Note: This is a freely accessible serialized version of Lab Leak Fever. Audio voiceover was AI generated for accessibility. Find an overview of all chapters here or consult the book website for further information.
“I was always the ‘zoonati’ they could talk to.” Professor Stuart Neil, a virologist from King’s College London, chuckled about his perceived diplomatic role for DRASTIC and some other lab leak proponents. “I intrinsically thought this was an interesting story,” he said, remembering how he became intrigued by various ideas of how SARS-CoV-2 could have come about via a research-related accident ever since the Mojiang miner story blew up in the summer of 2020. Scientists are not averse to discussing controversial ideas; quite the opposite, many are attracted by them. Stuart came across some of the online figures who promoted it, and he thought there “were some people of good faith in this grouping.” He still believes they “were at some point,” but now laments that ultimately, even polite discussion and good faith engagements were made impossible by the emergent group dynamics of increasingly radical believers. “Gradually, that all got poisoned.”
Open inquiry and exchange are as important in science as in a democratic society. When my sci-comm colleague Sam and I first interviewed him in November 2021, Stuart had advised us to “avoid people that think they have all the answers.” He was curious about new developments and cautious about drawing hard conclusions. In many ways, his attitude was a good sounding board for some of the more creative lab leak proponents to test their ever-new speculations, at least in the early years.
Stuart’s open attitude in this controversy is representative of the free-for-all, evidence-driven approach characteristic of most scientists. He does not care who makes the argument as long as they bring the evidence to back it up. This was always the power and promise of the internet: to prevent group thinking and empower unusual voices to be heard where traditionally they would not get the chance. Stuart enjoyed entertaining alternative scenarios and engaging in good-faith speculations. He might have often been a bit rough on others and undiplomatic when it comes to nonsense, but he would acknowledge if they had made a solid point. Of course, like the rest of us, he migrated between amusement, bewilderment, and eye-rolling about the quality of popular discussion on the topic.
However, as the scientific evidence continued to stack up against any type of research-related accident and uncertainty about what wild theories could still be entertained in good faith shrank, the fronts hardened beyond repair. Building bridges to reach lab leak believers became impossible. Eventually, even scientists who had previously established good rapport ended up being considered enemies by the lab leak camp. Soon enough, he found his safety threatened, his private communications FOIA’d, and even his kindergarten-age kids were stalked and doxxed. All for the simple sin of following the evidence.
I reached out to Stuart because I felt it was time to tackle the most misunderstood and polarizing element in the whole origin controversy: the infamous furin cleavage site (FCS). What is known, what is still unknown, and what we can learn from it.
The FCS is a short amino acid motif in the spike protein that is recognized by furin-like proteases, which are cutting enzymes. It is a dramatic functional element for those various factions captured by the idea that SARS-CoV-2 was somehow created in a lab. It is alleged to be the pandemic’s secret sauce. A trigger that was artificially inserted to turn an ordinary bat virus into the pandemic blight pathogen we have today.
Indeed, the FCS is critical in SARS-CoV-2; it increases virality and is required for efficient human-to-human transmission. Yet, it’s almost a mythological force in popular discourse that is entirely based on two misconceptions.
First, there is the common misunderstanding that the alleged introduction of a single genetic element has the power to create a pandemic pathogen. Second, and perhaps more importantly, there is deliberate deception about how likely it is that nature or human engineering came up with the nucleotides insertion that led to the FCS we observe in SARS-CoV-2.
We are going to have to get a bit technical to address both points. But it will be incredibly worth it to understand, so bear with us.
As a molecular virologist, Prof. Stuart Neil was well-positioned to talk intelligently about the nuances of this topic; his research lab focuses on host-encoded antiviral mechanisms, and he has investigated the role of the FCS in that context. Viruses need to constantly adapt to beat immune systems since immune systems tend to evolve to beat back viruses. Stuart’s lab studied how “this evolutionary arms race plays out in the context of HIV1 envelope proteins, so the spike equivalent,” and he elaborated on where he traditionally came from before SARS-CoV-2 entered the scene.
The vast majority of enveloped viruses need membrane fusion with a host cell to enter it, which can happen directly at the outside cell wall or after being ingested in a big bubble, also known as endocytosis. Either way, the viral proteins responsible for entering host cells need some help from the host environment.
“In general, a lot of them are activated by having a protease cleavage site midway,” meaning our own protein scissors do the deed for the virus. The virology professor used both hands to illustrate how a cleavage site usually liberates a hydrophobic (water-insoluble) part of SARS-CoV-2’s spike protein. This part “burrows itself into the nearest target cell membrane.” Imagine a harpoon ready to fire as soon as the gun ports are opened. In fact, talking about a single spike protein is a bit imprecise because, on the virus surface, three spike proteins are always intertwined, forming a type of trillium flower with three such spring-loaded harpoons that are ready to shoot once liberated by a protease cut. The activating cuts by a human protease prime the viral proteins to launch their invasion into a new cell. But where does this activation happen?
For many viruses, the most important part of that cutting activity often happens after the virus is already attached to the host receptor, which would be the human transmembrane protein known as ACE2 for SARS-CoV-2. Once the SARS-CoV-2 virion is bound to ACE2 on the outside of a new cell, a human protease called TMPRSS2, a molecular cutter also embedded in the host cell membrane, activates the bound viral spike proteins via multiple cuts. Then the membrane fusion starts.
That’s one way. However, there is another option. For some viruses, such as HIV1 or H5N1 influenza, activating cleavage of their invader proteins can happen even before they leave the production factory, so to speak. Their most critical cutting steps predominantly occur still inside a currently infected host cell, where the next fleet of virions is assembled to start a new invasion.
In this “pre-cleaved” scenario, the viral battleships come out, guns blazing and ready to attack. Stuart explained how that could be quite consequential for cell invasion:
Not only is the virus going out into the space and then getting activated and breathed out, it’s [also] being able to come out very rapidly and infect the next-door cell, or come out the back end of the cell, and that sort of... makes it go bang straight through that epithelial way.
Unfortunately for us, Stuart acknowledged, coronaviruses “live in this happy medium, where they can deal with both [activation scenarios], and that was always a worry.”
Coronaviruses can get both pre-cleaved in the factory or cleaved directly at a new host cell’s door. Even worse, Stuart said, “They can be very promiscuous about what protease they allow to do that.” This is where the FCS comes into play. Furin is a protease that sits inside a cellular compartment called the Golgi apparatus, basically the very last station of protein production where assembled proteins get various modifications, such as sugar shields (glycosylation). Having an FCS motif would allow SARS-CoV-2 to get pre-cleaved in the host cell, right on the way out.
Why that works so well for SARS-CoV2 is not completely understood and was largely unpredictable beforehand. Virology is complex. However, what we have since learned is that pre-cleavage of spike proteins by furin does two very specific things to SARS-CoV-2. First, it opens up the structure of the trimeric spike protein, which makes it a lot less stable in its 3D configuration but allows for better epitope binding to its host receptor, ACE2. Second, it primes the TMPRSS2 cutting work, making this specific route of viral cell entry through direct membrane fusion faster and favored over alternative entry routes.
This matters because it is not only the availability of host cell receptors that define what tissues in our body are most susceptible to a specific virus but also the concentration of activating proteases in the cellular environment. Higher ACE2 and TMPRSS2 co-expression are found in respiratory tissues. These tissues would find themselves especially vulnerable to spike proteins that were pre-cleaved by furin proteases.
Think of the whole thing as an efficiency hack. Pre-cleavage favors TMPRSS2-expressing respiratory cells and enhances virion entry via the direct membrane fusion route, ripping open the host cell membranes to deposit the viral cargo inside. The increased fusogenicity at the outer cell membrane has the additional benefit of dodging some intracellular antiviral defense mechanisms that would come into play via endosomal entry, such as virions being chewed up, sliced apart, or boiled in acid inside the endosome. This efficiency hack overwhelmingly favors respiratory tropism and circumvents some of our innate host defense. The end result of the tiny FCS motif in the spike protein of SARS-CoV-2, through a complex set of molecular processing, environmental factors, and host biology, is that it happened to aid both in pathogenicity and respiratory transmission of the virus, to our detriment.
This outcome was not given by any means. If you switch the host or environmental context, for example, by putting SARS-CoV-2 into human cell culture, the FCS motif is lost very quickly. In that similar yet different human cell context, harboring an FCS destabilizes the spike protein critically and slows down the virus, thus the FCS motif gets outcompeted. Same virus, same genetic elements, different outcome. Only real-world selection pressures during respiratory transmission seem to favor and preserve the FCS motif in SARS-CoV-2.
Stuart cautioned that we should not be fooled by overly simplistic notions, either. “Finding an FCS does not a priori mean respiratory spread,” the virology professor warned. Nor does an FCS turn a random bat virus into a respiratory virus. Multiple changes, large and small, must happen iteratively to reach a point where a bat virus can expand its cellular tropism. This includes stabilizing mutations to compensate for pre-cleavage lability and new structural or conformational changes; gaining or losing glycosylation sites; avoiding or surmounting prior host immunity; finding ways to subvert the cellular machinery of the host; and somehow using new routes for transmission. No researcher anywhere in the world has yet been able to optimize a virus for human transmission in a laboratory setting. The widespread notion that merely introducing an FCS will do the trick, or is even likely to do the trick, is simply false. Context is king when it comes to viruses.
