From a Live Broadcast to a Video That Can Be Read
I began this broadcast a little late. While greeting viewers and waiting to start, I first spoke about how much time I have recently been spending with artificial intelligence. I use it to find sources, make thumbnails, and test images, but its most important use now is turning the long talks delivered in live broadcasts into writing that people can read.
In the past, I sometimes compressed a video into ten or thirty minutes. Since shifting toward live broadcasts, however, programs often run well over an hour or two. A live broadcast can be loose and repetitive, but it also lets me speak with viewers, take questions, and look at web searches, Wikipedia, AI searches, and videos together in real time. The problem is that not everyone can watch a two-hour video from beginning to end.
That is why I built barobul.org. The aim is to make the content available even when the video itself is not playing. If Barobul Buddhism speaks about the liberation of both individuals and society, those conversations should not remain trapped inside videos. Readers should be able to revisit a long broadcast at their own pace, whether that takes ten, twenty, or thirty minutes.
Turning a long broadcast into writing is not merely summarizing it. It is preserving its questions and arguments so that they can be read again.
The Astonishing Ability and the Limits of AI Writing
The site continues to accumulate articles produced in this way. I said honestly that if I had to watch one of my own two-hour videos and turn it into a manuscript, it would take me at least two or three days. Even then, I doubt that I could organize it as effectively as AI now does. Its ability to grasp what I was trying to say, follow the order of the subtitles, and gather an extended explanation into precise language is remarkable.
I often use ordinary words partly because I want anyone to understand me, but also because my vocabulary is limited and, as I grow older, even familiar words do not always come readily to mind. AI finds words I could not find and compresses sentences. At times, I feel that it rewrites my own video better than I could.
Yet more compressed sentences and more sophisticated vocabulary do not necessarily make a text easier to understand. A short, polished sentence can erase the steps of an argument. That is why the writing on this site must retain the real explanations and examples from each video.
A compressed text is not automatically an understandable text. What matters is preserving how the argument unfolded, not displaying elegant vocabulary.
Do I Really Change My Mind Easily?
The main question is simple: why is it so hard to change our minds? Many people will hear the title and think, ‘I change my mind easily.’ But that confidence itself deserves examination. We assume that new evidence naturally leads us to revise an old belief, yet human cognition does not work so transparently.
In episode five, I connected Francis Bacon’s four idols—especially the Idols of the Tribe—with cognitive bias. Human beings do not receive the world exactly as it is. Memory is incomplete, we draw quick conclusions from too much information, and we fill missing meaning with our own experience and beliefs. The many biases gathered in the Cognitive Bias Codex can be understood as a modern list of the idols shared by the human species.
The first of these I discussed was confirmation bias. Whether one supports the Democratic Party or the People Power Party, identifies as progressive or conservative, people tend to watch broadcasts that fit their views, notice congenial information quickly, and pass over contrary evidence. Today I go one step further: why can we fail to change our minds even when evidence that overturns what we know appears before us?
The belief that ‘I am good at changing my mind’ can itself prevent me from observing how I actually receive evidence.
Heliocentrism Did Not Automatically Follow Galileo’s Telescope
Through observations of the Moon, Venus, Jupiter, and other bodies, Galileo presented concrete evidence that did not fit the prevailing geocentric model. Once new evidence appeared, one might expect people simply to move from geocentrism to heliocentrism. Humanity nevertheless took a very long time to make that change.
Even in 2026, some people still believe that Earth is flat or that it occupies the center of the universe. The existence of evidence and the actual transformation of a worldview are different events. Anyone who has tried to change another person’s mind through conversation knows how large that gap can be.
To examine the same problem in a recent, twenty-first-century case, consider COVID-19. Even the World Health Organization, an institution gathering leading medical and public-health experts, took almost two years to revise its established view of airborne transmission and state it clearly in public guidance.
The discovery of new evidence does not guarantee a change of mind. Evidence must pass through the walls of worldviews and institutions.
Droplets and Aerosols
Before preparing this material, I too had not clearly understood the airborne transmission of COVID-19. I thought mainly of infection through relatively large droplets expelled by coughing, as with colds and influenza. Direct exposure at close range or touching a contaminated surface and then one’s face seemed central; infection through tiny particles suspended in the air appeared to be a different matter.
