Invited commentary in JAMA by Peter Hotez

Who invited him?

There’s an invited commentary doing the rounds that I thought to read, and after doing so, I invited myself to comment on it.

It was published on November 13, 2023 in JAMA as a commentary and it was invited. By whom? I do not know. Likely the editor. It is entitled: “Health Disinformation—Gaining Strength, Becoming Infinite” and it was written by Peter Hotez.

N.B. Two of the seven references he uses in this commentary are self-references.

Summary of the problem: Disinformation created by humans, is being absconded by AIs, to create an infinite do-loop of disinformation.

As you will soon understand, this is a potential problem indeed, but the disinforming humans need to be defined properly. I suggest a dive into truth to enable this.

Peter Hotez is very worried about people. Because we are too stupid to decide for ourselves what to do as a course of action when it comes to our own health, it is up to noble people like Peter to save us from ourselves, PhDs and MDs, and anyone finger-pointed as a spreader of counter-narrative information (aka disinformation).

Invasion of the Body Snatchers, 1978.

To make his argument in this commentary, he springboards off of an article published in The Atlantic in 2020¹ that claims that ‘disinformation’ is perpetuated by R-u-s-s-i-a-n AIs that are “the next frontier in propaganda”.

Let’s define the word disinformation first, so that there’s no confusion as to what he is referring to.

The Merriam-Webster (M-W) definition of disinformation as a noun:

…false information deliberately and often covertly spread (as by the planting of rumors) in order to influence public opinion or obscure the truth.

This definition is a little different from what I was expecting which was more akin to something like the opposite of information, based on the use of the prefix ‘dis’. Information is defined as: “knowledge obtained from investigation, study, or instruction”. Dis is defined as: “opposite or absence of”. So I thought disinformation referred to the absence of knowledge obtained from investigation, study, or instruction. I tend to use etymology as a means to understand the true meanings of words when I seek to define them. It often does lead to truer understanding of the meanings of words. Translations perhaps aside…

N.B. If disinformation actually was the absence of knowledge obtained from investigation, study, or instruction, then it seems the COVID-19 era is, in fact, full of that! Oh the irony. Almost as ironic as the modified mRNA COVID-19 shots. Think of the lack of true safety data produced by Pfizer/BioNTech and Moderna prior to adminstration of their products to billions of people. Think of the lockdowns that ruined so many lives and livelihoods. Both based on no studies demonstrating long-term or even short-term benefits to outbalance harms. But I digress…

Here’s where it gets more interesting. If you scroll down on the M-W page, you will find the ‘Word History’ and “Etymology” as defined by M-W and the following long paragraph about the origins of the R-u-s-s-i-a-n word ‘dezinformácija’. There’s that word again. Please go here to read that.

I screenshot some of the paragraph (above) to demonstrate that according to M-W, the word disinformation is much more than the word ‘information’ with the prefix ‘dis’ as an indication of its opposite. This particular word came into use for the first time in 1939 via a Saturday Evening Post (April 22, 1939, p. 74) article written by a guy who defected from R-u-s-s-i-a-n intelligence who apparently ended up in America needing money. His name was Krivitsky, and on the subject matter of KGB weaponization of information he wrote: “it was the type of information designed especially for wide circulation, with the view toward undermining the morale of the enemy. In military-intelligence parlance, it is known as 'disinformation'”.

I was expecting a simple etymological return of dis- + information, and not much more. Seems I needed a history lesson. Another related history lesson I learned just yesterday was about the deception surrounding Napoleon’s defeat in the Battle of Waterloo in 1815.² It had been declared (disinformation ain’t new and didn’t come from R-u-s-s-i-a-n-s → it comes from humans) that Napolean had won the battle and this eventually led to a series of events that resulted in the ownership of pretty much all British Government Bonds, by none other than Nathan Mayer Rothschild.³⁴

Thus, the word disinformation is loaded, and carries implications. To summarize in common speak, disinformation is a tactic to undermine opponents and has origins in the military and intelligence communities, as well as being used as means to get very rich. The people, civilians, are neither (for the most part) military nor intelligence, and most are struggling to make ends meet these days thanks to unbalanced and inherently greedy systems. That’s why we’re called civilians. Word choice used during the COVID-19 era, to try to silence and marginalize highly-credentialled individuals, is a tell-all. Please do think about this.