For example, MERS-CoV from 2012 and SARS-CoV-1 from 2003 both caused epidemics, one with and one without an FCS. In the case of SARS-CoV-1, some experiments have shown that the artificial addition of an FCS does not increase virulence or pathogenicity. Additionally, SARS-CoV-1 without an FCS is no less infectious in respiratory transmission to ferrets compared to FCS-containing SARS-CoV-2. Adding an FCS to 96% similar bat virus cousins, such as RaTG13 or Banal-like viruses from Laos, does not turn them into human respiratory pathogens either, although they do tend to show increased fusogenicity. Feline coronaviruses, which actually tend to have a furin cleavage site as a baseline, are usually mild but become much more virulent when they lose their FCS, a direct opposite to SARS-CoV-2. Some of our endemic human CoVs—common cold viruses infecting the respiratory tract—have an FCS but do not even use ACE2 as an entry receptor, while others have no FCS but still use ACE2. The human CoVs without an FCS do no worse in respiratory spread either. The point is that we should not be overly focused on the FCS or any other single genetic element, as Stuart and other virologists have cautioned me over the years. In nature, no genetic element acts alone; they follow their own logic born from constant adaptation and genome-wide optimization.
“It is hard to explain to people that this is not something you design, and it just works,”
Stuart lamented. The reality is that for a pandemic virus, the right genetic elements, host, and environmental conditions have to fall together to create a perfect storm. We should also expect that any natural virus that is capable of starting a pandemic has a lucky combination of some unusual genetic tricks up its sleeve. Otherwise, every ordinary virus would cause a pandemic. So why was there so much fuss about this one genetic element?
As best as I can tell, the main reason why the short FCS sequence in SARS-CoV-2 (a 12-nucleotide out-of-frame insertion encoding four new amino acids) has been hotly discussed is because it stood out like a sore thumb in side-by-side sequence comparisons to known bat coronaviruses in early 2020. The llama in a supposed herd of viral sheep. We are pattern-recognizing creatures, and it was something we had not seen before that looked obviously out of place when lined up with bat viruses. This supposed uniqueness, coupled with the misconception that an FCS has the power to turn an ordinary virus into a pandemic pathogen, has since been perpetuated endlessly as evidence of human villainy.
And where there is villainy, there are those who believe they can become heroes.
Enter Columbia University professor of economics, celebrity academic, and political power broker Jeffrey Sachs. He has been an establishment figure for decades, most prominently involved in the economic “shock therapy” doctrine for post-Soviet Russia transitioning away from a planned to a market economy. “He was preaching privatize, privatize, privatize,” Dr. Carlos Morel, a Brazilian parasitologist and former WHO director who was very familiar with Sachs, once explained to me. Unfortunately, Sachs’s shock therapy mainly caused chaos and suffering, leading to millions of deaths from despair as well as seeding the roots of kleptocratic oligarchy, which he was later harshly criticized for, such as in The Body Economic: Why Austerity Kills. Sachs’s role as advisor to the United Nations Secretary-General, working on sustainability goals and poverty, has seemingly aged a bit better. “In 2001, he chaired the macroeconomics and health book,” Carlos explained. “For the first time, in this report, they showed that investment in health is good for economic development,” which shaped thinking and policies for developing nations. However, many of Sachs’s advocated policy initiatives in Africa, such as the Shiny Millennium Villages project, have had mixed results at best. No matter what Jeffrey Sachs gets involved in, he seems to relish exerting influence on world matters and hanging out with celebrities such as Angelina Jolie and Bono from the Band U2. According to a profile in New York Magazine’s Intelligencer: “Sachs’s ambitions are hard to overstate... His ultimate goal is to change the world—to ‘bend history.’”
Is it any surprise that after a traumatic pandemic, he anointed himself to litigate the origins of COVID-19 as well? The economics professor became the head of the Lancet Commission, which was convened by the prestigious medical journal The Lancet to collect lessons from the pandemic and make recommendations for public health policy. Initially, Sachs was against the lab leak theory, which he believes unjustly blames China, a country he often championed and has some undisclosed affiliations with. Yet it would not take long for him to be taken in by a very specific version of the lab leak myth, forcing Peter Daszak, whom Sachs himself had first invited to head the Lancet’s origin task force, to resign and disband his group.
“He is a very good speaker,” Carlos admitted. “But his role as the head of the Lancet commission… my God,” he shook his head and threw his hands in the air. He was happy to have not been on that commission. In his opinion, Sachs was very authoritarian. “Some of my colleagues who had been in the Lancet Commission said, ‘No, I quit. I cannot be in a commission where the chair orders what to do.” But that seems to be how politics often works these days. With the dissenting experts out of the way, Sachs could exert control over the origin narrative, which, surprisingly for people less familiar with his ideology, ended up pointing the finger of blame not at China but at the US instead, blaming “US biotechnology” for creating the virus. In parallel, he had written an opinion paper for the journal PNAS proposing, based on the most naive and cherry-picked sequence comparison to a human protein called ENaC, “that a ‘molecular mimicry’ between the FCS of SARS CoV-2 spike and the FCS of human ENaC” had taken place. He confidently asserted:
For a research team assessing the pandemic potential of SARS-related coronaviruses, the FCS of human ENaC—an FCS known to be efficiently cleaved by host furin present in the target location (epithelial cells) of an important target organ (lung) of the target organism (human)—might be a rational, if not obvious, choice of FCS to introduce into a virus to alter its infectivity, in line with other work performed previously. Of course, the molecular mimicry of ENaC within the SARS-CoV-2 spike protein might be a mere coincidence, although one with a very low probability.
To experts, the quality of these naive arguments is laughable; they are the equivalent of typing in a random string of six letters into Google and taking the inevitable positive search result as evidence that the word exists. It was the “HIV inserts” level of analysis all over again. “The paper takes a sharp turn and gets into some serious furin cleavage site trutherism,” Prof. Angie Rasmussen commented. Stuart Neil was even more poignant and less indulgent: “All these arguments of intelligent design without any actual evidence to support them are simply creationism by another name.”
Stuart was correct, of course. However, Jeffrey Sachs cared little for the virological details. They were pretense because he had much bigger fish to fry. He gets very explicit about what he wants, arguing:
A US-based investigation need not wait—there is much to learn from the US institutions that were extensively involved in research that may have contributed to, or documented the emergence of, the SARS-CoV-2 virus. Only an independent and transparent investigation, perhaps as a bipartisan Congressional inquiry, will reveal the information.
Sachs has always been critical of “US hegemony” while seemingly being comfortable cozying up to autocratic regimes. In recent years, that has apparently led him further and further towards fringe positions, such as denying the plight of the Uyghurs in China and arguing that NATO is responsible for the Russian invasion and war on Ukraine. Blaming the US and the genetic engineers at the University of North Carolina for the pandemic was very much par for the course. It was rather absurd to propose that US scientists created the virus and shipped or shared it to Wuhan where it broke out of a Chinese lab, even ignoring all the evidence in favor of zoonosis and against any laboratory involvement. But when somebody so influential speaks for the lab leak myth, allies in the media are easily found.
The Intercept would run “Jeffrey Sachs Presents Evidence of Possible Lab Origin of COVID-19” as a headline, charitably eating up Sachs’s creationist opinion paper. For them, it was a logical continuation of the DEFUSE proposal coverage, featuring the same story tropes and villains in the form of Ralph Baric’s lab but now adding a supposed “EnAC mimicry” story element as the supposed “inspiration” for the furin cleavage motif used by American engineers. Jeffrey Sachs, in turn, started doing media rounds, using the publicity around the Lancet Commission report and the opinion paper for PNAS that he had shaped for his self-serving cause.
No matter where he went, the skilled orator got away with claiming that he was “pretty convinced” that COVID-19 came out of US biotechnology. He made quite a few headlines, much to the pleasure of Chinese state actors it seems to me, who giddily amplified his message. In such a media environment full of motivated actors and power plays, finding our way back to evidence and fact-based discourse becomes increasingly difficult. But we have to try.
Stuart agreed on the basic tenet of why the FCS is unusual. “What was never seen among the sarbecovirus family, and this is obviously why it’s controversial, is the presence of one of these things in bats.” But he was not overly surprised by that. In bats, all coronaviruses, including SARS and MERS-related family members, are in general found to be gastrointestinal, not respiratory viruses. That is why Shi Zhengli and others have been collecting bat guano pellets in their multiyear surveys. Furin-cleavage sites in the wider beta-coronavirus family are relatively rare, but they do exist sporadically. Yet, so far, none have been found in close bat virus cousins of SARS-CoV-2, making the FCS unique for this particular branch of the family tree. The llama in the flock of viral sheep.
That being said, we already learned about very suggestive insertions at that critical S1/S2 region, which, for example, Alice Hughes and Supaporn Wacharapluesadee had discovered. The alpaca and guanaco within the viral herd, so to speak. Researchers in Germany around Christian Drosten have since discovered that these insertions at S1/S2 can indeed also contain polybasic cleavage motifs; however, they occur both at very low frequencies and do not seem to be maintained very well in bats. So, as best we can tell, these polybasic cleavage sites pop in and out of existence without giving an evolutionary advantage for transmission in bats. They are short-lived in that particular host and gastrointestinal context.
“I always thought that the FCS that SARS-CoV-2 picked up seems to me like it’s very recent,” Stuart argued. “It is a first pass. A first try. It works well enough as a toehold to respiratory spread.” However, that toehold does not come for free. Stuart’s research has found that these advantages come at the cost of protein stability, and the FCS also seems to play a role in how well pH differences are tolerated. This is because the FCS “is not just liberating the fusion mechanism,” he explained; “there can be long-range conformational effects on the whole of that protein” that might be beneficial or detrimental depending on the context. All these nuances and dynamics are incompletely understood and even harder to explain to non-experts.