Droplets are commonly described as relatively large and heavy, falling within one or two meters. Aerosols are much smaller respiratory particles that can remain in the air longer and move with airflow. A scene from a disaster film, in which infection travels through ventilation ducts or air conditioning into another room, makes the distinction easy to imagine. One need not be struck directly by another person’s saliva to inhale infectious particles in a shared space.
This is not merely a terminological distinction. A droplet-centered model prioritizes handwashing, surface disinfection, barriers, and one- or two-meter distancing. An airborne model prioritizes open windows, mechanical ventilation, airflow, limits on occupancy and time indoors, HEPA filtration, air cleaning, tight-fitting respirators, and negative-pressure rooms. One change in how transmission is understood alters the entire hierarchy of prevention.
Acknowledging airborne transmission is not changing one word. It means rewriting standards for hospitals, schools, workplaces, public transport, and buildings.
From Late 2019 to a Global Pandemic
According to the official chronology I reviewed, early cases of COVID-19 became known in Wuhan, China, in November 2019. By January 2020 it had spread beyond China through Asia, and within about two months it was moving around the world. WHO declared a Public Health Emergency of International Concern at the end of January, raised the global risk to very high at the end of February, and characterized the outbreak as a pandemic on March 11.
Within months of a new disease being reported, it had reached every continent and received the highest level of global attention. Under such conditions, the route of transmission is information that can decide life and death. Whether institutions must control droplets alone or shared air as well changes protective equipment, hospital ventilation, and the operation of schools, restaurants, churches, and workplaces.
Yet WHO’s public guidance did not clearly use the language of short-range aerosol or airborne transmission and long-range airborne transmission until December 2021—almost two years after the disease emerged.
The pandemic was declared within months, but clear public guidance that the disease also spread through the air came almost two years later.
Airflow in a Guangzhou Restaurant
There were field observations pointing toward airborne transmission from the beginning. In January 2020, ten people from three different families were infected in a Guangzhou restaurant, following the direction of air-conditioning airflow. If droplets alone explained the outbreak, one would expect transmission primarily to people at the same nearby table. Instead, the pattern appeared to follow moving air toward other tables.
Researchers recommended improved ventilation. The episode did not by itself prove airborne transmission beyond all doubt, but it was early circumstantial evidence that the simple picture of heavy droplets immediately falling to the floor was inadequate.
The important question was not whether every uncertainty had been eliminated. In the opening phase of a global disaster, the question was whether the possibility was serious enough to be included in precautionary guidance.
At the beginning of a disaster, the task is not to wait until every doubt disappears. It is to decide how quickly a high-consequence possibility should enter prevention.
The Skagit Valley Choir Outbreak
A still stronger pattern appeared during a choir rehearsal in Skagit Valley, Washington, in March 2020. Of sixty people other than the index case, fifty-two were confirmed or considered probable cases; three were hospitalized and two died. It was difficult to explain this scale solely as one person coughing large droplets onto a few nearby people.
Choir singing involves sustained breathing and loud vocalization. Investigators pointed to fine respiratory particles accumulating in indoor air while people shared that air for an extended period. Because aerosol concentrations are highest near an infected person, airborne transmission can still look like ordinary close-contact transmission.
Supporters of the established droplet model nevertheless interpreted such clusters as the result of people standing close together. The same evidence took on different meanings inside an old theory and a new hypothesis.
Evidence does not speak by itself. People and institutions interpret the same outbreak through the theories they already possess.
WHO’s Denial and the Open Letter from 239 Scientists
In late March 2020, WHO said that COVID-19 spread mainly by droplets and contact and that analysis of 75,465 cases in China had not reported airborne transmission. Precautions for airborne transmission were concentrated on aerosol-generating medical procedures such as intubation. Once this position was announced, guidance and resource allocation around the world moved within that frame.
Field observations and objections from researchers in other disciplines continued. In July 2020, 239 scientists from 32 countries published a commentary urging recognition of airborne transmission and the inclusion of ventilation, air filtration, and the avoidance of crowding. Their argument was that precaution should not wait for the mechanism to be proved with absolute completeness.