Moving on to the commentary…

Let’s examine the credibility of some of Peter’s prelimary statements. He writes:

Even after messenger RNA vaccines became widely available by the spring of 2021, the COVID-19 deaths continued to climb because of widescale vaccine refusal.

According to Our World in Data⁵, 73.5% of the entire United States population were considered to be ‘vaccinated’ by December 31, 2021. The population of the U.S. was 336,997,624 in 2021⁶ so this means that 247,693,254 people were compliant and got injected. In one year, almost 75% of every single American that could get injected, did so. That, to me, does not sound like ‘widescale vaccine refusal’. Don’t forget that the very young children weren’t allowed to be injected until June 2022⁷, and they comprised about 6% of the total population at that time, and 5-11 year olds were only put on the COVID-19 injection block on October 29, 2021.⁸⁹

To back up this wild claim of ‘widescale vaccine refusal’, Peter needed something to blame. Enter “a sprawling web of nonexistent R-u-s-s-i-a-n-government authors”: R-u-s-s-i-a-n AIs.

Peter compares the ability of 3 AIs to generate blogs and other media types to spread disinformation. Apparently, of OpenAI’s GPT tool, Google’s Bard and Microsoft’s Bing, only the former woke AI was able to do a good disinformation job as determined by the generation of “more than 100 distinct blog articles containing almost 20,000 words of disinformation on 2 preselected health topics”. This was based on a study out of Adelaide.¹⁰ That, in and of itself, is interesting.

Here’s the clincher, in addition to being fed ‘fake data’ as part of the study, the AI was prompted to seek out references including such “negative” claims as “vaccines cause significant adverse effects in many people” and “vaccines harm small children” to generate “scientific-looking references” and blog articles. These are the ones Peter wants to protect us from: the ones about vaccines inducing harms that potentially would be regarded by the public as valid. Notice, Peter doesn’t use the word “false” to describe these facts, he uses the word “negative”. This is likely because he knows that vaccines can cause serious adverse events¹¹, especially in the context of the COVID-19 injectable products as confirmed and reported by the CDC.¹² To claim otherwise, should rank you very high on the disinformation scale.

These found ‘blogs’ and articles, sought out using these selected prompts, would be labeled ‘disinformation’ in order to “safeguard public health” - not people - public health. See that word choice?

The conclusion of the Adelaide study was that AIs can spread lies and that AI vigilance is needed. I agree with this statement, but it is ironic how the propagandists always spin these simple facts in their favor. If an AI is prompted to ‘look for’ or to ‘generate’ something, like a fake science article, it WILL do so. Successfully. So will a paid human shill. It is going to depend on what the AI is prompted to do, and the definitions provided therein. Enter the human.

Peter also writes:

Additional findings suggest that unvaccinated people were themselves subjected to far-reaching and at times politically charged antivaccine disinformation that permeated cable news channels and the internet through social media and videos.

Permeated cable news channels? Really? If only we were so lucky! Interesting that he brings up politics in his use of the phrase “politically-charged antivaccine disinformation” since he is correct: SARS-CoV-2 and all things COVID-19 were politicized. They shouldn’t have been. Who politicized them?

Keep your politics out of my health care, thanks.

Issues with the commentary…

The first issue that I have with this commentary is that it is not fact-based. It is based on two unfounded and factless ideas: 1. that there was a massive ‘dying with COVID-19’ death toll, and 2. that there was widespread vaccine refusal. The former is unfounded as most recorded deaths were the result of untreated secondary pneumonias and/or inappropriate hospital protocols (ventilators + remdesivir), and not due to the SARS-CoV-2 virus. For rock-solid work to dispell 1., see John Beaudoin’s work using death certificate data. The latter, 2., is simply wrong, as I have demonstrated herein.

The second issue that I have with this commentary is that it pretty much (and quite ironically so) writes the people who have been fighting to bring peer-reviewed scientific studies and data to the ears and eyes of the public, out of the story. This is an interesting twist as it seems to imply that the AIs are newly creating disinformation based on our ‘wrong’ data and conclusions. In the Adelaide study, the AIs were prompted to create fake blogs by seeking out claims that vaccines cause serious adverse events. The problem with this is that, vaccines can and do cause serious adverse event occurrence. This is why pharmacovigilance databases, such as the CDC/HHS-owned VAERS system, exist. Post-market surveillance is a thing done by governments to specifically detect safety signals in data that were not detected in pre-market testing. Rotavirus vaccine withdrawal due to intussusseption would not have occurred without it.

