alabama black men who originally took part in the study, 200 or so were allowed to suffer the disease and its side effects without treatment, even after penicillin was discovered as a cure for syphilis,” AP reported.

This was the final straw. In 1974, Congress passed the National Research Act, Public Law 93-348, which became the first federal law mandating IRBs.

Jalen Coates was born in 2000 to a Jewish mother, Kim, and an African-American father, Cori. Diagnosed with severe autism as a young boy, Jalen never learned to talk, and had frequent temper tantrums that became harder to manage as he became stronger. One day, when he was 8 years old, his therapist told Kim and Cori about a doctor nearby in Maryland who was offering a chance of cure - Mark Geier. 

“Dr Geier’s office was beautiful. Him and his son were very nice to us,” recalled Kim and Cori in a recent zoom call. “Jalen had a temper tantrum right in their office. We were scared he would break their furniture. But the Geiers were so patient. ‘Just wait, he will quiet down,’ they said.”

The Coates do not remember signing any informed consent document. In fact, they were not aware they were participating in a clinical trial at all. “If I knew this was a clinical trial, I would have never done it,” Kim said. 

But even if they were provided with the informed consent form, the one prepared by the Geiers was grossly inadequate. A short two pages, it failed to explain what the study was about, and failed to mention important side effects or the glaring conflicts of interest. 

The clinical trial protocol the Geiers followed was even shorter - only page, containing none of the required elements that the IRB would typically look for to determine whether the clinical trial is acceptable: What are the objectives of the study? What is the rationale? What is the exact patient population - do certain patients have contraindications? How to manage toxicity? 

To obtain insurance reimbursement for the cost of Lupron, the Geiers entered their patients’ diagnosis not as autism, but as precocious puberty - an FDA-approved indication. “Lupron was covered by our insurance,” Kim recalled. 

With Chemet, it was more complicated - the drug is only FDA-approved for acute lead toxicity, so insurance would not cover long-term use. “We could not afford Chemet, so Jalen only received Lupron.”

This, in the language of clinical research, is called a “major protocol deviation.” The protocol calls for two treatments - Lupron to block testosterone, and Chemet to remove mercury (however ridiculous the rationale). By skipping one of the components, the Geiers managed to deviate even from their 1-page protocol - and for an absolutely egregious reason - the family’s inability to pay.

Lupron did help with temper tantrums - an expected consequence of lowering testosterone. But Jalen was “crying all the time,” Kim said, “and we did not know why.” He gained weight and developed gynecomastia - breast enlargement. All are common Lupron side effects. 

Investigational drug product is obviously a key element of any clinical trial. Jalen was on Lupron for 3 years. The first two, the Coates obtained Lupron from a pharmacy. Except it was not Lupron, but a generic version, leuprolide, made by Teva pharmaceuticals, Kim recalled. 

“Then, for the last year, the Geiers told us to pick it up from their house,” Kim recalls. “It came in glass bottles with a home-made label that barely stayed on the bottle,” she said. “Once, I went down to the Geiers’ basement to use the bathroom,” she recalled. “I saw several ladies packaging drug bottles into boxes. And in the bathroom I saw what appeared to be equipment from a chemistry lab.”

What was the chemical Jalen was receiving - was it even leuprolide? Did the Geiers make it in their basement? Or was it just water? We don’t know. 

“Did Jalen’s side effects get better,” I asked Kim. She said they did not, so presumably there was some version of leuprolide that Jalen was getting in the self-made bottles. 

Here was the third mechanism by which the Geiers profited from prescribing Lupron (or some version of it) to their clinical trial subjects: in addition to the patent and the licensing fees paid by other doctors, they also ran a pharmacy and charged insurance companies for the Lupron their patients received. 

After three years, the Geiers stopped their treatment to allow Jalen to enter puberty. Not surprisingly, he was not cured of his autism. His temper tantrums came back. 

The Coates stopped seeing the Geiers, unaware that Jalen may yet suffer long-term consequences of Lupron use. Some patients develop osteoporosis or dental issues. 

“The idea of using GnRH agonists to treat autism, especially based on such a questionable theory, is deeply troubling,“ Dr Lars Dinkelbach, a Clinician Scientist at the Department for Pediatric Endocrinology at the University of Essen in Germany and an expert in puberty and its medical modification, told us. “Pubertal timing plays an important role in neurodevelopment. That makes me think we need to be especially cautious when considering interventions like GnRH agonists, particularly in vulnerable populations. The long-term consequences of such treatments, including potential effects on fertility, bone health, and mental health, are still not fully understood.”

The Geiers ultimately published their clinical trial in two scientific papers, both with David Geier as first author. The first article, called “A clinical trial of combined anti-androgen and anti-heavy metal therapy in autistic disorders,” was published in 2006 and reported data for 11 of the trial participants.

Adverse events - a critical aspect of any clinical trial - are not mentioned at all in the “results” section of the paper. The discussion section states, “It  was observed  that  the  treatment  protocol  employed in the present study resulted in minimal significant adverse health effects… The patients’ parents also reported no  significant  adverse  health  effects  of  the  treatment  protocol  employed.”

We don’t know whether Jalen was one of these 11 patients, but he alone had three adverse events - crying, obesity and gynecomastia. And the contemporaneous online chatrooms are full of patients’ stories of side effects from the Geiers’ Lupron+Chemet treatments: increased behavioral problems, “brain fog,” urinary incontinence, yeast infections, and others.

The other publication of this clinical trial’s results came in the form of a paragraph in another article by the Geiers, again with David as the first author. “In our own clinical experience we have observed that leuprolide acetate (LUPRON®) administration to nearly 200 patients diagnosed with ASDs significantly lowered androgen levels and has resulted in very significant overall clinical improvements in socialization, sensory/cognitive awareness, and health/physical/behavior skills, with few non-responders and minimal adverse clinical effects to the therapy.”

Neither paper has been retracted. Since their publication, case reports have appeared in the literature of autistic children treated with Lupron around the world. It’s not a surprise - to a physician coming across these papers, the treatment looks promising - tremendous efficacy and (if one trusts the authors) “minimal adverse effects.”

The Geiers treated over 300 autistic children, and some adults, with their “Lupron Protocol.” They were ultimately stopped thanks to Kathleen Seidel, the mother of an autistic child and the creator of Neurodiversity Weblog (now archived as Neurodiversity.net). In 2006, Seidel noticed the reference to an IRB in one of Geiers’ publications. Through a Freedom of Information Act request, she obtained the membership of this IRB. What she received was concerning.

“Whereas the IRB was registered in March 2006, the research described in the article was conducted between November 2004 and November 2005. Further, according to The Common Rule, Mark Geier [IRB chair] and David Geier would be ineligible to vote on any of their own research proposals. Anne Geier would be ineligible to vote on any research proposed or conducted by her husband or son. Rev. Lisa Sykes would be ineligible to vote on any study in which her son is a participant. As a co-investigator with Mark and David Geier in their Lupron research, John Young, too, would be ineligible to vote on any IRB supervising that research. Of the seven members of the IRB, only a minority of two — Kelly Kerns and Clifford Shoemaker — would be eligible to vote on the research described in the article — and only if they are free of any personal or financial interest in its outcome.”