“You can’t think of these proteins as rocks,” he advised. “They are living breathing things that are moving around like this,” he said as he waved his fisted hands upside down while explaining how spike proteins are configured very flexibly, “and the receptor binding domain sits on the top of that spike and goes up and down.” These dramatic conformational changes can work to a virus’s favor or detriment, depending on the environment.
Stuart elaborated, “But you know if you're going through an intestinal compartment, you're going to be exposed to a much greater pH and temperature, or whatever it might be in the way an upper respiratory tract is not.” This means that maintaining spike stability in the bat gut may exert significant evolutionary pressure on bat viruses; ergo, having pre-cleaved virions around that are less stable just might not work well in the merciless competition for survival these sarbecoviruses find themselves in. They are short-lived because selection does not favor them. However, if this harsh bat gut environmental context were to ever switch, let’s say after spillover into an intermediate animal, new evolutionary pressures would come into play. That’s where Stuart would put his money—that the FCS can start to grow out and become evolutionary favored. Keep this in mind for later. Because the question posed is not whether nature can come up with it, which it obviously can, but rather whether or not humans had their hands involved in creating it.
Here, my own viral vector cloning experience from my PhD came in handy to judge the evidence. Because from a genetic engineer’s point of view, using the unusual and likely ineffective FCS motif (RRAR) we observe in SARS-CoV-2 never made much sense. Why choose this crippled sequence when one could use a perfectly canonical sequence like RRSRR instead, which would guarantee efficient cleavage? Inserting four new amino acids—rather than substituting existing amino acids—is another thing genetic engineers tend not to do because it unnecessarily risks disrupting the overall protein structure, making the virus unviable. In fact, scientists have introduced furin-cleavage motifs through substitution multiple times, for example, in 2006 in the original SARS virus. Other labs have also worked on this over the years, all using substitutions of existing sequences. No one has ever documented a single example of a de novo motif insertion like we see in SARS-CoV-2. Lab leak proponents, such as Jeffrey Sachs, not only propose that the FCS was inserted (which is not even accurate, the staggered nucleotide insertions create PRRA, which forms a minimal FCS motif RRAR only with an already existing R in the viral backbone), but that an extra amino acid up front, a helix-breaking proline was thrown into the mix for no good reason. Beyond odd.
Anybody who knows anything about genetic engineering will admit that the FCS we observe in SARS-CoV-2 was always unlikely to be rationally designed. Because in addition to its odd, crippled cleavage motif and a superfluous proline amino acid that risks the integrity of the whole protein, the insert itself has a very high GC content that genetic engineers try to avoid. Even more tellingly, all the nucleotides are inserted out-of-frame, oddly breaking the insertion in staggered nucleotide triplet ends. No engineer would do that either; there is simply no reason other than to confuse oneself for fun. However, natural mechanisms that can create sequence insertions, such as viral recombination, are completely reading-frame agnostic. With recombination, the odds are two out of three that new inserts will be out-of-frame, too. To me, it all made the notion of a rational designer rather far-fetched. Yet merely pointing that out in my first blog article about the topic in 2021 got me into this hot mess, with lab leak proponents feverishly arguing that we cannot presume to know what Chinese (or American) engineers would do just because textbook genetic engineering and some experts would do things differently. We can never know their motives, they assert; therefore, we can never exclude that the FCS was artificially introduced. Apparently, experimental precedent, basic biology, experience, and parsimony did not count to judge the odds.
Luckily, we do not need to dive into the supposedly twisted minds and motivations of mysterious Chinese or American engineers, because nature is more creative and intricate than our naive human suppositions.
Unequivocal evidence that no human could have designed the FCS emerged from subtle mechanistic studies published in 2022. Researchers from the University of Texas in Galveston had investigated a short amino acid motif called QTQTN right upstream of the FCS motif. It's next-door neighbor, so to speak. QTQTN had garnered curiosity because, when SARS-CoV-2 was cultured in human cell lines, this motif would get lost over time, just as the FCS motif does. This was considered odd.
The two seemed to be a package deal: keep both or lose both. When researchers deleted just the QTQTN motif but kept the FCS intact, they found that the FCS could not be cleaved very well anymore. They discovered that the QTQTN forms a little loop on the side of the spike protein, positioning the FCS motif in just the right spot so that proteases like furin can cut it efficiently. Imagine holding it out like a ribbon so that the molecular scissors can quickly fly by and cut it. As a result, the deletion of the QTQTN loop made the FCS less accessible, reducing the proteolytic processing of the spike protein and, with it, the ability of virions to enter new cells. It was a symbiotic interaction, one genetic element empowering the other.
Further investigation revealed that there is a third component to that particular package deal: a glycosylation site (sugar shield) that would be assembled on the threonine amino acid (the second T in QTQTN). When the researchers in Galveston kept both the QTQTN loop and the FCS intact but removed the ability for the sugar shield to form on it, the spike protein would get pre-cleaved by furin but somehow had trouble interacting with the TMPRSS2 protease on the next target cells, reducing viral infectivity as well. The sugar shield served as a homing device for emerging pre-cleaved virions to find TMPRSS2. It became clear that SARS-CoV-2 needs the full molecular trifecta of glycosylation, loop extension, and cleavage motif to make the FCS efficiency hack work. This finding, although hidden deeply in the nitty-gritty of molecular virology, is absolutely remarkable and devastating to any engineering hypothesis.
Co-dependency and synergy between proximal genetic elements are hallmarks of evolutionary selection. These types of complex interactions are virtually impossible to design even with perfect knowledge of all structural components, which nobody anywhere in the world had in 2019. Also, consider that the QTQTN motif is an uncommon but natural element that is occasionally found in SARS-CoV-2-related bat viruses. So how does a symbiotic interaction—a package deal—come together between a supposedly “artificial” FCS and its natural neighbor? In 2019, there was simply no conceivable way that engineers could have ever anticipated, designed, or stumbled upon the observed, but hitherto unknown, mechanistic synergy in any laboratory setup. For me, this mechanistic discovery was the final nail in the coffin of the “engineered” FCS speculation.
Other differences were also found. While the FCS was one critical element for respiratory transmission that made headlines for its uniqueness, it was only one among many unique genetic elements in an intricate virus full of surprises. Since this discovery, researchers have found other such unique respiratory adaptations, including two single amino acid mutations called A372T and N519H. None of the related bat sarbecoviruses seem to have these, but SARS-CoV-2 maintains them strictly. Why?
We talked about three spike proteins forming a trillium flower; what we have not yet explained is that there is an open and closed state of that flower. When closed, the viral particle is better protected against the host’s immune system but cannot bind to the ACE2 receptor as well. To bind better, it has to adapt to an open conformation. One way to picture this is that the trimeric spike proteins oscillate between open and closed conformations, with mutations defining how much time the trillium flower spends in either. When a virus jumps from its natural host to a new, "naive" species without existing immunity, mutations favoring the open state are often selected. This is what we observe in SARS-CoV-2.
In short, researchers found that, for example, the A372T adaptation (a mutation of a single amino acid, unique to SARS-CoV-2 among the sarbeco bat viruses) destroyed yet another glycosylation site in the receptor-binding domain. This site is basically an anchor for a sugar shield that covers parts of the spike. The loss of this highly conserved (in bat viruses) glycosylation site, in turn, prompted a conformational change in the larger trimeric spike protein assembly, causing the RBD to turn outwards into a more respiratory-infectious “open” position. Great for facilitating ACE2 binding on respiratory cells. In lab experiments, reversing this single mutation decreased viral titers (a mark for infectivity) by fiftyfold. Similarly, another recently discovered respiratory adaptation called N519H is a flexibility mutation, lowering the energy barrier to overcome the conformational change between “open” and “closed” conformation. The closed form in bats protects against the host immune system, but makes binding to ACE2 more difficult compared to the open form, which reduces infection efficiency. Once SARS-CoV-2 jumps from the bat into an immune-naïve host animal, mutations that flexibly favor open confirmation are advantageous for transmission.
None of these 3D conformation-altering genetic kinks could have been anticipated by any engineers, nor could any of these host-context-dependent impacts have been reasonably stumbled upon in any laboratory setup. Apparent “uniqueness” based on limited comparisons to known bat viruses is not a good argument for a genetic element to be “artificially created.” It is a non sequitur. Nature is far more creative and capable of developing these genetic innovations because it runs trillions of such exploratory experiments each day, and when the circumstances are right, some viruses will hit the jackpot. For viruses capable of causing a pandemic in humans, the presence of such unusual genetic elements is fully expected, and it is not suspicious if related bat viruses do not contain them while inside their bat hosts.
Most lab leak proponents are perfectly capable of following the above reasoning based on the emerging evidence. Unfortunately, none seemed willing or able to let the myth of the engineered FCS go.
Since then, conspiracy theorists have proposed that Shi Zhengli must have had an unknown but perfect progenitor to SARS-CoV-2 in her lab. One that not only already included the receptor binding domain found in bats in Laos in 2021 but also had all these other synergistic genetic elements, conformation changes, stabilizing mutations, etc., required for respiratory spread already in place. Basically a perfectly capable human respiratory pathogen where Chinese engineers only needed to splice in an unusual and crippled furin cleavage motif that would ordinarily not even work, for some reason introduce it out-of-frame (something no engineer would do), add an extra proline spacer (that would usually break the whole protein structure) for fun (or to extend the loop length they did not know about), all so they could stumble upon a hitherto unknown mechanism by luck that made the FCS trifecta package deal all work together magically. Oh, and of course, Zhengli has kept all of that amazing molecular virology work worthy of a Nobel prize, as well as the hundreds of intermediate steps in between, secret for years before the pandemic started for no particular reason.
