On Propaganda Part 2

A Twitter argument over a shared interview prompted this sequel on propaganda.

Dr. John Ioannidis, a Stanford meta scientist who audits how well studies are constructed, is the epidemiologist most worth trusting on Covid data.

The case fatality rate is converging toward flu levels for the general population but is far worse for at risk groups, according to Ioannidis's interview.

A Twitter reply falsely claims millions died from the economy during the Spanish flu, then twists his shared interview into an admission that policy must proceed without data, though the interview actually argues quarantine could not have stopped that pandemic's spread.

His real point, backed by the data, is that infections are far more widespread and far milder than assumed, and strong confidence despite little real knowledge is exactly what his prior video defined as the propaganda test.

Pandemics strike only every twenty to thirty years and each one is a unique snowflake, which is why even experts have little accumulated experience treating them, and why Ioannidis is praised for having no political agenda and simply following the data.

Antibody studies already showed one to three percent of the population had been exposed before March quarantines even began, undercutting the Spanish flu comparison, while Italy's case fatality rate is skewed since deaths with other serious conditions were still counted as Covid deaths.

Borrowing Rumsfeld's known and unknown framing, the known costs of quarantine already include a suicide rate that rises about one percent for every one percent rise in unemployment, growing domestic violence, mass job losses, and delayed elective medical care that is itself causing suffering and death.

The quarantine is premised on an unproven twenty to fifty million death Spanish flu scenario now contradicted by data points like Sweden and the Stanford antibody studies, with real evidence based drug trials expected by May or June, unlike the observational fast medicine relied on so far.

He compares it to choosing between getting cancer today or twenty years from now, since most people would rather get sick later once treatments and knowledge have improved, and applies that same logic to the timing of this pandemic.

Government's real job is weighing known risks against unknown consequences dispassionately, and the argument that reopening means wanting people to die ignores that ordinary activities like driving already create statistical risk to others.

He illustrates the risk with unknowingly giving the flu to his ninety one year old grandfather, a Marine veteran with asbestos damaged lungs, who could die from an infection Hoskinson himself would barely notice.

He cites Dan Crenshaw's tactical retreat framing for the original lockdown decision, but says real world data now show empty ICUs far below the CDC's projected caseloads, and predicts officials will still claim the Imperial College model would have been right had nothing been done.

The actual study argues little could have stopped the spread regardless of policy, splitting outcomes into a young and healthy pool with flu like results and an older, unhealthy pool with a case fatality rate above five percent, and argues targeted protection like locking down nursing homes could have replaced the trillion dollar blanket shutdown.

He calls the blanket lockdown anti science given no data supports the Spanish flu death toll, and dismisses finding a single healthy thirty five year old who died, or a teenager with cancer, as a strawman numbers game that ignores the far larger pattern.

Confident opinions on topics people barely understand are installed by propaganda rather than reasoned out, and the constant messaging demands accepting rising suicide, domestic violence, divorce, and economic damage, alongside eroding constitutional freedoms of assembly, religion, and speech, without question.

YouTube's CEO announced videos contradicting World Health Organization guidance would be censored, a policy that would have caught his own earlier video on human to human transmission before the WHO itself acknowledged it.

Epidemiological models are riddled with bad inputs and unrealistic assumptions, the same problem he sees in backtested trading models that fall apart once real capital starts moving the very market they were built to predict.

Once people know a disease is spreading they start social distancing, washing hands, and wearing masks on their own, a behavior shift the models never accounted for, which is why predicted deaths were off by orders of magnitude, millions in some places instead of tens of thousands.

He recommends Scott Page's book The Model Thinker, from the University of Michigan, on how to build models properly, since no model ever perfectly represents reality and every added assumption pulls it further from the truth.

Media and policymakers treat these flawed models as holy canon and attack anyone who produces contrary evidence, while the real known consequences, mass poverty, homelessness, abuse, imprisonment, and lost access to health care, could scale to millions or billions worldwide.

Vaccine efficacy is in doubt since viruses mutate, so a vaccine matched to the original Wuhan strain may miss the version actually circulating by the time it ships, and it is still unknown how long immunity to this virus lasts, though the original 2003 SARS immunity lasted a few years before fading in one small study.

Some earlier coronavirus vaccine candidates actually worsened the disease through antibody dependent enhancement instead of preventing it, and contact tracing hits a combinatorial explosion of contacts in dense cities, workable in low density places like Cheyenne, Wyoming, while Apple and Google are building phone based tracing with no stated end date for when it turns off.

Society has already signed up for enormous costs, lost constitutional rights, higher suicide and domestic violence, catastrophic economic damage, based on models that were never correctly built and have already been proven wrong, yet no one goes back to reassess them; the same dynamism, people changing behavior once they know what is happening, is what makes modeling diseases as hard as modeling trading or games.

Doctors and nurses are being laid off during a supposed pandemic because elective procedures are frozen and hospitals sit empty, while data from New York City, Santa Barbara, Sweden, and Amsterdam all contradict the doomsday model's predictions.

Norway's lower Covid death rate compared to Sweden ignores population density; a fairer comparison is Wyoming, whose density resembles Norway's and which has one of the lowest Covid death rates in the world.

He urges people to think critically for themselves and be honest about the limits of their own expertise, warning that propaganda victims lash out at others on Twitter and Reddit with lazy labels like team red or team blue.

Highly qualified domain experts like Ioannidis get attacked despite having done the actual research themselves, while medical school professors have separately criticized remdesivir studies, a drug that began as a failed Ebola treatment around 2014.

University of Vermont pulmonologist Josh Farkas, on his PulmCrit blog, called an early remdesivir study a dumpster fire over its flaws and conflicts of interest, and drug efficacy also depends on where in a patient's disease progression it is given, from prevention through ICU ventilation.

Randomized controlled trials are hard to design fairly since no two patients are truly comparable, an extreme illustrated by Josef Mengele's twin experiments, and he says the jury is still out on remdesivir while hoping it succeeds as a prophylaxis for frontline health care workers.

Preliminary remdesivir studies may have cherry picked patients likely to recover regardless of treatment, and he criticizes people who watch fifteen minutes of an hour long interview before tweeting attacks claiming domain expertise.

Domain competence is not transitive, since being a brilliant doctor does not make someone qualified on physics, and he traces the repeat it until it is believed propaganda technique, seen in the endlessly repeated Spanish flu death toll, to Joseph Goebbels' big lie.

Anyone advocating for reopening the economy is being smeared as anti science, a Nazi, a Confederate, or a racist, and his informal survey of media coverage of reopen protests found imagery of Nazi and Confederate flags used to frame the movement.

The right response to positive Stanford style data should be relief that there is a path out, not accusations of having blood on your hands; roughly half a million people die from influenza worldwide every year on average, and he compares the overlap between flu deaths and Covid deaths to Hiroshima survivors who fled to Nagasaki only to be killed by the second atomic bomb.

He warns that prolonged lockdown risks a global depression, wars, and famine, points to Rwanda, Sudan, and Somalia as examples of what happens when social and governmental order collapses, and closes by urging people to resist propaganda and not give away their freedoms.

He signs off with sympathy for families affected by the coronavirus and optimism that the crisis will push society toward better ways of solving problems together.