Mr Shoemaker is a vaccine injury lawyer, and Mrs. Kerns is an anti-thimerosal activist, Seidel noted.

A reader of Seidel’s blog conveyed these findings to the Maryland Medical Board, which conducted a multi-year investigation and ultimately concluded:

Dr. Geier has displayed in this case an almost total disregard of basic medical  and  ethical  standards  by…  providing “informed consent” forms that were  misleading  and  in  at  least  one  case  blatantly  false.    He  provided treatments  supposedly  according  to  an  investigational  protocol,  but  the investigation was approved only by a sham IRB, and he applied protocols to patients who did not fit his own profile  ...  At the same time, he profited greatly from the minimal efforts he made for these patients.  In plain words, Dr. Geier exploited these patients under the guise of providing competent medical treatment…”

Mark Geier’s license to practice medicine was permanently revoked in 2012; he petitioned for judicial review but the revocation was upheld.  

The saga of unethical conduct does not end there. To circumvent the loss of his license, Mark Geier asked at least one colleague, and possibly others, to prescribe Lupron to his patients on his behalf. He then moved his operation to his franchises other states, and continued until each of the states, one by one, stripped him of those licenses as well. 

Mark Geier was the only research mentor his son ever had. This alone would be concerning. But worse, David Geier worked side-by-side with his father in the clinic, sometimes even seeing patients by himself, ordering and interpreting tests, and even forging his father’s signature on at least one test order. He was sanctioned and fined for practicing medicine without a license. His petitions to overturn this ruling, like his fathers had, failed.

Were all the hundreds of patients who ultimately received the “Lupron Protocol” participating in the original clinical trial? Or was the trial completed at some point, and slowly morphed into clinical practice, without a full publication and accounting of results? I asked Seidel this question. “I do not recall ever seeing any indication that some patients were included in the Geiers' "clinical trial" and some were not. Honestly, although they sometime used the term "clinical trial," I suspect that the Geiers had little conception of what is involved in a true clinical trial; rather, they used the term "clinical trial" to put a scientific gloss on what was in fact uncontrolled pharmaceutical experimentation on autistic children conducted without ethical review,” she said.

Jalen is now 25. I got to meet him recently - a handsome young man, he shook my hand when I was leaving.  He remains non-verbal and is now on 4 medications to control his temper tantrums, with some success.

Nina lives in a group home with two similarly disabled roommates and 24-hour care, paid for by Medicaid. Dr. Bernard Carabello, a long-time resident of Willowbrook who became a disability advocate in New York State, told us he is worried about Nina and others like her. “Medicaid cuts will have a devastating effect on the services they and scores of other disabled people are receiving,” he said.

And David Geier? He is now at the HHS,  in charge of Kennedy’s “autism study.”  

"Autism Epidemic:" The Origin of an Idea.

With the attention drawn to the concept of an "Autism Epidemic," we gathered four top voices in this field - Mitzi Waltz, Gil Eyal, David Mandell and Greg Robinson - to explore the origins of this idea. Who coined the term and why? Is autism a new phenomenon? Is it a "disease" that requires a "cure"? And could autism be what makes us human?

Korney Chukovsky (pronounced cor-NEIGH choo-KOF-ski) is described as a “Russian Dr Seuss.” Born out of wedlock and kicked out of school for being of “lower class,” Chukovsky taught himself several languages and became the beloved founder of Russian children’s literature. In the 1960s he defended political dissidents such as Anna Akhmatova, using his tremendous reputation. One of his most famous poems, written in 1921, is Tarakanishe, or “Giant cockroach.” In it, happy existence of various (mostly large) animals is interrupted when a scary cockroach appears and threatens them with his dictatorial rhetoric. They all run and hide. Attempts at rebellion prove unsuccessful. Finally, a sparrow, oblivious to unfolding drama, comes and swallows the dictator. 

This fairytale left a strong impression on me growing up. First, because I suffer from katsaridaphobia - intense fear of cockroaches (from Greek, κατσαρίδα (katsarída), cockroach and -φοβία (-phobia), fear). But mainly because I thought of this poem as an allegory for Stalin’s rule. (In fact, the poem predates Stalin’s rise to power and instead was inspired by the pervasive fear during World War I.)

I was reminded of Tarakanishe in light of our current public health crisis. Look around, and you will see our fear of Kennedy and his team. 

You will see it in the deafening silence from the scientific community. Where are the autism experts speaking out against Kennedy’s proclamations of an “autism epidemic that started in 1989?” Where are the women’s health experts speaking out against Makary’s falsehoods about hormone replacement therapy (HRT)? In his bestselling book, Blind Spots, he incorrectly declares: “The data are clear: HRT saves lives.” He cites a 20-year-old article that was subsequently refuted. Bhattacharya was accused of publishing a “deeply flawed” anti-lockdown paper sponsored by an airline, an accusation he never fully rebutted. Yet The Washington Post considered them “credible scientists.”  Makary recently repeated his HRT falsehoods in an FDA roundtable, in which several sycophantic experts, uncomfortably seated in a circle, agreed with him, looking straight into the camera. 

You will even see fear in the most brazen Op Eds such as the one signed by nine prior CDC directors. Everything published by the media is carefully reviewed by their lawyers to make sure they cannot be sued for defamation. Certain topics are therefore taboo. One is talking about Kennedy’s “plans” - since planning to defraud the government is illegal, and calling it a “plan” would amount to accusing him of a crime (pardon my amateurish explanations which could be inaccurate). Two, any mention that Kennedy is motivated by money is stripped by the lawyers as well, unless it can be carefully and precisely documented, which we don’t have enough details to do. And three, current HHS employees are threatened by HHS lawyers not to “leak” lest they be criminally prosecuted for breaking confidentiality agreements that remain in force after they leave, according to confidential sources (this has not been reported for HHS, but has for other parts of the government). So even the people who resigned or were fired from the CDC cannot reveal the details of Kennedy’s plans. 

And you will likely see fear, if you look closely, in the Senator's questions during Kennedy’s upcoming Senate hearings on Thursday. First, Congress has no power over Kennedy. He has not been responding to their ongoing questions, and there is nothing they can practically do to compel him to respond short of a lawsuit that will last years. And I know from a confidential source that Senators are scared of Kennedy.

They are afraid to ask him questions about vaccines for fear that he will know more than they do on this topic, and the questions will back-fire.

So we are left with general statements such as “undermining trust in vaccines.” Imagine a cabinet member is seen building an army of robots in his garage. We cannot say he is “planning to take over the world.” The Senators will be afraid to question him - “he knows more about robots than we do.” And we cannot speculate on his motivation. The resulting headline - “He is undermining trust in robots!”

Why is this fear of Kennedy a problem? Several reasons - it hampers our ability to anticipate and “prebunk” his strategy. It shifts the news cycle into reactive mode, just counting losses. It prevents us from thinking rationally, further confuses the public, and further stokes our divisions. 

I hope the Russian fairytale of lions and hippos hiding from a cockroach will open our eyes.