Stuart Neil and his two coauthors, Peter Jacobs, a climate scientist and communicator familiar with the dynamics and arguments of science deniers, and Stephan Lewandowsky, a professor of cognitive psychology focusing on conspiracy theories, would write in a prominent article for Scientific American:
In normal scientific inquiry, as evidence emerges, the remaining space for plausible hypotheses narrows. Not so with conspiracy theories and pseudoscience. One of their hallmarks is that they are self-sealing: as more evidence against the conspiracy emerges, adherents keep the theory alive by dismissing contrary evidence as further proof of the conspiracy, creating an ever more elaborate and complicated theory.
As Stuart Neil would say:
This is always my dividing line… Scientists go with data. And these guys never go with the data if the data is not going in their direction. If it doesn’t keep their little pet theory on the table. Then clearly, there is fraud going on. Let’s get their emails. Let’s cherry-pick. Let’s blah blah.
From his voice and expression, I could tell he was exhausted and haunted by the last few years.
Some details may differ, but Stuart’s trajectory has played out with dozens, if not hundreds, of outspoken scientists during the pandemic. It doesn’t matter if they engaged in discussions about pandemic origins, masks, vaccines, or supposed alternative miracle cures like Ivermectin. Their clinging to evidence made them targets of mobs that believing in evidence-free narratives.
In times of crisis and trauma, when myth, manipulation, and motivated reasoning are all too appealing, our ability to have informed discussions is diminished. Yet, following the facts and building a basis for shared reality, even with people we disagree with, seems like something worth preserving. For all our sake.
Somehow, we seem to have lost that ability forever.
§
2022 and 2023 marked two important years in origin research but also brought up the ugly side of science denial and anti-science activism. The lab leak myth had become too powerful to ever let go. Having lost just about every battle on the evidence front and in the scientific literature, lab leak proponents were in dire need of new narratives to justify their beliefs and identity. Yet, instead of trying to integrate new knowledge into their shifting speculations, many switched tactics entirely. They now fully immersed themselves in discrediting science and scientists while gaming the referees of public opinion: journalists, influencers, and politicians. Focused on creating a plethora of new counter-narratives against a consolidating scientific consensus, many seasoned opportunists and clout chasers saw an opening to get their name on the map.
As the dominos start to fall, there is a very real chance that the people we thought were heroes were actually villains. There is also a real chance that the researchers with the most-cited papers are guilty of the worst research conduct of the century [...] they pushed an overconfident narrative to the mass media, lying to the entire world and the scientific community to cover for Fauci.
The computational biologist Dr. Washburne would post this on Twitter on September 6, 2022, in what I would personally describe as his usual delusions-of-grandeur fashion. I had been paying attention after his profile began to rise.
When I first looked into his Twitter history, there was none discernible before 2022, which seemed odd to me for somebody with a sizable following. Despite that, the former postdoc had cultivated an influential network on Twitter. Somehow, he seemed to be connected with many of the contrarian academics, such as Dr. Robert Malone, Dr. Martin Kulldorff and Dr. Jay Bhattacharya, who had seen their number of followers explode during the pandemic. One for pushing “spike toxicity” anti-vaccine propaganda, the other two for advocating for the mass infection of citizens to supposedly get quickly to “herd immunity”, known under the pompous name the “Great Barrington Declaration”. Two dominant narratives on social media, unsupported by scientific evidence, that caused a lot of harm and confusion to public health.
Dr. Washburne, in contrast, seemed more like a data guy, posting models of epidemic case developments, preprints, and similar information that was somewhat typical for scientific conversations on Twitter. During exchanges, he presented himself as a polite scientist, a “deaf guy who is great at listening.” He constantly posted inspirational quotes about science, portraying himself as somebody who enjoys learning from others and respecting their perspective.
I appreciate folks’ diverse views and respect those who disagree with me. In my life, it’s the folk I disagreed with - but gave benefit of doubt & listened to - who’ve taught me the most. Happy Saturday, and I love you all!
That charitable tone was not extended towards the zoonati, though.
Looking into his Twitter history, it appears that on September 5th 2022, Dr. Washburne came across an idea that had been cooked up in the conspiratorial fever swamps of DRASTIC and related amateur sleuths that he could turn into something much bigger. For over a year, amateurs had discussed how the FCS could have been inserted or even how the whole genome of SARS-CoV-2 could possibly have been assembled by using various combinations of restriction sides. This “discussion” had been inspired by the work of Dr. Ralph Baric, who had developed a protocol to synthesize a known bat coronavirus genome directly from a sequence. Unlike Shi Zhengli’s group, Ralph Baric’s team never isolated and cultured a live coronavirus from bat droppings, so his team needed a synthetic approach to construct a viral genome. The amateur sleuths believed that SARS-CoV-2 was created the same way. Because CoV genomes are large, such a synthesis approach required first creating a handful of smaller pieces that needed to be stitched together, sometimes with the help of restriction enzymes (molecular scissors) and ligases (glues). By the time Dr. Washburne came across the idea, its most fervent proponents, Dr. VanDongen, an associate professor in pharmacology, and Dr. Bruttel, a German cancer researcher at an obstetric institute, already had a rather elaborate speculation in place. Dr. Washburne likely saw an opportunity to bring them to the next level. In my opinion, rather than being scientifically astute, he was an expert in upselling—taking a piece of content, polishing it, and repackaging it to make it look very attractive. The trio would soon set events in motion that would cascade to the top of the attention economy.
“At the time, I had no idea who any of these people were.” Dr. Alex Crits-Christoph told me. He was another computational biologist with a background in microbial genomics and would emerge as their sharpest online critic. Debunkers of scientific misinformation often come from the same field or community as the falsehood spreaders because they are usually best suited to follow the particular vocabulary and methodology used to mislead bystanders.
As a computational microbiologist and sequencing guy, Alex had been around the origin discussions online from near the beginning, starting with the flawed “HIV inserts” preprint that went viral in late January 2020. “That was just a classic bioinformatics error,” he explained. “I wrote a comment on bioRxiv,” the preprint server where it was uploaded. He then commented on Twitter, explaining both the scientific methodology and why some results just did not withstand scientific scrutiny, even superficially. He had his debunking cut out for him. “Then there was this RaTG13 sample… Somebody said it was fake,” he explained, not remembering the details but already involving some known lab leak proponents. “Well, that is interesting,” he thought. “It is actually really hard to fake sequencing data correctly. There are tools to simulate it, but it never really looks like the real thing,” he elaborated. So, he had a look at the raw sequencing data and debunked the assertion. Whatever you might think of RaTG13 and Shi Zhengli, this was a real sample and not fabricated. “They were not happy with my responses,” he remembered.
A major flaw with conspiracy theorists is that they never discard debunked ideas, constantly rewarming them if they fit the larger emotional story. Of course, Scarlett, the supposed bioweapon whistleblower, would later again claim that RaTG13 a fake because it suited her agenda. The next thing Alex recalled was a conspiracy theory that Jesse Bloom and Alina Chan promoted, related to the hype surrounding the Mojiang mine story and supposedly secret viruses named RaTG15 collected there, which they thought could have been SARS-CoV-2. A family tree including RaTG15 had been shown in a talk Zhengli gave, so of course, both Jesse and Alina, as well as most of DRASTIC, thought Zhengli was hiding pertinent sequences, with the latter strongly implicating there was evidence of a cover-up.
Alex challenged them: “But there are totally normal reasons to share incomplete sequences in talks.” He tried to explain to them that their comments were irresponsible because “based on the tree/alignment, these sequences aren’t recent ancestors” of SARS-CoV-2 and “don’t indicate lab origins.” Unlike with journalists, it was harder for the motivated contrarians to bullshit somebody like Alex who understood the flimsy arguments they were making.
“And of course, later they [Zhengli] published RaTG15 work… obviously it was not SARS-CoV-2… and everybody just forgets,” Alex described the dynamic. No accountability for the contrarian falsehood spreaders. No revisiting of previous beliefs. Alex was mostly ignored. But that particular debunking and Alex’s consistent pushback had gotten him a bit more on the radar with a different set of scientists. The serious ones. Kristian Andersen once told me, “Alex has these amazing sleuthing skills.” He explained how they started working together sometime down the road, inviting him as an outsider to work on a critical review of the evidence for the origins of SARS-CoV-2 alongside almost two dozen other scientists. “It is interesting, and it is important as well; that’s how I kinda stuck with it,” Alex explained, describing his continued engagement in correcting the many falsehoods the various factions of lab leak believers would bring forth.
This brings us back to Dr. Washburne and his collaborators.
On October 20th, 2022, Dr. Washburne announced that he had uploaded a new preprint with some rather dramatic conclusions:
We examined whether SARS-CoV-2 was synthesized in a lab.
We studied a common method for synthesizing CoVs in the lab.
This method was thought to not leave a fingerprint.
We found the fingerprint.
That fingerprint is in the SARS-CoV-2 genome.