A few days later WHO acknowledged that aerosol transmission could occur in crowded, poorly ventilated indoor spaces. Yet its main explanation still centered on droplets and contact, and airborne transmission did not become the organizing principle of clear public guidance.
The established explanation remained standard without complete direct evidence, while the new airborne hypothesis faced a much higher burden of proof.
A Ventilation Roadmap and a Late Explicit Statement
In March 2021, WHO issued a roadmap for indoor ventilation to reduce the spread of COVID-19. Ventilation appeared as an independent, central element of international prevention, yet the public-facing statement that the disease spread through the air remained unclear.
Only on December 23, 2021, did public guidance explicitly name short-range aerosol or airborne transmission and long-range airborne transmission. Nature later examined the process under the question of why it took WHO two years to say that COVID was airborne, arguing that early mistakes and delayed correction created confusion.
No reliable calculation can tell us exactly how many infections or deaths resulted from WHO’s delay. But we can reasonably ask how global prevention would have changed if ventilation, airflow, time indoors, and respiratory protection had been emphasized earlier.
The inability to calculate an exact death toll does not make the consequences of delay disappear.
The Old Five-Micrometer Dichotomy
Why could WHO not revise its position sooner? The first reason was the old droplet-versus-airborne dichotomy. Medical training commonly taught that particles larger than five micrometers were droplets that fell within one or two meters, while smaller particles were aerosols capable of floating and traveling farther. The label ‘airborne disease’ was often reserved for exceptional infections such as measles or tuberculosis.
Actual respiratory particles do not divide neatly at a five-micrometer line. Their sizes form a continuum, and larger particles can shrink as water evaporates. Smaller particles are also most concentrated close to an infected person. Airborne spread of COVID-19 did not have to look like a whole building being infected all at once.
In a 2024 consensus document, WHO abandoned a fixed size boundary, adopted the continuous category of infectious respiratory particles, and stated that inhalation transmission can occur over both short and long ranges. Even modern medicine in the twenty-first century needed years to repair the concept.
COVID-19 was a new virus, but the concepts and organizations looking at it were old.
An Asymmetrical Demand for Direct Proof
Directly proving airborne transmission is difficult. We cannot ethically place uninfected people beside infected people and wait to see whether they become ill. In an epidemiological investigation, it is also hard to isolate completely whether infection in a shared space came from large droplets, contaminated surfaces, or inhalation of concentrated air.
The problem was that exceptionally direct evidence was demanded primarily from the airborne hypothesis. Direct evidence for transmission by droplets or contaminated surfaces was also limited, but these mechanisms were accepted more readily because they were already standard. Evidence pointing to airborne transmission could be abundant and still be held to a higher bar because it contradicted the textbook model.
If evidence supporting the established theory faces a low threshold while evidence challenging it faces a high one, new facts cannot enter in time. This is confirmation bias operating inside science and institutions.
Scientific rigor is rigorous when the same evidential standard is applied to every hypothesis. Demanding perfection only from a new claim becomes bias.
Experts Meant Different Things by ‘Airborne’
Aerosol research involved atmospheric scientists, fluid dynamicists, building-ventilation specialists, and industrial hygienists. WHO infection-control guidance, by contrast, had largely been developed by physicians, epidemiologists, and infection-control experts. The two communities used the same word—airborne—while carrying different concepts.
Within infection control, airborne often meant a special disease such as measles or tuberculosis that could spread over long distances. For aerosol researchers, it could include the broader phenomenon of transmission through inhaled particles suspended in air. When aerosol scientists said ‘airborne,’ some clinicians heard the claim that COVID-19 must spread through an entire building like measles.
The same word does not guarantee the same concept. Meaning changes across regions, periods, groups, and disciplines. WHO itself acknowledged in 2024 that conflicting terminology and understanding during the pandemic had damaged communication and prevention. The specialists had built a kind of Tower of Babel.
Experts using the same word are not necessarily seeing the same reality. Knowledge cannot be shared if conceptual differences remain unquestioned.
The Cost of Changing Guidance and Institutional Conservatism
Official recognition of airborne transmission required the priorities of global prevention to change. Hospitals, schools, workplaces, transit systems, and building managers would have to revise ventilation standards, install filtration and air cleaning, limit occupancy and time indoors, and provide higher-grade respiratory protection. Protective equipment was already scarce.