By Alex Morozov

August 27, 2025

Here is a simple explanation for Susan Monarez's firing.

Recall, we and others are predicting that Kennedy will shortly announce the results of an HHS-sponsored study showing that aluminum-containing vaccines cause autism. This study will likely be fraudulent, but we won't necessarily be able to prove that it is.

Kennedy then will pass this data to the newly restaffed and loyal ACIP to vote. And judging by their rehearsal in their first meeting, they will vote to recommend to take all aluminum-containing vaccines off the schedule. By law, this recommendation then goes to the director of the CDC for endorsement.

Kennedy initially nominated Weldon to CDC director, the mercury-causes-autism guy and a mentee of Dan Burton, the original anti-vaxxer in Congress (more on that in a future story - let's just say Andrew Wakefield was his star witness).

But several republicans woke up and opposed Weldon's candidacy - Cassidy, Collins and Mirkowsky. We all wish they had done that for Kennedy - that would have made a difference. Opposing Weldon accomplished nothing, as you will see in a second.

When that fell through, Kennedy (1) moved the autism study to NIH, and (2) was forced by the Senate to appoint a well-credentialed scientist, Susan Menarez.

But she was clearly in the way because the ACIP recommendation would go to her. So before he unveils the “aluminum causes autism” news, he needed Monarez to be out.

With this firing, Kennedy and Trump are essentially telling the Senate - you can vote for whoever you want, and then if they get in the way, we just fire them.

What is such an important thing that they are planning, that would risk them blowing up the CDC?

The "aluminum plan."

By putting together multiple lines of circumstantial evidence, we have uncovered what we are calling The Kennedy-Siri-Bigtree Plan to Defraud America - proclaiming in September that “Aluminum Causes Autism.”

The crux of the plan was articulated by Kennedy in 2021 when he stated, “Under the Vaccine Injury Act, you cannot sue a vaccine company for an injury... The only exception is that if they knew of an injury that is caused by a vaccine and they failed to list it on their manufacturer’s insert.”

He was referring to the US National Childhood Vaccine Injury Act of 1986 which shields vaccine manufacturers from lawsuits, and instead sets up a program called Vaccine Injury Compensation Program (VICP). It was a win for patients who now had a much easier process to receive compensation, and for pharma who were on the brink of stopping vaccine development.

But for one group this law was a devastating loss: antivax lawyers. Typically medical malpractice lawyers charge a 25-40% “contingency fee,” meaning that they do the lawsuit for free but take a cut from the ultimate award to the patient. But in VICP cases, contingency payments are prohibited.

However, there is one loophole that permits VICP to be skipped. Section 42 U.S.C. § 300aa–22(b) of the Act provides an exception if “the injury or death resulted from a vaccine not properly prepared or was caused by the manufacturer’s failure to provide proper warnings or instructions.”

Since Kennedy first mentioned aluminum in 2005, in his Deadly Immunity article, the question remained: how can aluminum in vaccines be toxic, if it's ubiquitous in our environment?

The first clue came when Meryl Nass, (an MIT graduate and Maine former doctor whose license was suspended during COVID) was researching the Gulf War Syndrome in the early 2000s. She zeroed in on the anthrax vaccine (which contains aluminum) as the cause. To find further evidence, she needed a neurotoxicologist, and found one in none other than... Christopher Shaw (See Aluminum Grand Finale). The question was, can we reproduce Gulf War Syndrome by injecting the anthrax vaccine into mice? But they could not procure the actual vaccine, Shaw recalls in his book (with Foreward by Kennedy). So he just injected aluminum. Mice developed neurotoxicity (note, to this day, all papers showing neurotoxicity of aluminum in mice have Shaw as an author).

This caught the attention of Leslie Manookian, who in collaboration with Claire Dwoskin was producing an anti-vaxx documentary, The Greater Good. Manookian interviewed Shaw, who is in the film describing his experiments, voicing over a cartoon of a dying mouse (The film premiered in April 2011). Manookian then introduced Shaw to Claire Dwoskin, and the aluminum plan was set in motion.

The plan is as follows.

1.    (Done) In 2011, anti-vaxxers – Andrew Wakefield, Barbara Loe Fisher, Claire Dwoskin, etc, start to collaborate with the “aluminum family” of international researchers - Christopher A. Shaw, Lucija Tomljenovic, Guillemette Crepeaux, Romain K. Gherardi, Christopher Exley, Lluís Luján, etc (See Aluminum Grand Finale part 1)

2.    (Done) Anti-vaxxers - primarily Dwoskin, an aluminum-focused anti-vaxx philanthropist - sponsor research by the “aluminum family” demonstrating harm of aluminum vaccines. This research results in dozens of scientific publications supporting the harms of aluminum, that remain unchallenged and un-retracted. No systematic effort by the mainstream scientific community has yet been undertaken to evaluate the claims made by the supported “aluminum family.”

3.    (Done) Discuss with Trump. When Kennedy met with Trump in January 2017, they likely discussed aluminum. But their plan would have to wait until HHS leadership would be replaced with loyalists.

4.    (Done) File petitions with the HHS, in 2017 and 2021, to warn of aluminum risks, to document that they were warned. The “aluminum family” above is cited in the petitions.

5.    (Done) Proclaim that there is an “epidemic of autism” starting at the “red line” year of 1989. (HiB and HBV vaccines, both aluminum-containing, were introduced in the 1980s). Kennedy said "EPA scientists" came up with this "red line" of 1989. If you are wondering what he is talking about, here is Wakefield discussing the paper in the 2011 Jamaica meeting.

6.    (Done) Announce an HHS-sponsored “autism study” – a number-crunching exercise on vast amounts of data previously collected from Americans - that will “find the cause of autism by September.” (Note, Kennedy also announced the 50-million-dollar grant competition, what's called the Autism Data Science Initiative, ADSI - that is delayed, the winners will be announced imminently but obviosuly will not be in time for September. This is probably what Kennedy refereed to as "in the next 6 months we will have more.")

7.    (Done) Attack mainstream journals. Since mainstream journals will not publish Kennedy’s fraudulent study, they need to be discredited in advance, providing an explanation for why this paper would be published in a Kennedy-aligned journal, such as Journal of the Academy of Public Health, with Makary and Bhattacharya on the editorial board.

8. (Done) Restaff ACIP, and fire Monarez, so when the aluminum data goes to the ACIP for a vote, their recommendation then goes not to Menarez, but directly to Kennedy.

9. (Done) Rig the HHS study to ensure aluminum is “discovered”:

a.    Nominate David Weldon, an anti-vaxxer with experience with the mercury-vaccine link, as the head of CDC.

b.    When that fails, take the study out of the CDC and into NIH, where trusted Kennedy loyalists are charge: David Geyer, Jay Bhattacharya and Martin Kuldorff.

c.     Set up a single government-owned database in which this data crunching is be performed, while not allowing outside access, thus preventing independent confirmation of the findings;

10. “On Track” for September -> Falsify data or analysis to arrive at a fraudulent conclusion: “HHS study identifies aluminum adjuvants as the cause of autism.”

11. Last step - Unleash lawsuits against vaccine manufacturers and the previous government for failing to act on the risks of aluminum.