It was a clear, simple, repetitive message with eye-catching formatting. The accompanying material appeared substantiated enough for non-experts, with math and complicated-looking figures that tried their hardest to give the impression that thorough science had gone into it. Maybe it was the fact that some influential names seemed to immediately jump on the preprint, such as the contrarian virologist Dr. Francois Balloux, who endorsed it right out the gate as “an important piece of work” that “looks both solid methodologically and conceptually,” and, of course, Dr. Feigl Ding, who was never shy about amplifying preprints with sensationalist implications and seemed to have learned nothing about the “HIV insert” story. Then there was a set of COVID-19 contrarians and anti-vaccine influencers that also seemed to jump on the preprint right off the bat. Dr. Washburne’s posts and the Bruttel et. al. preprint went viral on Twitter quickly.
“There are many kinds of ‘wrong’ in science, but this preprint is False,” Dr. Alex Crits-Christoph wrote at the time. Alex had seen the same half-baked theories about synthesis floating around for over a year, and he had explained to these same authors that these ideas are entirely contradicted by the fact that coronaviruses undergo recombination, and we had already found informative relatives in the wild.
Yet the Bruttel et al. preprint, as it became known, was quite something else. Polished by Dr. Washburne with shiny statistical figures, it was flashy. It looked really good. It was also scientifically unsound on multiple levels. To many experts, including me, it appeared as if the authors had deliberately constructed it to reach a predetermined conclusion. It started by cherry-picking two restriction enzymes that produced a somewhat evenly spaced pattern among many possible options. Then they applied what many experts and I would consider multiple inappropriate and misleading statistical tests to claim that this pattern is so unusual that only a human engineer could have come up with it. They also deliberately excluded a bat virus named RpYN06, one which Alice Hughes had collected in Yunnan. That “oversight”, if it was one, however would prove that the loss of a particular restriction site happened through recombination and not through genetic engineering. Alex Crits-Cristoph had even told the authors about RpYN06 and how its existence contradicted their assertions months before they posted their preprint, which seemingly purposefully excluded the sample. So it would be hard to come to a different conclusion other than that the authors must have known this genome existed yet chose to ignore it because it contradicted their conclusions. In science, such exclusion acts could be seen not only as highly misleading, but might even constitute deliberate scientific fraud.
Confronted with RpYN06 later, Dr. Bruttel falsely asserted that the virus was fake, that is why they excluded it. Supposedly all part of the cover-up. This behavior was not too surprising to me, given that Dr. Bruttel seems to have rarely encountered a viral genome that he did not believe to be engineered or faked in his mind. In my opinion, he seems to have fallen deep down the rabbit hole some time back. Over the years, he has for example falsely asserted that HIV, Ebola, and the Omicron variant were all man-made viruses, too. Not a trustworthy conduct for a scientist. Add to that cherry-picking, inappropriate methods and statistics, and the exclusion of contradictory evidence in the preprint, and what Dr. Bruttel, Dr. Vandongen and Dr. Washburne put to paper was not science but something else. In essence, it was exactly the same type of motivated reasoning covered in scientific language that intelligent design proponents use to discard the theory of evolution.
Despite the – what I would personally characterize as - “statistical paint job” the preprint received from Dr. Washburne’s efforts to in my opinion obfuscate its pseudoscientific reality, many virologists quickly dismissed the preprint as the nonsense they saw it to be. Kristian Andersen would tweet:
The study is a clear example of motivated reasoning with a heavy dose of technobabble to make it sound legitimate - but it’s nothing more than poppycock dressed up as science. In plain language - this is uninformed nonsense, and it’s simply not worth engaging with this bullshit.
Unfortunately, for the average bystander, the preprint and fanfare around it proved more persuasive than the harsh dismissal from experts. They bought into it in droves, to the surprise of scientists. “That one is always a head-scratcher,” Alex Crits-Christoph rubbed his chin, thinking about the preprint’s impact. “Anybody who is qualified to comment on it thinks it is a total waste of time, and so won't comment on it. They think it’s a total joke.”
“Unfortunately, while that is true, it leaves open a space,” Alex rightly recognized. In the current information environment, if scientists do not immediately refute such bullshit with hard data, which often takes an order of magnitude more work and effort than producing it (a phenomenon known as Brandolini’s law), the bullshit can develop a life of its own. “It also then creates this weird dynamic where journalists think nobody is pushing back on that… especially ones that are just on Twitter, which is a terrible place to get information from,” Alex observed. “And this is a dynamic that happens a lot.”
When Natasha Loder, a science reporter for The Economist and Kelsey Piper, a writer for Vox, saw the waves of engagement and bickering on Twitter, they likely believed there was a genuine controversy at hand around the preprint. On top of that, both seemingly got annoyed by Kristian Andersen’s harsh tone and dismissal. “I don't think this kind of snideness and contempt really serves public communications,” Kelsey Piper tweeted at him. “Ditto,” Natasha replied right under. They decided to take matters into their own hands. After all, if there’s so much smoke around the preprint, there must be some fire.
Natasha was first out with her story. “Controversial new research suggests SARS-CoV-2 bears signs of genetic engineering,” the google headline read, much to the horror of scientists who had seen the pathetic work floated around on Twitter in various forms before. How could a serious news outlet take it up with such a declarative headline? The world reacted, thinking that the renowned British news outlet had gotten a scoop of momentous proportions. Natasha, who I consider a proper journalist, later explained the headline came out unintentionally wrong over google, and had it changed very quickly. Nevertheless, the preprint (that never made it through peer review even years later, unsurprisingly) and its coverage was hailed by many lab leak believers as all but finally proving that SARS-CoV-2 was created in a lab. But why was this pseudoscientific preprint so successful in the first place?
“It really surprised anybody how much that junk blew up.” Alex rolled his eyes. I believe this might have had something to do with the well-connected Dr. Washburne, who was not an entirely unknown entity on Twitter. I later learned he had inhabited a different controversy bubble before he became engaged in the origin topic. Best I could find, he was part of a network aggregating around a shady think tank called the Brownstone Institute, funded by libertarian billionaire and reportedly pro-child labor advocate Jeffrey Tucker. It was created to oppose various measures against COVID-19, including lockdown measures, masking and vaccine mandates, as well as financing academic contrarians that were involved in the “Great Barrington Declaration,” a public health proposal advocating for lifting all restrictions and letting the virus rip unmitigated through society. Dr. Washburne is listed as author writing for the Brownstone Institute. He had also been sitting on panels with prominent anti-vaxxers and activists affiliated with the think tank, producing what I personally would describe as akin to “merchant of doubt” fan fiction, mixed with ideological talking points that aligned with the GDB authors.
“He was in the middle of a Twitter network, where people would see his stuff,” Alex noticed as well. It seems to me personally that Dr. Washburne was somewhat influential with a certain anti-pandemic measures crowd because he appeared to be this polite, data driven guy that somehow always found a scientific-sounding argument to justify their beliefs. Looking deeper into his “scientific” work, I found that Dr. Washburne had created, as best I can tell, misleading statistics to justify the policy decisions of right-wing politicians in Florida and claiming that the pandemic had reached “natural endpoints” due to herd immunity in the summer of 2021. A bogus claim scientifically, yet possibly contributing to the overall popular libertarian narrative that the danger of the virus was starkly overblown. A fatal misconception for too many Americans in red states, who also simultaneous were manipulated by too many anti-vaccine activists into not taking vaccines, ultimately leading to a large disparity in deaths between Republicans and Democrats. In my opinion, between his scientifically unsound publications, Dr. Washburne was mostly occupied with shaping the public perception of his persona. It appears that he had deleted his Twitter account and posting history multiple times, removing old tweets that aged poorly, partisan predictions, junk science, and outbursts that exposed his persona, seemingly re-emerging with a new angle when he returned. “He kinda gamified the system a bit by posting all types of feel-good stuff that people then followed him” Alex gave his opinion. To me, it appears most followers were not realizing there might be more than meets the eye. With his engagement in COVID-19 origin speculations, the contrarian finally seemed to have found his niche.
“It really caught people off guard and managed to break through in this really unfortunate way,” Alex recalled. While the “synthetic origin” preprint and The Economist news story about it were gaining steam, debunkers like Alex reacted. Within less than 24 hours, he put together a list of arguments and refutations, including code where he reproduced some of the alleged “statistical patterns” and showed how they are non-discriminative nonsense. Multiple others joined in as well. Even virologists at the University of Würzburg, Dr. Bruttel’s home institute, felt compelled to tear down the viral nonsense in a press release that must have stung Dr. Brutal and the institution with embarrassment. Yet the falsehood was out making its rounds.
Much later, even Dr. Ralph Baric, the world-leading expert in using reverse genetic systems to build coronaviruses, the literal inspiration for the “synthetic origin” preprint that went viral, would describe Dr. Washburne and his coauthors’ preprint as “a pathetic piece of work” under oath. Why was he under oath? Because Dr. Baric was called in to give testimony at closed-door hearings by US lawmakers. That a US congressional committee had interrogated him about motivated junk science in the first place might, however, be the most damning tell on how our information ecosystem increasingly shapes reality perception. It certainly confirmed how viral nonsense cooked up by fringe figures can penetrate even the thickest walls of power.
The same cannot be said about the corrections and debunkings by dozens of scientists. Their arguments against the flawed preprint were mostly never heard by the average citizen or policymakers. No newspaper made a story about how The Economist got played by activists not shying away from upselling pseudoscientific nonsense. Instead, all the virologists and debunkers got was more harassment from crowds that saw their belief that the virus had been created validated.
As scientists, we participate in communication platforms that fail so utterly that millions have already seen this obviously flawed work and will never hear it’s wrong. More people have seen this straightforward lie and internalized it as true than maybe any paper you’ll write.
Alex wrote about the experience. A reality of our modern information ecosystems where engaging fictions outcompete elaborate facts.