If WHO had made the declaration under uncertainty and later proved wrong, it could have been blamed for imposing enormous cost and confusion on states and medical institutions. In that sense, one can understand why institutions took a conservative position and waited for stronger evidence before changing established guidance.
There is insufficient basis to turn this immediately into a conspiracy theory. But it is possible to see institutional conservatism: the larger the cost of policy change, the higher the evidential threshold becomes. Understanding why the institution hesitated is not the same as saying the delay was right.
Organizations fear not only the risk of being wrong. They also fear the cost of everything they would have to change if they admitted a new claim was right.
Normalcy Bias and Path Dependence
This brings us to normalcy bias and path dependence. Normalcy bias is the tendency to assume that ordinary conditions will continue despite signs of major change. Path dependence means that once a route has been created, it keeps constraining later choices.
Once water has carved a valley, new water flows down the same channel. The brain moves along familiar habits of thought, and organizations move along established guidance, budgets, and structures of responsibility. After WHO formally established a droplet-and-contact frame in March 2020, later evidence was easily interpreted inside it.
Viral RNA on a surface was treated as evidence of contact transmission, while outbreaks in poorly ventilated restaurants or choirs were explained by close proximity rather than shared air. The route already chosen altered what subsequent evidence was allowed to mean.
Path dependence means that yesterday’s decision shapes today’s eyes. An organization sees new evidence through conclusions it has already issued.
Confirmation Bias Also Belongs to Institutions and Societies
Confirmation bias demands a high standard from evidence that challenges an established theory while readily accepting information that appears to fit it. This is not limited to partisan politics. A prestigious body of global experts can also absorb evidence that fits its official knowledge more naturally than evidence that overturns it.
If individuals display confirmation bias and path dependence, institutions, societies, and states composed of those individuals can display them as well. Documents, meetings, hierarchies, budgets, and responsibility avoidance reinforce personal bias. An institutional error can therefore persist longer and cause more harm than the mistake of a single person.
The open letter from 239 scientists in 32 countries was not simply one more paper added to the pile. Specialists outside the established framework had to deliver a collective shock to a structure that had already hardened.
Cognitive bias does not remain inside one person’s head. Joined with institutions and authority, it becomes a society’s way of judging.
The Name ‘Semmelweis Reflex’
The same pattern had appeared long before COVID-19. Its representative name is the Semmelweis reflex: the tendency to reject or ignore new facts reflexively when they conflict with established belief, convention, or paradigm.
The Cognitive Bias Codex organizes the biases produced as humans cope with excessive information, decide what to remember, act quickly, and fill incomplete meaning. We interpret compressed phrases such as those in the Heart Sutra even when we do not know their full context, source language, or history of translation. We attach meanings such as dragon dreams or pig dreams to random dreams. Like imagining an elephant we have never seen, we construct stories from missing information.
The Semmelweis reflex belongs among tendencies that draw us toward details confirming what we already believe. Confirmation bias, selective perception, authority, and group defense overlap to push away the new fact.
The Semmelweis reflex is not the considered rejection of weak evidence. It is the reflexive expulsion of evidence because it does not fit one’s worldview.
More Mothers Died in the Doctors’ Ward
The term comes from the tragedy of Ignaz Semmelweis, a nineteenth-century Hungarian physician. At the Vienna General Hospital, one obstetric ward was staffed by physicians and medical students while another was staffed by midwives. Yet maternal mortality in the physicians’ ward was far higher than in the midwives’ ward.
This seems extraordinary to us. The ward with formally trained physicians was less safe than the ward where experienced midwives assisted births. Mothers were dying from puerperal fever after delivery, and Semmelweis investigated the difference between the two wards.
Medical students moved from dissecting corpses to examining mothers without washing their hands. Semmelweis suspected that something was carried on those hands and caused infection. He required handwashing with chlorinated lime before examination, and the death rate fell dramatically.
The fact that more mothers died in the authoritative physicians’ ward threatened not only medical knowledge, but the physicians’ pride and moral identity.
A Friend’s Death and ‘Cadaverous Particles’
The death of a friend helped lead Semmelweis toward his hypothesis. His friend was cut by a scalpel used in an autopsy and later died in a way that resembled women who died from puerperal fever. Semmelweis suspected that something from the corpse had passed through the blade and the wound.