We estimate there to be about 1 million new autism diagnoses since 1989. The total award amount could thus be in the trillions of dollars, thus resulting in an enormous transfer of funds from the government coffers to Trump supporters and Kennedy-aligned lawyers.

Kennedy himself has filed many lawsuits claiming damage from Gardasil (which contains aluminum) and will also benefit tremendously. In fact we hypothesize that there exists a secret agreement between Trump, Kennedy, Del Bigtree and Siri, that the bounty will be divided between them.

This plan, if executed, will lead to a public health disaster – infections and cancers will rebound, while autism rates will not go down, and may even go up, since certain maternal infections increase the risk of autism.

By Alex Morozov

August 20, 2025

See also our Opinion out today in The Hill.

In our last our Opinion piece called “RFK Jr. is targeting aluminum in vaccines — and children could pay the price,” we revealed Kennedy’s apparent efforts to orchestrate a strategic attack on aluminum-containing vaccines. We described a recent Danish study that demonstrated no association between the amount of aluminum children received and their risk of various chronic conditions, including autism. 

This study is in the way of Kennedy’s scheme. Predictably, it was immediately attacked by Children’s Health Defense, an antivax nonprofit founded by Kennedy. 

Then Kennedy himself as a solo author published a lengthy and impassioned yet deeply technical rebuttal of the Danish paper, called “Flawed Science, Bought Conclusions: The Aluminum Vaccine Study the Media Won’t Question.” The rebuttal was published in an obscure trade publication, Trial Site News, rated as “conspiracy-propaganda” by MediaBias. 

Kennedy's rebuttal contains a multitude of scientific falsehoods that were debunked by the scientific community. To its credit, Trial Site News published Dr Hviid's response to Kennedy that I might summarize as the "adult in the room" - "Many of these [criticisms] are related to specific study design choices that are reasonable to raise and discuss, preferably in a respectful manner..."

However, a few aspects beyond the science are noteworthy. 

First, the Danish study was fundamentally quite routine - it confirmed the safety of aluminum-containing vaccines which have been in use for decades. There is no indication from HHS that any new safety concerns with these medicinal products have been identified recently. In this context, it is difficult to explain such an abrupt, fierce and personal attack by the Secretary of HHS, unless he is indeed planning to “discover” aluminum-containing vaccines as the cause of autism as we predicted in our Opinion. 

Here is Kennedy the lawyer – attacking science through intimidation and deceit, as though it was a courtroom opponent. Intimidation is a legal technique that "creates an atmosphere of fear, coercion, and uncertainty, compelling the opposing party to make concessions under duress." And attorney deceit "may take many forms, from lying to clients to concealing facts from the court."

The intensity of the attack was such that it even included a demand on the journal, Annals of Internal Medicine, to “immediately retract the badly flawed study,” an unprecedented attack on an independent, non-government journal by the Secretary of HHS. The journal courageously refused.

Kennedy even attacked Denmark as a country, referring to “Gold Standard Science” as “standards that will determine with whom the United States will collaborate or do business going forward.” This is a worrisome example of the US blackmailing another country into twisting conclusions of an independent scientific study to its will. Kennedy went even further, asking for Danish laws to be broken to release the underlying data used in the publication.

This demand is reminiscent of the “quid pro quo” arrangement Trump proposed to Vladimir Zelensky in the Fall of 2019. This time, Denmark is being offered a deal too: follow Trump’s version of “Gold Standard Science” as outlined in his recent executive order, or risk losing “business and collaboration” with the US. 

Such threats with international policy implications would not be done without Trump’s approval. He appears to be in on the aluminum scheme. It appears likely that Kennedy discussed aluminum with Trump when they met in January 2017. Kennedy was already in discussions with both Exley and Shaw at that point, as they told us. And Kennedy brought up aluminum several times when he was interviewed about his meeting with Trump. 

Second, a lot of Kennedy’s statistical criticisms are similar to those recently made by James Lyons-Weiler, a long-time friend of Kennedy. Lyons-Weiler is now serving as Kennedy’s scientific advisor, Christopher Shaw, one of the founders of the aluminum toxicity field, told us recently.  According to Shaw, Lyons-Weiler recently called him and asked for his ideas for further research on toxicity of aluminum. 

On January 27, when Scott Gottlieb proposed in an interview that Kennedy may be planning to attack aluminum-containing vaccines, Lyons-Weiler responded ominously, “We tried to warn you. We are glad you are finally listening.”

Lyons-Weiler is also the owner of the scientific journal, Science, Healthcare Policy and the Law, which in January published a paper by Mawson and Jacob that Mr Kennedy cited in his confirmation hearings to support the vaccines-autism link. This paper, (which did not make its underlying data publicly available,) was of such poor quality that an external article review by the Journal of the Academy of Public Health (with NIH Director Dr Bhattacharya and FDA Commissioner Dr Makary on the editorial board) concluded that “it should be forgotten.” Yet Kennedy did not voice any methodological concerns when he brought up this paper in the Senate hearings. Such a double-standard - forgiving flawed studies that support one’s agenda, and criticising those that do not - is dangerous for any scientist, let alone for the Secretary of HHS. This is Kennedy’s vision of “Gold Standard Science” he is intimidating Denmark into adopting. 

Third, Kennedy mentions in his rebuttal “mountains of contrary literature documenting the neurotoxicity of aluminum.” He fails to mention, however, that much of this literature comes from the “aluminum family” and was sponsored by Claire Dwoskin, a long-time board member of the anti-vaccine organization National Vaccine Information Center. In 2011 she founded an aluminum-focused anti-vaccine nonprofit, Children’s Medical Safety Research Institute, CMSRI which provided over 1.1 million dollars in research funding to Christopher Shaw alone, according to a CV he provided to us. Another $860,000 came to Shaw from the Luther Allyn Shourds Dean estate, described as a “private fund that supports vaccine-critical research.” 

The specific paper Kennedy cites in his rebuttal is also authored by the “aluminum family.” Among the authors is Christopher Exley. Kennedy personally offered Exley research funding, as Exley describes in his book (his institution declined the funding due to Kennedy’s antivax views). Another paper Kennedy alludes to in support of his criticisms of the Danish study lists CSMRI and Luther Allyn Shourds Dean estate among its sponsors.

Forth, Kennedy and Children’s Health Defense made multiple contrived accusations of conflicts of interest against the Danish study. First, Children’s Health Defense quoted Exley as blaming the aluminum industry for influencing the study authors. However, privately Exley told us that it’s unlikely that the study authors could be in direct contact with the aluminum industry - “Indirectly for sure, for example through pharma,” he said. Shaw privately acknowledged that the amount of aluminum in vaccines would be negligible for the aluminum industry, but he proposed a possible rationale: fear of negative perceptions of aluminum in general. “Next thing you know, people will stop buying aluminum boats,” he conjectured.

Now, Kennedy is resurrecting CHD’s old criticism of Dr Hviid being influenced by “pharmaceutical giant Novo Nordisk,” failing to mention that vaccines are not part of the company’s pipeline.