There are, of course, many reasons why the “synthetic origin” story went viral; it wasn’t just the crafty acts of the media manipulators themselves. Luck played a role, as did timing. The “synthetic origin” preprint was a perfect fit for the mood of the moment. First, it was a narrative sequel to the big DEFUSE story as well as Jeffrey Sachs’s pointing the finger at Ralph Baric’s lab, which he had been promoting for weeks. Second, with the US midterm elections at stake, the right-wing myth-making machinery was ready to give this polarizing origin controversy another go. The politicians realized that they could present themselves as the party of accountability, ready to punish those scientists who demanded masks and lockdowns, who forced vaccines into freedom-loving Americans, and who had created the virus in the first place. The much-hated Dr. Fauci, who smirked and made their president look stupid, rapidly became the focal point of their ire in these overlapping conspiracy myths. All the politicians needed were crafty content creators like Dr. Washburne, who could provide some flimsy support for their vendetta.
“It is often the stupider and simpler stories that really go big,” Alex Crits-Christoph noted dryly. Politicians are masters at recognizing their potential. You don’t persuade the hectic online masses to engage by advancing smart technical arguments, no matter if they are about science or policy. These often make people feel inadequate when judging a topic, and they discourage them from participating.
What works for winning over online tribes are simple messages and slogans—an exciting new angle to a common story trope that makes the complicated seem intuitive. Something that channels our emotional engagement and justifies our biases, where we can signal our allegiance and put our own spin on it for endless repetition. Coincidentally or not, these are the ingredients that can dominate the news cycle, too.
Unfortunately, as the scientific evidence accumulated entirely on the side of a natural origin hypothesis, more and more such motivated fictions completely detached from ordinary scientific inquiry started to gain momentum.
Right after The Economist went viral with the restriction enzyme bullshit bingo coverage, Katherine Eban had another, in my opinion, trope-laden story to sell to the flagship outlet ProPublica. Based on faulty translations of innocuous Chinese memos, Katherine built up a Republican Senate staffer into a codebreaking genius who could read between the lines of supposedly secret CCP party-speak. Together, the duo supposedly unearthed (or, more accurately, just fabricated) that a “grave and complex” biosafety emergency at WIV had taken place.
If this sounds silly and improbable, that is because it was. Countless native Chinese speakers, including Jane Qiu, were outraged by the motivated translation errors and apparent maliciousness of Eban’s story. So were many non-Chinese experts—including translators, sinologists, and China-based foreign correspondents—some of whom are fierce critics of Chinese governments. Even worse, Eban’s fabrication was tied in and coordinated with a later Republican Senate report attacking China, which would also be pulse-giving for a similarly titled “A Complex and Grave Situation” origin report from Senator Marco Rubio’s office. In his political fanfiction, Rubio offered a “political chronology of the SARS-CoV-2 outbreak” that blamed the Chinese communist party for creating the virus, again to increase tensions. That supposedly independent journalism outlets like ProPublica had been setting the stage for such blatant partisan PR stunts was a bit too on the nose for many journalists.
They say journalism is what makes democracy work. We should all ask ourselves: Have we helped to make democracy work? Or have we helped perpetuate stereotypes and existing narratives, exacerbate mistrust and polarization, and make the world a more dangerous place?
Jane Qiu, in response to the ProPublica story.
Even worse for ProPublica, they had received a five million dollar “grant” from the cryptocurrency fraudster Sam Bankman-Fried in early 2022 to do dedicated biosecurity reporting. Sam Bankman-Fried, as a leading light of the “effective altruism” movement, has dispensed a lot of money to investigative news outlets to do this type of coverage. The effective altruism movement had also been a community that fully bought into the lab leak idea as an existential risk and donated hundreds of thousands of dollars to the anti-science attack group USRTK, according to their own statements. This of course invites scrutiny. Did any of that “existential risk” money play any role in ProPublica’s choice of coverage or focus? Did it commission or facilitate any aspect of Katherine Eban and Jeffrey Cao’s misleading article?
I don’t know, but the amount of content that was created on the supposed existential risk of virus hunting from various outlets led to public pushback on virus discovery work and grants, resulting ultimately in the cancellation of pandemic prevention efforts such as DEEP-VZN, a program that was previously highly rated before online commentators got wind of it. It certainly seemed from my vantage point that putting large sums of money behind skilled storytellers was well-invested if one wanted to shape public discourse on the supposed existential risk of lab leaks. But is all of this existential risk fearmongering about virology in the field or the lab based on solid factual ground?
Few had time to ponder. Yet another media frenzy was started by The Wall Street Journal end of February 2023, and it subsequently captured the news cycle for days. They reported that the US Department of Energy had internally revised their assessment of a lab leak to likely, albeit with low confidence. No evidence for that assessment was ever provided, and as best one can tell, it was political. CNN learned that the DoE did not, however, blame the WIV; rather, it blamed a lab from the Wuhan CDC that was closer to the Huanan market. What a sensation! Doesn’t that finally explain the Huanan market connection? Two days later, Fox News followed up with FBI director Christopher Wray, who told them that “The FBI has for quite some time now assessed that the origins of the pandemic are most likely a potential lab incident in Wuhan.” Again, no evidence was offered, but it did not matter. Either way, it cascaded through society, with journalists dutifully reporting about pseudo news events that had little basis in reality.
Some lab leak proponents were already holding victory speeches. On Twitter, Alina Chan claimed she felt “validated” with the intelligence community seemingly on her side now. And they must know, right? Well, a few months later, a dry report from the Office of the Director of National Intelligence (ODNI), which oversees 11 US intelligence agencies, clarified that the whole of the US intelligence community had no special insights into the origin of the pandemic. They had no evidence of Shi Zhengli ever possessing a SARS-CoV-2 progenitor nor any evidence of WIV workers getting sick with COVID-19. In fact, most agencies agreed that the virus was not engineered, with four agencies assessing that a zoonotic origin was most likely and others not commenting. The FBI and DoE had been more of an outlier in their assessment, which seemed mostly based on their interpretation of Chinese obfuscation rather than any pertinent knowledge or insight. So much ado about nothing.
The only thing to learn from all these episodes is that our ability to create content for a belief has been democratized to the point where anything, no matter how wrong, motivated, or flimsy, can reach the peak of the attention economy and shape the news cycle. Especially when there is some dubious money and political desire behind it. Instead of a battle of facts—or at least interpretations of facts—the origin controversy has become a battle of viral narratives, cascading from the bottom to the top of the attention economy.
This bothers me enormously. Just as with the intricate biology of the furin cleavage site, I believe our incomplete understanding of what makes things go viral is, unfortunately, a problem that haunts us also in the technological realm.
§
We previously discussed how the arrival of the “winner takes all” attention economy has inadvertently changed the way information is valued in society. Instead of using information to inform and educate ourselves, we increasingly treat it as a product—something we consume or share for popularity, persuasion, profit, or power, like the influencers we have come to admire. When information is abundant, we pick and choose what we want to hear, not necessarily what we ought to hear. We have strong preferences to seek and amplify information that either confirms our preconceived notions, provides us with opportunities for social networking, or engages us emotionally. The accuracy, factuality, context, or topical relevance of the information we share is a secondary issue at best. As a result, certain information products (i.e., a funny meme, a tweet we like, an outrageous video, or a sensationalist article) have an easier time spreading through society than their competition.
How fast and how far information spreads is called information velocity. We want some relevant information to have a high velocity; think about an immediate crisis like an earthquake or school shooter warnings, where quick engagement and spreading can save lives. However, information velocity is not related to the veracity of the content of information. Misleading, false, or harmful information about an issue can have a much higher velocity than competing good information on the same issue. A meme showing 2+2=5 might spread a thousand times faster than a meme showing 2+2=4 for a variety of reasons, including sarcasm, humor, and irony. Information velocity is also dependent on the environment. Something that might go viral on TikTok might not go viral on Facebook, given dramatically different audiences and formats.
Perhaps most importantly, information velocity is a property that can be optimized externally without changing the underlying content. Two identical articles can have very different velocities depending on their headline. Two identical videos can have very different velocities by just changing the thumbnail. Two identical tweets can have very different velocities depending on the time they are first posted. Two identical memes can have very different velocities depending on who has written them. Messengers also play a role. If Taylor Swift writes “haha” in a tweet, it will be shared by millions. For most of us, the same tweet will go nowhere.
Today, there are multiple industries trying to figure these dynamics out, most prominently marketing agencies, advertising companies, and political campaigns. They seek to constantly optimize their information products, or the ones of their clients, to have higher velocity on various online platforms. Search engine optimization, keyword rank monitoring, link building, and nurturing are all intended to game the algorithms that curate our feeds. However, the most impactful velocity strategies are the ones that hack our human psychology and get us to participate and share the content we are exposed to. There are more and less legitimate ways to do so, with a big gray area in between. For example, a pro-vaccination campaign might increase the information velocity of its content by making complex scientific information more accessible to laypeople through compelling visualization, simplification of messages, inclusive language, and community engagement. This can really help public health officials get their message about vaccines out. Most of the world is, however, neither that noble nor benign.
A historical gray area is when headline writers use sensationalism or shock to get more people to click and share their articles. “If it bleeds, it leads” is an old adage attributed to William Randolph Hearst about this very phenomenon. With the rise of the attention economy, some of us have gotten very good at figuring out what the algorithms and our fellow humans want to see and amplify. We are a story-telling species, so narrative structures, including heroes and villains, work fantastically to package information products into bespoke worldviews with a unique selling proposition.