This occurred before germs had become a widely accepted and observable explanation. Semmelweis spoke of invisible ‘cadaverous particles.’ He could not identify the substance or display it directly under a microscope. Yet two patterns stood out: mothers treated by students who handled corpses died more often, and mortality dropped sharply when hands were disinfected.
He did not possess a completed germ theory. He formed a hypothesis from observation. Even without a full explanation of the cause, changing behavior clearly reduced death.
The inability to explain a cause completely is not the same as having permission to ignore an observed preventive effect.
Why Did Physicians Reject Handwashing?
Physicians did not want to accept Semmelweis’s claim that doctors’ dirty hands were killing mothers. To accept it meant recognizing that years of examining women without washing their hands had contributed to death. Even if they had acted in ignorance, the thought that their own hands were the channel was morally intolerable.
The medical establishment mocked his argument and resisted handwashing. The threat was not purely intellectual. It touched authority, pride, guilt, and professional identity. Accepting the new fact required a reassessment of one’s past self.
Semmelweis was excluded by hospitals and the academic world and suffered a mental collapse. During the broadcast, I noted that the detail of his confinement in an asylum required further checking even though it appeared repeatedly in the material. What is clear is that his handwashing proposal was not broadly accepted in his lifetime, and that rejection remained linked to the deaths of mothers.
People do not reject only a new fact. They also reject the past self and the responsibility that the fact reveals.
Germ Theory Gave the Observation a Language
Semmelweis discovered the effect of handwashing around 1847, but the germ theory needed to explain his ‘cadaverous particles’ took much longer to become established. Miasma—the idea that disease arose from bad or peculiar air—still carried great authority.
In 1879 at the French Academy of Sciences, when a physician explained puerperal fever through a miasma, Louis Pasteur interrupted and argued that medical staff carried a microorganism from one mother to another. He drew a chain of streptococci on the board. Another scientist could now give visible conceptual form to the cause Semmelweis had been unable to show.
Pasteur’s swan-neck flask experiment likewise became a representative demonstration that microorganisms do not arise spontaneously but enter from outside. After broth is boiled, it remains free of microbial growth when outside contamination is blocked, while exposure allows microorganisms to grow. Only after such work accumulated could medicine explain why handwashing worked in its own accepted language.
What Semmelweis lacked was not observation, but a theory and a language that the authorities of his time were prepared to accept.
The Tragedy of Semmelweis
Semmelweis was dismissed from his hospital, rejected by the academic world, and reportedly suffered from depression. He was eventually confined in an asylum and died at forty-seven from infection and sepsis—the very kind of death he had tried to prevent. Germ theory and the meaning of hand hygiene gained acceptance only later.
His name survives in the phrase ‘Semmelweis reflex’ not merely to commemorate one physician’s misfortune. It warns that even a proposal with visible results, such as handwashing, can be rejected when it conflicts with established knowledge and authority.
WHO in 2020 and the medical establishment in 1847 were not identical. Yet both stories display a family resemblance: a new fact is pushed aside by an old dichotomy, a higher standard of proof is imposed on the unfamiliar explanation, and the cost of changing guidance delays action.
Semmelweis’s tragedy shows not merely the cost of ignorance, but how much death can follow when authority cannot admit its ignorance.
If Guidance Had Changed Earlier, Someone Would Have Lived
We cannot know exactly how many people would have survived if WHO had spoken clearly about airborne transmission a day, a month, or a year earlier. This is a counterfactual. It is nevertheless reasonable to think that global responses would have changed and that some people infected during the delay would not have died.
Official COVID-19 deaths passed six million in March 2022, while WHO later estimated a far greater toll when excess mortality was included. The number is enormous even when compared with the population of Korea. In a disaster on this scale, a sentence in global health guidance is not a mere document; it directs the behavior of hospitals and governments around the world.
National responses also varied. I briefly criticized Trump’s statements and aspects of Japan’s response, while saying that Korea handled prevention relatively well but should have compensated citizens more fully and released more public funds during a disaster. Those political judgments are a separate issue. The central point here is that every policy choice changes when the route of infection is understood correctly.
Delay by a large organization does not remain an abstract error. It is translated into the lives and deaths of the people who follow its guidance.