These implausible accusations do not amount to a conflicts of interest, in sharp contrast to Mr Kennedy’s real one - the prospect of personal enrichment for himself and other anti-vaccine lawyers. Even Daniel O’Connor, the owner of Trial Site News where Kennedy published his rebuttal, is on the Board of REACT19, a nonprofit “currently engaged in critical litigation on behalf of the entire COVID vaccine-injured community.” O’Connor also runs an investment firm with a portfolio of pharmaceutical startups. One of them is Vaxxine, an Australian company that directly competes with the established vaccine manufacturers Kennedy is attacking. 

Kennedy told Tucker Carlson on June 30th that he expects “initial indicators” from his “autism study” to be released in September. He is likely not referring to the 50-million-dollar grant competition, as those winners have not been announced yet, according to an anonymous source. Instead, Kennedy is likely referring to the database David Geier, an antivaccine researcher with no doctorate degree who was accused of experimenting on autistic children, is setting up. Kennedy touted Geier as the “only person outside of the CDC who has ever had access to Vaccine Safety Datalink.” (In fact Geier was accused of stealing the data twice). “David was always a mercury guy,” Shaw told us. He is referring to Geier’s body of literature attacking mercury-containing vaccines that experts describe as “junk science.” Now the world awaits with baited breath to see how Geier will attack his next target: aluminum. 

This story provides additional background, including interviews with members of the "aluminum family," to supplement our MSNBC Opinion piece out today - RFK Jr. is targeting aluminum in vaccines — and children could pay the price.

On the Northern coast of Jamaica, just West of Montego Bay, the rolling hills are occupied by the Tryall Club - dozens of privately-owned villas that, if you are lucky to find one for rent, go for tens of thousands USD a week. One of the villas, called Twin Palms, was owned for decades by Albert Dwoskin, a Northern Virginia real estate developer, and his young wife, 20 years his junior, Claire Dwoskin. “They travel to Jamaica 5 or 6 times per year with their two children to enjoy the warm, sunny, idyllic setting in the Tryall Club,” said  a travel magazine at the time. 

Back home in Virginia, the Dwoskins were big supporters of Democratic causes, hosting fundraisers for the Clintons at their sprawling riverfront property, a short drive from the heart of DC, and even earning an invitation to a State Dinner at the White House in 2000. 

But after Claire became a mother in 1999, she developed a singular, unusual focus: aluminum. 

“I organized a small conference, inviting about 20 scientists from all over the world, and they presented all their research,” she recalled in a recent podcast. “And the conclusions of the scientists were that aluminum toxicity is linked to so many different conditions. We have been taught that aluminum is benign. But in fact it’s insidiously unsafe.”

The “small conference” Dwoskin is referring to is the 3-day “Vaccine Safety Conference” she organized at the Tryall club in early January, 2011. The Dwoskins rented several villas for the occasion to house the guests comfortably - a separate bedroom for each, as Romain Gherardi, a French neuropathologist turned aluminum researcher who was at the meeting, describes in his 2016 book, “Toxic Story: Two or Three Embarrassing Truths About Vaccine Adjuvants.”

In attendance were most of the world’s top researchers on toxicity of aluminum, including the field's founder, Christopher Exley, who is often referred to as "Mr. Aluminum." In his recent book, "Discussions with Mr Aluminum" (published by Skyhorse, the publisher of " books other houses consider too controversial to publish," such as Kennedy's books), he coins the term “aluminum family” to describe the group. Christopher Shaw, Lucija Tomljenovic (who now works for Children's Health Defense), Romain Gherardi, Yehuda Shoenfeld and a few others were in attendance. 

But this was no ordinary scientific conference. And not only because of the unusually luxurious setting.

Normally, scientists themselves would organize such conferences and set the agenda and the goals.  But in this meeting, opening remarks were delivered by an anti-vaccine activist, Barbara Loe Fisher, a long-time friend of Dwoskin and the founder of National Vaccine Information Center (NVIC) for which Dwoskin was a funder and board member since 2007. In the video of the conference, still on youtube, you can see Dwoskin at the head of the table, then Fisher, then Vicki Debold Pebsworth - Fisher’s long-time friend and NVIC board member (also since 2007), now on ACIP - and then Meryl Nass, an NVIC “Medical Advisor” and an expert in Gulf War Syndrome who played a key role in sparking interest in aluminum (more on this later). 

Also strange for a scientific conference: two lawyers were in attendance. 

And there was Andrew Wakefield, another NVIC “medical advisor” and the disgraced British doctor who was stripped of his UK medical license for experimenting on autistic children. When things went south in his home country, he came to the US in 2001, and low and behold - here he is, giving a talk to other scientists on the “research strategy” for autism. In fact, during the “conference” Wakefield learned that his 1998 Lancet paper was finally retracted - a shocking 13 years later - due to “intentional fraud.” He gave an exclusive interview to Anderson Cooper right from Tryall Club. Little that Cooper new - Wakefield was again discussing “autism strategies” with like-minded colleagues.

It’s clear that Wakefield had a key role in the meeting. Shaw told me that Dwoskin insisted on inviting Wakefield. He is interviewed prominently in the “highlights” video of the meeting. He gave a 40-minute “strategy” talk. Why was he there? Aluminum researchers denied to me that Wakefield conducted any aluminum research himself. But Exley and Shaw both called him a friend and referred to him as “Andy,” and Crepeaux, another member of the “aluminum family,” said she was in recent communication with him.  And Wakefield and Fisher go way back. At a 2002 NVIC conference Fisher organized, soon after Wakefield’s arrival to the US, he gave several presentations and a keynote address. (Wakefield also helped Kennedy pass confirmation hearings in January, as I learned in my lunch with Anthony Mawson. Wakefield’s shadow role supporting Kennedy and Del Bigtree - in a future story.)

Finally, unlike a regular scientific conference, the conclusions of the Jamaica meeting were pre-ordained. Dwoskin was not fully transparent when she described the Jamaica meeting as a revelation to her on the toxicity of aluminum. While she was listening to the talks, a documentary she sponsored called “The Greater Good” was submitted to the Dallas Film Festival, where it premiered 3 months later. (The New York Times called it an “ emotionally manipulative, heavily partial look at the purported link between autism and childhood immunization” - those were the days when such films were still reviewed by the mainstream media). In the film, Shaw voices over a cartoon of a mouse being injected with aluminum, depicted as an ominous white cloud spreading over its body. The mouse starts to swing its head violently side-to-side and dies. A zoom-in of its brain shows the white cloud killing its brain cells. The film was probably a couple of years in the works by the time of the Jamaica meeting, Shaw told me. He was interviewed over a year earlier.

The Fringe Science of Aluminum

This is a story - based on recent interviews with several members of the “aluminum family” and review of archival materials - of a massive effort between 2011 and 2017, largely sponsored by Dwoskin, to produce an entire body of literature - from animal studies to descriptions of human syndromes caused by aluminum-containing vaccines - making aluminum the next target in vaccines after mercury moved to the background.