No matter how flashy, sharing a shady preprint that supposedly shows how SARS-CoV-2 contains fingerprints of genetic engineering isn’t going to go viral without the proper story tropes, emotional context of a gain-of-function panic in place, and a network of amplifiers giving it the initial boost. I learned that Dr. Washburne had meticulously arranged for sympathetic influencers, such as Dr. Balloux, to have advanced knowledge of the preprint before he announced it and received his endorsement. He had also contacted multiple journalists, including Natasha Loder from the Economist, and provided them with and advanced copy and “expert commentary” from credentialed pro-lab leak academics, making it very easy for journalists to get supportive quotes for their coverage, if they choose to pick it up. All these efforts increased the velocity of the “synthetic origin” preprint, regardless of its pseudoscientific content, in the information sphere.
Most influencers, grifters, snake oil salesmen, activists, politicians, and other media manipulators are brilliant at increasing the velocity of information products that work in their favor. There are hundreds of tricks that help game recommendation algorithms. The most impactful velocity hacks, however, are usually those that target our human vulnerabilities. The most unethical velocity hackers are using moralizing language to steer outrage, capitalizing on trigger words, jumping on hashtags, clout-chasing celebrities, and piggybacking on the news or traumatic events. Some examples include posting partisan talking points right after a school shooting or fabricating violent images with deepfakes to blame immigrants after a horrendous terror attack. Nothing is too toxic, too sacred, or too soon for velocity hackers chasing the next big hit. The velocity of their information product defines who wins, not the quality, accuracy, or truth of the content itself. It is an all-versus-all fight. From self-made influencers to mainstream media, from activist groups to marketing companies, from political campaigns to foreign influence operations, hacking the velocity of their information products has become the key business model for information merchants of all kinds. It is their ticket to popularity, persuasion, profit, or power in the information age. Those who get all the ingredients to velocity right—through luck, timing, and ruthless calculation—will catapult their content into virality.
Unfortunately, velocity hacking is extremely harmful to society and the public good. Virality not only shapes public discourse and what information people get to see, it also has a pernicious side effect that few of us have yet had time to wrap our heads around: it rewires our social networks.
Nobody understood this better than my now embattled acquaintance Renée DiResta, a disinformation researcher from the Stanford Internet Observatory who has recently become a target of a smear campaign. We were both trying to make sense of the craft of an overlapping cast of velocity hackers that Renée would end up naming “invisible rulers” in her book. She started innocently enough, trying to understand how propaganda and rumors spread online and shape public opinion. It wasn’t long before she became the target of a propaganda campaign herself.
“One of the things that social media does is you feel like you can trust the things that you're seeing because they’re coming from your friends. Right?” Renée explained. When our network shares something, we are more likely to engage with it, comment on it, and amplify it further. We take their collective behavior as a cue and act in a predictable fashion. In a sense, whether we engage with a piece of content is not fully our own decision; it depends on how others in our environment act and what decisions they made before us. This phenomenon is not new to social media by any means but is well studied, for example, in behavioral economics, and known under the term “information cascade.” The Lehman Brothers Bank is a good example of an information cascade.
In 2008, the US bank Lehman Brothers announced a massive asset write-down as a result of sub-prime mortgages. For Lehman’s management team and the majority of institutional investors (who understand how to value a company), this wasn’t a major problem. ... However, individual investors began to sell shares of Lehman stock as they feared their equity was in danger. Taking this behavior as a clue, more and more investors, seeing the stock price falling as a result of other investors unloading shares, decide to sell shares themselves based not on intrinsic valuation but on the general panic of the market. This eventually led to a massive drop in Lehman Brothers’ stock price over a very short period of time, eventually forcing the operationally-sound company to the point of insolvency and bankruptcy. The end result was the largest bankruptcy in US history and the further descent of the US economy into recession.
From the Dutch Tulip mania to bank runs to sudden cryptocurrency implosions and other investment bubble bursts, information cascades are events when large amounts of individuals take the same decision in a sequential fashion based on the decisions of others before them rather than their own personal knowledge or assessment of the situation.
On social media, we constantly mimic these cascading behaviors by making the binary decision to buy or reject information products in the form of engagement and sharing. When an influencer engages with a piece of content, their followers see this and might be nudged to make the same decision to engage. If a piece of content has been visibly engaged with by multiple people in our social circle, our decision to engage with it becomes more likely. After all, we want to participate in the conversation with our peers and signal our allegiance. At some point, after seeing almost everybody engaging with a specific information product, our supposedly private choice to engage becomes a near certainty. Our social network compels us to engage, and our decision to do so further influences others after us to engage as well. By being part of a specific social group and engaging with a specific information product the same way, we have become part of an information cascade amplifying the information, a stepping stone on its way into virality.
However, information cascades on social media do more than make things go viral; they teach algorithms what content and associated social relationships influence our decisions to engage. And inversely, the algorithms learn which of our connections and friends are less relevant to create ever higher likelihoods of us participating in such social information cascades. This leads to a radical social sorting over time.
Network studies showed that participation in “Tweet cascades increase the similarity between connected users, as users lose ties to more dissimilar users and add new ties to similar users” and that “Twitter users who follow and share more polarized news coverage tend to lose social ties to users of the opposite ideology.” It’s a dynamic system that nudges us towards a social amplification network while trimming our social ties that might interfere with viral amplification; you know, those low-velocity buzzkill people who ask questions, fact-check information, and caution others not to spread false rumors. Debunkers like Dr. Alex Crits-Christoph and other scientists who care about getting it right.
Fact checks, diversity of opinion, nuanced discussion, and appeals to deliberation all create friction and slow down the amplification and spread of viral narratives. Platform and ranking algorithms have long figured out that the social engineering of our networks removes these frictions and compels us to become better hosts in viral amplification cascades. “Think about it as an ecology: The online ecosystem lends itself at a particular moment to a particular species, which thrives and in turn reshape their environment,” Renée explained.
She hopes that “understanding how recommender systems sort people into networks can help us be more cognizant of polarizing factions” that inevitably arise through this social engineering process that makes new information cascades ever more likely. The consequences of such information cascades can be no less dramatic on social media than in finance. Instead of a bank run leading to a historic bankruptcy, we get pile-ons, mobs, and witch hunts. We might also see, let’s say, a vibe shift in media coverage and elite belief that causes a geopolitical escalation with a foreign superpower.
According to Renée:
Vibe shifts happen when you have a lot of people who are seeing the same thing. The narrative becomes very, very common in the community, and it becomes like... the tipping point. You just have that moment where you see it all around you, and it's much more about the sentiment than the facts.
Talking to Renée, my mind rushed back to Nicholas Wade, whom we discussed in a previous chapter, putting the blame on his peers in the media. “Can you produce a vibe shift by targeting specific journalists and media tastemakers?” I proposed to her. “Yes, I think the answer is yes,” she jumped in halfway through my question. “You even see this in the Russia 2016 data.” Renée had studied Russian propaganda and election interference in the 2016 US election. “Who do they mention in their tweets? Right? They want to get retweeted by prominent people, and so they’re actively soliciting and replying to people who have the power to boost them.” Sometimes, foreign agents even directly pay influencers to boost certain narratives. They do this because they want to reach a critical mass, a tipping point in the discourse. Causing vibe shifts is their game; that’s how media manipulators try to shape social networks in their favor and exert influence that can become relevant on the geopolitical stage. Today, a sizable proportion of Americans buy into Kremlin propaganda that has been promoted by paid American influencers.
All the false news stories about a supposed lab leak that went viral over the years flashed past my inner eye. The motivated and needy conspiracy theorists are tagging scientists, politicians, and journalists, trying to get them to comment and engage. All the carefully crafted pieces by gifted storytellers, multiple genuine disinformation campaigns, and the many plots, ploys, and players who aimed to sway tabloids and news outlets, politicians, mainstream journalists, and influencers to participate in lab leak speculations. How they all found there was something to gain by participating and co-creating these viral narratives surrounding a supposed unnatural origin of the pandemic. In turn, the ensuing virality in our information ecosystem socially aligned them more and more with each other until they became an unstoppable force to be reckoned with scientifically, politically, and socially. “Each individual act of clicking or resharing may not feel impactful, but in the aggregate, those acts shape conversations, beliefs, [and] realities,” Renée summarized aptly.
It is a phenomenon bigger than ourselves and individual actors.
Virality is the driving force that shapes the topology of our social networks. It gradually defines our relationships, drowning out acquaintances, friends, or even family that might think a bit differently about some topic. In return, it rewards the most efficient velocity hackers and elevates their network to become our new guiding stars. This leads us further away from reality because, while fiction can be optimized for maximum velocity, facts usually cannot. The best velocity hackers inevitably veer into self-serving falsehoods to win the battle for our attention. Over time, an asymmetry is created where the most efficient information cascades form around the most viral falsehoods supplied by content from the most successful velocity hackers. This is the world we have created today.
Science cannot compete. By the time scientifically accurate information reaches the majority of us, we have already been sorted into polarized factions where we clash about identities and worldviews, often unable or unwilling to concede an inch even when the other side makes a good point. No wonder that Prof. Stuart Neil, “the zoonati they could talk to,” now laments how “gradually, that all got poisoned.”
Renée DiResta has observed that:
People initially come to participate in online crowds because of a mix of algorithmic nudging and personal interest. Being part of a political faction can be fulfilling—there is a cause and a mission. Fighting a common enemy creates camaraderie and a sense of belonging.