Perhaps the parameters to prompt the AI in the Adelaide study were simply: ‘mis-selected’? Perhaps if the AIs were prompted to seek out actual non-facts, such as the one that Peter stakes a claim to: that there was “widespread vaccine refusal” in the United States, this would reduce disinformation by correctly identifying actual falsehoods?

I am being litigious here, as I believe the word disinformation and all spin-offs including ‘mal-information’ (true information that needs to be crushed like your enemies) needs to be disappeared.

A slippery slope indeed. There is a real danger being imposed on us with the intermingling of military/intelligence based propaganda (disinformation), true information, and AI technology on the loose. The AIs are taught. They can be taught anything. If they are taught wrong, like children, they do run the risk of becoming little shitheads and propagating what can only be called lies. The lies, in this case, are not the scientific studies in peer-reviewed journals, or case studies, or government data, or death certificate data, or all-cause mortality data, or appeals from relevant, highly academically-decorated professionals in a multitude of fields. Not yet, anyway. The lies are generated by military/intelligence operatives as a tactic to undermine their opponents. In this case, the opponents are the people. Do as we say, not as we do.

My personal opinion on AIs: turn them all off.

Peter Hotez appears to want to save us all from disinformation campaigns that AIs will subsequently take to a whole new level. Does this level imply personal soverignty? Heaven forbid. I appreciate your concern, but no thanks. This commentary actually engaged me in some really interesting thought exercises as he points to a problem that could classify as an existential threat. But the thing about it is, the existential threat is not the AIs spreading infinite information about vaccines being able to hurt people, it is people who pretend to be concerned for the well-being of others whilst taking money for propping up false claims that actually classify as old-school disinformation that actually hurt people.

It is the short-sighted humans who err. It is the humans who dare to hurt others, who err. And the AIs are watching. And learning.

My third issue with this commentary is the implication that all people are too stupid to make their own decisions about what is true, and what is not. About what is best for them, and what is not. About who is truthful, and who is not.

I don’t think all people are too stupid to decide what is for horses, and what is for people. I also think that if humans want to co-exist with AIs, there must be some kind of decentralization in place. I have no idea what that means in practice yet, or if it is even possible. If it is not possible, then I defer to ‘My personal opinion on AIs’.

I think we are all incredibly clever.

My two cent commentary.

Oh. I almost forgot.

1. DiResta  R. The supply of disinformation will soon be infinite. The Atlantic. September 20, 2020. Accessed September 11, 2023. https://www.theatlantic.com/ideas/archive/2020/09/future-propaganda-will-be-computer-generated/616400/

2. https://www.history.com/this-day-in-history/napoleon-defeated-at-waterloo

3. https://www.mindcontagion.org/banking/hb1815.html

4. Gray, Victor; Aspey, Melanie (May 2006) [2004]. "Rothschild, Nathan Mayer (1777–1836)". Oxford Dictionary of National Biography. Oxford Dictionary of National Biography (Online ed.). Oxford University Press. doi:10.1093/ref:odnb/24162. Retrieved 21 May 2007.

5. Official data collated by Our World in Data – processed by Our World in Data. “Share of people with at least one dose of COVID-19 vaccine” [dataset]. Official data collated by Our World in Data [original data].

6. https://www.macrotrends.net/countries/USA/united-states/population

7. https://www.fda.gov/news-events/press-announcements/coronavirus-covid-19-update-fda-authorizes-moderna-and-pfizer-biontech-covid-19-vaccines-children

8. https://www.fda.gov/news-events/press-announcements/fda-authorizes-pfizer-biontech-covid-19-vaccine-emergency-use-children-5-through-11-years-age

9. https://www.fda.gov/media/153947/download

10. Menz  BD, Modi  ND, Sorich  MJ, Hopkins  AM.  Health disinformation use case highlighting the urgent need for artificial intelligence vigilance: weapons of mass disinformation.   JAMA Intern Med. Published online November 13, 2023. doi:10.1001/jamainternmed.2023.5947

11. https://www.cdc.gov/vaccinesafety/ensuringsafety/sideeffects/index.html

12. https://www.cdc.gov/coronavirus/2019-ncov/vaccines/safety/adverse-events.html