The interviews offer new insights into how the shift from mercury to aluminum occurred, and why it took so long. They explain the years-long gap between Kennedy’s first mention of aluminum - in his famous 2005 “Deadly Immunity” piece embarrassingly co-published by Rolling Stone and Salon - and his apparent resurgence of interest in aluminum in 2017 that probably led him to rename World Mercury Project into Children’s Health Defense in 2018. And they help us answer the key question in this story - is the “aluminum family” just a group of ordinary scientists doing ordinary research that produced results that helped antivaxxers, and that the “mainstream” scientific community did not want to hear? Or do the scientists themselves hold extreme views?

One clue comes from the literature itself. It is full of holes, short on independent validation and not accepted by the mainstream scientific community - what scientists usually call “fringe.” 

For example, all studies showing that rodents develop neurological dysfunction when injected with aluminum have Shaw as an author. 

Another example: Macrophagic myofasciitis (MMF), a syndrome discovered by Gherardi in a saga he describes in his book. According to Gherardi, MMF is an exceedingly rare condition with vague symptoms that can develop in adults years after receiving an aluminum-containing vaccine. And Gherardi showed that if you biopsy their deltoid - the shoulder muscle where the vaccine was injected years earlier - you do find traces of aluminum. There is only one problem. He never biopsied people without symptoms. His excuse is that this would be unethical. He is clearly wrong. Even sham surgeries have been done in the interest of research. The real reason he did not biopsy healthy individuals is clear: they, too, would likely have traces of aluminum at the site of vaccination. It’s what we call in medicine a “red herring” - a totally irrelevant finding.

Of course, this is not to diminish in any way the symptoms and the suffering of the patients. Gherardi’s research, his book and his activism led to the patients in France forming their own association called E3M. Its website says that the group “does not oppose vaccination in principle, but advocates for the use of aluminum-free vaccines.” In 2013, when the French Ministry of Health refused to fund Gherardi’s work, E3M members went on a hunger strike until the Ministry relented. When I spoke to Crepeaux, Gherardi’s former student, she denied that Gherardi was involved in organizing the strike. But regardless, as a physician, I am not sure that these patients are helped by tying their symptoms to a vaccine they received years earlier. 

Shoenfeld also invented a syndrome. He modestly calls it the Shoenfeld syndrome, or ASIA - Autoimmune/inflammatory syndrome induced by adjuvants. He was not shy at the Jamaica meeting to say that he invented it for the express purpose to help win antivax lawsuits. 

“How did I come to the story of vaccines,” he said in his talk. “Being an immunologist and doing basic research, I was called by lawyers to defend or support claims for vaccine-induced autoimmune diseases… I saw this as a challenge whether I can convince the lay people… I have gained a lot of experience and appeared in many courts. They wanted me to show that this was a real, well-defined autoimmune disease. If it’s lupus, it should fullfill all 4 of 11 criteria [a requirement to make a lupus diagnosis]. And in many of the cases, I knew that it’s lupus. It was lupus-like disease. So it will be lupus in a year. But if it was 3 of 11, the judge said, listen, it’s not lupus. And if it’s not lupus, you don’t have a case… So this gave me the impetus to [describe] what we called 'ASIA syndrome'.”

Enter Kennedy, Del Bigtree and Trump

While the aluminum research was ongoing, Kennedy was kept in the loop. Shaw said he spoke to Kennedy occasionally starting around 2011. Exley said he knows Kennedy for “almost 10 years,” which would be around 2016. 

2015 is when a new actor enters the scene: Del Bigtree. He was at the time a producer on a daytime CBS show “The Doctors.” “At an event in a backyard in the Hollywood Hills, Del Bigtree met Andrew Wakefield and was, like Jenny McCarthy, Robert DeNiro and model Elle Macpherson who is his current girlfriend, taken in,” according to an online magazine.

Wakefield was in LA pitching a documentary based on secret recordings of a CDC employee made by Brian Hooker, an antivax activist who recruited Wakefield earlier for help making a documentary. They needed a real hollywood producer. Del Bigtree left CBS and joined them. He soon started ICAN, Informed Consent Action Network, which would go on to become one of the most successful antivax nonprofits, earnings tens of millions. 

But big money was not in donations - it was in antivax lawsuits. Repeating Wakefield’s UK playbook, him and Bigtree found not one, but two antivax lawyers - Kennedy and Aaron Siri. 

Wakefield provided the link to Dwoskin’s meeting.  Once enough aluminum publications were accumulated by 2017, reliable or not, Del Bigtree’s ICAN, cited them in a brief warning of the dangers of aluminum filed with the HHS on October 2017. Aaron Siri was the lawyer for ICAN at the time, and Kennedy may have been as well (was definitely by 2018). The brief included letters of support written by senior members of the aluminum family - Shaw, Exley and Gherardi, proclaiming aluminum as the cause of autism. Exley and Gherardi use identical language. Shaw’s is different but the same points are made. 

In my interviews with Exley and Shaw, these letters were not something they were eager to talk about. I understand why - they were shameful misrepresentations of science by researchers on antivaxers’ payroll, reminiscent of Wakefield’s misgivings. Exley told me he first met Del Bigtree, the legal brief’s originator, and Aaron Siri, the lawyer who prepared the brief, in 2018. When I reminded Exley of the letter, he hesitated and said that Shaw was the one who coordinated letter writing by the three. Shaw feined poor memory.

The letters are dated June 2017. A few months earlier, in January, Kennedy met Trump. The media broke the news at the time - “Trump proposes to create a vaccine safety commission headed by Kennedy.” 

But the real topic, unknown to the media, was most likely aluminum, as Kennedy let in during an interview a few months later. In the interview, Kennedy brings up aluminum several times (despite being interrupted by the journalist), and says that Trump sent him to meet with the heads of relevant agencies at the HHS. One of them confirmed to me that indeed aluminum was the topic. Shaw also agreed that this was likely the topic. 

But nothing came of it. It’s no surprise - given the quality of the science, HHS leaders were not interested. 

(Andrew Wakefield met with Trump earlier, in 2015, and gave him a pre-release copy of his documentary, Vaxxed. Also in attendance - Mark Blaxill, an antivaxxer, a self-proclaimed expert in the “autism epidemic,” and a co-author of a paper with Walter Zahorodny, who recently testified alongside Kennedy press conference announcing the “autism epidemic” - more on this in a later story.)

At this point, Kennedy and Trump probably decided to wait until Trump’s second term and a more friendly HHS. 

And the science and medicine establishment? They forgot all about aluminum as a potential antivax target. This was a critical mistake. As a result, most of the literature on aluminum in vaccines now comes from Dwoskin’s financing of the “aluminum family.” These studies now are sprinkled across scientific journals, largely uncontested and un-debunked. They pop up any time you google “aluminum and autism.” 

And now Kennedy is in charge of HHS, poised to deal aluminum-containing vaccines a decisive blow by “discovering” aluminum as the cause of the “autism epidemic” in a massive HHS study conducted, fully or in part, behind closed doors.