However, once we are part of a tribe, we tend to defend our co-created narratives with all we’ve got and subsequently get trapped in a vicious circle. “Participation in factions may lead to entrenchment, more extreme beliefs, or stronger and more belligerent partisanship.” According to Renée, platforms need to be held accountable because their “design decisions now play a huge role in determining whether groups online are going to behave like civil communities or mobs.”
In my opinion, the extreme polarization today is largely a result of “winner-take-all” engagement algorithms, mercenary velocity hackers, and viral information cascades that stoke our emotions and nudge us to participate in conflicting online tribes. Once our social networks have been restructured along the lowest common denominator lines, such as partisanship or conspiratorial worldview, building bridges and good-faith discussions between opposing camps becomes almost impossible. “You don’t consort with the other side while you’re at war! And in the gladiatorial arena of social media, there is always tension,” Renée observes.
All platforms and their ranking algorithms show users emotional and engaging content to keep them hooked, but there are differences in how they are designed to optimize user participation and how they restructure networks. Platforms like Twitter highlight extreme members and representatives of other factions to provoke outrage in users about what despicable opinions others hold. Facebook creates closely knit interest groups that serve as echo chambers, driving people deeper into a niche worldview with ever more radicalizing content and very little contradictory information penetrating that bubble. “At its worst, Twitter made mobs, and Facebook grew cults,” Renée would summarize succinctly. Either way, restructuring social networks is dangerous for society.
Neither mobs nor cults are known to be very amenable to scientific conversations or any type of rational discourse that is necessary to live together in a pluralistic democratic society. Yet because of their “asymmetry of passion” for a topic, these extremist crowds make for great activists and keyboard warriors, who are then favored by algorithmic amplification. They have a disproportionate impact on our public discourse; that is why velocity hacking influencers need to cater to them. These crowds are their ticket to hijack engagement algorithms that boost them to the top of the attention economy—the ultimate social velocity hack. Influencers, no matter if secular gurus, political pundits, heterodox podcasters, or snake oil salesmen, who have harnessed and directed the energy of these crowds have subsequently gathered the biggest platforms.
It does not surprise me that the heterodox- and secular guru-sphere are all connected to Joe Rogan, who constantly circulates its guests among the same group of in-network podcasters. The algorithms brought them together because they all knew how to virally elevate emotionally salient culture war issues created by motivated partisans and conspiratorial crowds. The amplification dynamic in turn shaped public discourse and their own audience and contributors in the process. Increasingly, this builds up a bespoke information reality around us that feels real to us. As Renée puts forward in WIRED magazine:
What I find most alarming is that people have the ability to just create reality by making something trend, to reinforce over and over and over again these conspiracy theories. You do have this increasingly divergent set of realities where there’s a deep conviction built up over many, many years of reinforcing the same tropes and stories. You can’t just correct that with a fact check.
The consequences for public health were dramatic. “The most prominent influencers in the conversation managed to frame every conceivable aspect of a global pandemic not as a fight of humanity against a viral invader but as culture war battles about identity and values,” Renée writes in her book. “And institutions, unfortunately, were ill-equipped to participate.”
During the pandemic, many scientists and institutions talked about the role of misinformation or disinformation in shaping public perception. The WHO spoke of an infodemic of false and misleading information. Everyone has tried to wrap their heads around why so many false rumors—about the pandemic, NPIs, vaccines, masks, and so on—have sabotaged public health responses in many societies worldwide. Public health communicators often sleepwalked into a fragmented minefield they did not know existed.
The reality is that the deck had been stacked against them from the start. For years, online culture warriors and tribal gladiators have carved up society into factions with the help of algorithms and asymmetrically engaged crowds. Engagement algorithms made sure to remove the friction for information flow within these socially restructured amplification networks while building barriers to them for outsiders. Scientifically accurate information rarely reaches people unfiltered by influencers anymore. The socially engineered crowds, in turn, looked for influencers and conduits who could channel their feelings into viral narratives while their activists and true believers were working to sway, subvert, or discredit the traditional gatekeepers of information flow, such as institutions or media. A remarkable synergy for virality ensued. The ultimate efficiency hack to spread harmful falsehoods: a participatory anti-science ecosystem.
Just like with the genetic trinity of loop extension, glycosylation, and pre-cleavage around the FCS efficiency hack we discussed earlier, the trinity of crowds, algorithms, and influencers achieved together what one part alone could not: creating cascading outbreaks, not of biological agents, but viral anti-science narratives that have developed their own lives and become self-sustaining in our information ecosystem. The bombardment of information cascades, in turn, altered our public discourse and society in ways we do not yet fully understand.
Since the beginning of the pandemic, I have witness how viral information cascade after viral information cascade about an “unnatural” pandemic has washed over society. How these narratives were rearranging the topology of our social networks, infecting new immune-naive citizens, and re-infecting old bespoke communities alike with ever-new versions of the lab leak origin myth. Just off my memory, the various tales went something like this over the years:
It was a pretense to stop pro-democracy movements in Hong Kong. It was released by the Chinese. Or the Americans. No, it was planned because a vaccine patent was about to expire. Alarm! The virus had HIV inserts. No, it was a bioweapon built around bat viruses ZC45 and ZXC21 collected by a military academy. Or maybe they isolated it from the Mojiang mine, where people died of a mysterious disease we believe was COVID? They must have collected the virus in secret and cultured it. They had poor biosafety. Three WIV workers got sick. Actually, the virus wasn’t collected but a chimera, constructed in a reckless gain-of-function experiment gone awry. Project DEFUSE was the blueprint. Maybe they tried to mimic EnAC’s furin cleavage site since the virus was likely constructed in Ralph Baric’s lab in North Carolina. Actually, a preprint just came out explaining how the whole virus was stitched together in a computer and synthetically assembled with restriction sites. So many possibilities! No matter what, these memos in secret Chinese party-speak tell us that there was a grave biosecurity situation at WIV. Did you know the Chinese military was deeply involved with Zhengli’s lab? Maybe they did bioweapon research after all. Breaking news! The Department of Energy now thinks a lab leak is most likely. But actually, they do not believe it was WIV, but rather a different Wuhan laboratory closer to the Huanan market. Oops. Seems like gain-of-function is off the table. But the Wuhan CDC had moved there in December 2019; maybe something went wrong during the move? Wait, the FBI director claims the agency is moderately confident in an engineered virus leak since 2020, so it’s back to the WIV! What about all those intelligence agencies who have classified information? Did DARPA not fund EcoHealth? When the US military is involved; it goes to the highest levels of the US government. Daszak is likely a CIA spy; that’s why they protect him. They funded EcoHealth to keep tabs on Chinese bioweapon development. But the ODNI report says they have no special knowledge; does that mean the intelligence community is covering up the truth as well?
Nobody bought into all the stories; many are contradictory and deeply steeped in conspiratorial ideation. We are not so irrational. But few of us were able to resist every single one of those compelling narratives. Especially when “the power to create pseudo-events has been democratized,” as Renée would put it. We would always find something in the news that was persuasive to us. The information sphere always delivers, and that includes our own desires. News cycles can be fabricated from thin air to the point where influencers can find apparent media amplification for any remote beliefs, no matter how flimsy, fraudulent, or far-fetched the evidence. Like an economics professor talking about “molecular mimicry,” shady contrarians supposedly discovering “synthetic fingerprints” in SARS-CoV-2’s genome that all the virologists did not see, or Republican staffers being supposedly able to decode secret Chinese party-speak that evaded over a billion native speakers. I guess virality online is no less dynamic and context-dependent than in biology.
As we have seen during the pandemic, those virus variants that spread faster than the competition are the ones that would sweep the world, overwhelm our innate and acquired defenses, and make people sick. Viral narratives are not that different; they just make people sick with false perceptions and beliefs. As a consequence, just like real viruses, viral narratives have the potential to harm our societies severely. Depending on what communities we were part of and what motivated us to participate, our beliefs were gradually shaped as the consensus of our social network was taken over by them.
We started to shift away from rooting our understanding of the world in scientific evidence toward what we wanted to believe. We took our cues from high-status individuals we trusted and believed, whether they were popular gurus, relatable influencers, aligned politicians, or smart-sounding journalists. In that environment, I am afraid, science and institutions can no longer even compete for our hearts and minds.
Without well-connected and influential spreaders who can amplify narratives and make them go viral, without activist crowds creating pseudo-events, and without social media algorithms constantly pushing new, unwitting citizens towards them, science seems at a loss in the attention economy. Even the most worthy and relevant scientific information will have a hard time winning against competition for our attention. Going forward, how would citizens ever learn about real scientific progress? How can they get the context of new discoveries being made? And even if scientists ever had a scientific breakthrough that went viral and reached the masses by lucky coincidence, who would be able to accept new evidence given the hardened and polarized fronts of our fragmented realities?
In March 2023, we were all about to find out.
Adapted from Lab Leak Fever: The COVID-19 Origin Theory that Sabotaged Science and Society by Philipp Markolin.
Copyright © 2025 by Philipp Markolin. All rights reserved.
Continue reading chapter 11 here.
Note: If you want to download, print-out, share or otherwise collect this chapter, run it through an LLM or just store for record keeping; here is a high-quality pdf version as well:
Find more background info, chapter footnotes and video interviews at www.lab-leak-fever.com. If you want a physical copy of the book, kindle ebook, or support my work, you can buy directly on Amazon. (for non-US readers, please check your regional Amazon such as amazon.br or amazon.in as ebook prices may differ dramatically)
How did you feel about this chapter? Please let me know any feedback or comment below. Make sure to share.