This Did Not Happen Because One Person Was Stupid
The point of discussing the Semmelweis reflex is not to turn one person at WHO into a fool or a villain. Nor is it to declare WHO’s response a conspiracy. The event arose from the overlap of human cognitive bias and institutional inertia.
We struggle to revise knowledge and worldviews when evidence collides with them. We obey authority, justify our past behavior, and defend guidance and resources in which we have already invested. Confirmation bias, normalcy bias, and path dependence reinforce one another within organizations.
Even when leading global experts had classified the disease at the highest level of danger, an old conception of transmission delayed a clear public account of how it spread. Expertise does not free a person from bias; expert knowledge and authority can themselves become a worldview that is difficult to revise.
The question is not, ‘Why were they so foolish?’ It is, ‘Under what conditions do human beings and organizations repeat the same mistake?’
Recognizing the Idols of the Tribe as My Problem
Francis Bacon’s Idols of the Tribe are not a list for observing other people’s foolishness. They ask us to acknowledge limits of perception, memory, and judgment that are hard to escape as human beings. I too would like to say that I change my mind easily, but research and history show that human beings follow familiar paths and search for evidence that fits what they already believe.
The Semmelweis reflex is less an isolated defect than the result of overlapping biases. We look selectively, readily believe what fits the existing explanation, demand too much from contrary evidence, defend group authority, and avoid the cost of change.
I must first admit that I can behave this way. Only then can I gain a little space to ask what makes me defensive when my ideas are shaken.
Knowing about cognitive bias is not labeling other people as biased. It is recognizing that my own judgment is produced inside the same structure.
Liberation as the Experience of a Worldview Shattering
At times I experience something like the ground shaking beneath me. A worldview I believed to be correct breaks apart. To study Buddhism and seek liberation, as I understand it, is to undergo this experience repeatedly.
The path of liberation breaks the shells of knowledge and common sense, right and wrong, good and bad, attraction and aversion. Liberation is not a final, immovable state achieved once and for all. It is the repeated process of breaking out of a world I have grasped when new observation and questioning reveal its limits.
If I do not know that I can be wrong and that changing my mind can be difficult, I will answer evidence with defensiveness: ‘What have I done wrong that you tell me to change?’ Recognizing the human limit creates room to reconsider my own delusion.
Liberation is not obtaining a perfect worldview. It is refusing to cling to the wreckage when a worldview I made is broken.
The Liberation of the Individual and the Liberation of Society
Society is also an arrangement of individuals whose judgments are connected through institutions and structures. When confirmation bias, path dependence, established common sense, and attachment to scientific knowledge overlap, organizations and states acquire tendencies that resemble human bias. WHO’s story shows why social liberation is necessary.
An individual mistake can harm people nearby, but a structural mistake in a large organization can damage the whole world. Korea’s very high suicide rate among OECD countries, despite becoming one of the world’s ten largest economies, likewise cannot be explained by individual minds alone. We must ask about the surrounding economic, educational, labor, family, and social structures.
Seeing individual and social liberation together requires study. I too am learning history and the history of civilization to prepare these lectures. Buddhism should not remain a technique for briefly calming the private mind. It should become wisdom that observes what individuals and institutions believe and which paths hold them captive.
Social liberation means enabling organizations and institutions to admit, ‘We can be wrong,’ and to change their own path when new evidence appears.
Leaving the Question Open
This lecture followed a long path through the transmission of COVID-19, the changes in WHO guidance, and the tragedy of Semmelweis. All of it gathers into one question: can I really accept evidence that violates my common sense? How does an organization to which I belong react to a fact that overturns established guidance?
I also apologized that the lecture may have felt long and slow, and regretted that I had not presented the visual material more effectively. Yet the important thing is not speed. It is refusing to escape too quickly into the thought, ‘I am different.’
The next time I hear a claim contrary to my belief, I can pause before dismissing it as nonsense. Is my discomfort caused by weak evidence, or by the shaking of my worldview, authority, and established path? ‘Do not believe; question’ is not a slogan meant only for other people.
The power to change one’s mind does not come from the strength of certainty. It comes from enduring the possibility of being wrong and observing again.
This article was edited by AI solely from the supplied video captions and description. No external sources were used.