“Capacity to Excrete Metals”

On April 30, 2025, Kennedy was a guest on Dr Phil’s talk show. They turned to the subject of what Kennedy called an “autism epidemic” and what may be causing it (in fact, there is no epidemic - more on this in another story. To be reminded how tragically wrong Kennedy is when he says that there was no autism in 1960s, just google “willowbrook.”). Kennedy said,

“Well, whatever’s causing the epidemic is not genetic, as I said. You know, people who smoke cigarettes, one out of every four to five died of lung cancer or got lung cancer. That means there was four of those five who didn’t. So they were okay. So you’d say, is it genes that are causing the epidemic of lung cancer? No, it’s not. It’s an environmental exposure. And that environmental exposure disproportionately impacts people with different genetic makeups.

And we know a lot about the genes among autism kids. We know about some of the genetic differences, that there tends to be less glutathione production, there tends to be higher testosterone, there tends to be less of a capacity to excrete metals. And many, many of them, in one case, 100%. In one study, 100% of the kids had mitochondrial disorders. And that indicates that the cause could be an accumulation of materials in our environment that are affecting mitochondrial function and that the kids who already are in the race with low mitochondrial function, metabolic dysfunction are the ones who are the ones who get hurt.

…I mean, a couple things we know is that we know that it’s a toxin that became widespread around 1989 because Congress asked EPA, what year did the autism epidemic begin? And the EPA scientists came back and said, it’s a red line. 1989. So something happened in 1989, some toxic exposure that affected every demographic in our country, from Cubans in Key Biscayne to Inuit in Homer, Alaska. And we also know that it impacted boys at a 4 to 1 ratio to girls.

You need to find a toxin that became pervasive around 1989, and that has that sexual dimorphism in its impacts.”

Incidentally, the EPA study Kennedy is referring to was described eloquently by Andrew Wakefield in the Jamaica meeting. 

On June 4, as I was relistening to this podcast for the second time, it suddenly hit me - metals… metals… what metal could he be talking about? 

Dwoskin meets Fisher

“I struggled to have my children, and I was so blessed to be lucky enough to have my twins, a boy and a girl,”  says Dwoskin.  

“When I first took our twins for their 2-month pediatrician visit, they were given a vaccine that then at 4 months had already been taken off the market because of safety concerns. And I wondered, how could a vaccine be given one month and not the next - it made me question, how rigorous was the safety? But then I continued to vaccinate my children. My son developed speech and language problems, motor skill problems. So we were in the therapy offices. And I saw so many kids who were struggling and could not live their lives in a fulfilled way, and how it was affecting the parents… 

“I began looking into the safety of vaccines, and I probably did five years of reading, talking to others, going to conferences, FDA meetings and Institute of Medicine meetings, and learning. And then I realized that I can make a bigger contribution by starting to fund some science.”

This quest led Dwoskin in 2006 to meet Barbara Loe Fisher, the founder of the National Vaccine Information Center (NVIC), the oldest existing anti-vaccine organization in the US, founded in 1982. 

The DTP vaccine: public health triumph, but at a cost

Fisher started her activism against one specific enemy: the DTP vaccine, and particularly the “P” - the whole-cell pertussis vaccine. 

According to the NVIC website, Fisher “graduated from the University of Maryland with a B.A. in English and was a writer and community relations professional at a teaching hospital before becoming a mother in 1978. She has three grown children and a grandson. Her oldest son suffered a convulsion, collapse and brain inflammation within hours of his fourth DPT shot in 1980 when he was two and a half years old. He was left with multiple learning disabilities and attention deficit disorder and was confined to a special education classroom throughout his public school education.”

The pertussis vaccine was an incredible public health breakthrough. In a heart-wrenching video from a lecture on pertussis at the NIH, at about 9 minutes in, you can see a 7-week old infant having a bout of whooping cough in a scene from a documentary shown by the lecturer, Dr Alexandra Freeman. "He was due to be vaccinated against whooping cough in the next week," says the narrator. "But the germs got to him first." "What is really hard about this is that there is no great therapy when you are like this," says Freeman. Antibiotics don't help the patient, because it's not the bacteria, but the toxin they release, that damages the lungs.

In 1940s, the pertussis vaccine was combined with tetanus and diphtheria, to make DTP, highly effective against three dangerous diseases.

But DTP did have side effects. Fisher’s concerns were not unjustified. According to a historical review, 

DTP vaccines were very reactogenic and concern regarding DTP safety grew in the 1970s and 1980s. A high proportion of infants experienced significant injection site reactions following DTP immunization (redness [37%], swelling [41%], and pain [51%]).

Fever occurred within hours after DTP in many infants: 47% of infants had temperatures > 38 C and 6% had temperatures > 39C. DTP vaccines were also associated with serious systemic reactions, including febrile seizures, hypotonic hyporesponsive episodes (temporary shock-like state), persistent crying and whole limb swelling. Although none were associated with serious long-term sequelae, these adverse events contributed to increasing public concerns about the safety of the vaccine. Especially concerning was the suggestion that some serious cases of encephalopathy were attributable to DTP; these were referred to as “pertussis vaccine encephalopathy.” Studies ultimately detected no true association between DTP and encephalopathy.

Nonetheless, reservations about DTP safety, and to lesser extent, the effectiveness of selected vaccines, had global impact. In the 1970s and 1980s, Japan suspended their pertussis vaccination program because of safety concerns. In Sweden, effectiveness  concerns  led  to  a  suspension  of  Sweden’s  pertussis vaccination program. 

In the US by the mid-1980s, lawsuits related to vaccine safety led several manufacturers to withdraw their DTP vaccines and paved the way to the US National Childhood Vaccine Injury Act (NVCIA) in 1986. This act provides funds to compensate for adverse events following immunization. Faced with such widespread apprehension surrounding DTP safety, great efforts were made to develop acellular pertussis vaccines.

According to the NVIC website, Fisher’s activism played a key role in raising awareness of the DTP toxicity and in passing the 1986 NVCIA.

In 1991, a new, safer pertussis vaccine was developed - the acellular pertussis vaccine (aP), and the DTaP vaccine replaced DTP by about 2000. This was likely the “discontinued vaccine” that Dwoskin remembers from her kids’ 2-month pediatrician visit in 1999.

This DTP story is significant because it was the first aluminum-containing vaccine. Since it was introduced in 1940s, the question becomes - home come the “autism epidemic” only started in 1989 according to the EPA study Kennedy and Wakefield cite? Doesn’t this alone eliminate aluminum as the cause of autism, if, as Kennedy claims, there was no autism in 1960s? Unfortunately, it’s not a simple argument to make, given the toxicity of DTP and the fear of “encephalopathy,” however unfounded. 

Kennedy recently attacked the DTP vaccine, which remains in use in low and middle-income countries due to its low cost. He wants us to ignore it in the aluminum-autism discussion. And if we do that, the next aluminum-containing vaccines - HBV and HiB - were introduced in 1980s, which fits the 1989 start date perfectly.

But first, we need to touch on the mercury era to understand why attacks on aluminum came so much later.

Mercury - the next frontier

Once DPT was replaced by DPaT, the antivax community focused on the next target: mercury. This included NVIC, SafeMinds (whose leader, Lyn Redwood, is now heading up vaccine safety at the CDC), and World Mercury Project (later renamed Children’s Health Defense). “SafeMinds put all their eggs in the mercury basket,” said one of their members recently. 

Why not aluminum?

When I was growing up Russia, and I would get sick, my mom would pull out a glass thermometer with a tiny tube filled with mercury in the middle. If we dropped it, it would break and I watched in fascination as little shiny droplets of mercury would roll on the floor, splitting and coalescing. My mom would get on her knees and catch each one with a piece of cotton. “It’s poison,” she would tell me.

She was right - mercury is a known toxin. And not just to humans, but to microbes as well. A preservative was developed about 100 years ago that has a sulfur group (thio), mercury (mer), and a salicylic acid (sal) was called Thi-mer-o-sal (in the US) or Thio-mer-sal (ex-US). This became a go-to preservative used in vaccine vials that had more than one dose in them. Such vials are much cheaper per dose to make, but there is a problem: when the first dose is drawn by sticking the needle into the rubber stopper, bacteria and fungal spores would be introduced that would start multiplying while it’s sitting in the refrigerator. Hence the need for a preservative.

To a lay person, removing mercury from vaccines might make a lot of sense. For a child prior to 2001, there were only two main sources of exposure to mercury: vaccines and fish. 

Aluminum is a different story. 

Part II to follow. 

On July 21, TIME Magazine published a story called “The World’s Richest Woman Has Opened a Medical School.” This story has so much misinformation that it’s worth fact-checking. 

But in addition, let’s examine whether this story adheres to journalistic principles of independence and neutrality. After all, TIME “seeks to be the most trusted media company in the world.”

Society of Professional Journalists has a detailed Code of Ethics. In particular, two principles stand out:

1. Act Independently

• Refuse gifts, favors, fees, free travel and special treatment, and avoid political and other outside activities that may compromise integrity or impartiality, or may damage credibility.

• Deny favored treatment to advertisers, donors or any other special interests, and resist internal and external pressure to influence coverage. 

• Distinguish news from advertising and shun hybrids that blur the lines between the two. Prominently label sponsored content. Native advertising” – also known as “advertorial” or “sponsored content” – is advertising designed to look like genuine journalism.

2. Ethical journalism should be accurate and fair. 

Alice Park, the author of the story, is a Senior Health Correspondent at TIME. (Neither she nor TIME responded yet to a request for comment.)

It is clear that Park has known Alice Walton - the World's Richest Woman - for a while. Walton lists Park among “the best and brightest minds” she invited to her January 2025 event called “Think Health” at the site of the new medical school in Denville, Arkansas. Its slickly designed opening video has someone say, “Our mission is to implement Alice’s vision.” 

Then in May, Park wrote a glowing profile of Walton, one of 2025 TIME100 Health winners. 

Now, as Walton opens a new medical school, she likely hired a PR agency. What a PR agency might do can be reconstructed from this advice shared by pr.co:

1. Make a media list. 

2. Build trust with the journalist and editors.” "It’s called media relations for a reason." 

3. The best days are usually Tuesday or Wednesday. The reason behind this logic is that you should avoid Fridays and weekends because pitches get pushed over for the following week. Therefore, Mondays are not optimal because inboxes tend to be flooded. “Publishers may receive 600 pitches per day.” 

4. Send your pitch right after lunchtime. By that time, the journalist will have already taken a break but still isn’t focused on afternoon work - which makes this window of time perfect for a compelling pitch.

Now, let’s examine the story. 

...Alice Walton wants her school’s graduates to keep patients healthy by practicing something that most doctors today don’t prioritize: preventive medicine and whole-health principles, which involve caring for (and not just treating) the entire person and all of the factors—from their mental health to their living conditions and lifestyle choices—that contribute to wellbeing.

“most doctors today don’t prioritize preventive medicine”

These are reminiscent of MAHA talking points. The MAHA EO states, “America’s healthcare system is largely focused on treating chronic illnesses rather than preventing them.”

In fact it’s the exact opposite. The majority of medical interventions over an individual’s lifetime are aimed at prevention, starting with prenatal vitamins, childhood vaccines, pediatric well-visits, anti-tobacco campaigns, HPV vaccine, pap smears and other cancer screening, blood pressure medications preventing cardiovascular complications, etc.

“Caring for, and not just treating…” 

This is a common trope of the wellness movement. And unfortunately, people are starting to believe it. Instead of doctors, they turn to friends, internet and social media - very dangerous. 

But the issue is not in education - it’s in care coordination, which was a big priority for the Agency for Healthcare Research and Quality (AHRQ). Tragically, this agency has now been subsumed into another agency and defunded. Under AHRQ was USPSTF, US Preventive Services Task Force, whose evidence-based prevention recommendations every provider knows (or should know) by heart. "These systems are not failing; they are inadequately funded and supported," wrote a former USPSTF Chair.

Park continues. 

Those aren’t new ideas, of course, but traditional medicine has only paid lip service to them. Experts have noted that while as much as 80% of medical education focuses on biology, about 60% of premature deaths are due to behavioral factors including lifestyle habits like diet, exercise, and smoking. “I applied to 34 schools, and nowhere else are they doing this,” says Ellie Andrew-Vaughn, who arrived in Bentonville from Ann Arbor, Mich. 

 traditional medicine has only paid lip service to them - see above. Wrong and dangerous rhetoric that risks to worsen the already low trust in doctors.

80% of medical education focuses on biology, about 60% of premature deaths are due to behavioral factors

The reference Park links to, from 18 years ago, does not talk about education at all, and estimates "premature deaths due to behavioral factors" at 40%, not 60%.

Medicine in the U.S. has long incentivized doctors to respond to people’s symptoms—by ordering many rounds of tests and procedures, to name two cost-driving examples—rather than trying to prevent them in the first place. The doctor-patient conversations that should be at the heart of effective medical care are rare today, and patients are saddled with exorbitant fees that haven’t always contributed to better health outcomes. 

This perpetuates another trope, that the high cost of care in the US is due to over-use. In fact it’s due to prices, as an aptly named study, "It's the Prices, Stupid," describes. 

Where do you start if you want to recreate health care from scratch? There isn’t a single solution, and any strategy needs to account for not just how doctors are trained and practice medicine, but also the financial incentives that currently drive those practices. In 2019, Walton founded the Heartland Whole Health Institute, located steps from the new medical school, which focuses on research, health advocacy, and education about the policies and financial systems necessary to advance preventive care. With AWSOM, she is turning her attention to finding a better way to train the people who will populate that system: future doctors.

recreate health care from scratch

As mentioned above, this rhetoric is very dangerous.

financial incentives that currently drive those practices

Indeed doctors have many financial incentives - many of them are good things. Here are a few examples. But the pejorative tone of the article only sows mistrust.

If successful, AWSOM could prove that medical school should, and can, be about more than just biology and anatomy. It can also be about what drives a person, and what feeds them—literally, figuratively, spiritually. 

A recent study shows that the majority of medical schools have extensive medical humanities programs. So to imply that AWSOM will be first to go "beyond biology and anatomy" is not accurate.

In summary, there is a simple way to fix this article - it just needs to say “advertisement” at the top. 

COMING SOON - the crisis of journalism.