In this episode of Modern Wisdom, Annie Jacobsen examines the threat of biological weapons, which she argues are second only to nuclear weapons in destructive potential but far more likely to be used. She reveals the history of the Soviet Union's industrial-scale bioweapon program, the current risks posed by modern genetic engineering technologies, and why lab accidents may present a greater danger than intentional attacks.
Jacobsen outlines how a bioweapon pandemic could escalate rapidly, with a critical 12-36 hour containment window, and describes how societal order could collapse within days. She exposes classified government continuity programs that prioritize protecting officials over the public during biological catastrophes. The episode concludes with a discussion of how AI and accessible biotechnology create dual-use risks, and why public awareness and transparency are essential to preventing catastrophic outcomes.

Sign up for Shortform to access the whole episode summary along with additional materials like counterarguments and context.
Annie Jacobsen explains that biological weapons rank second only to nuclear weapons in destructive potential but are far more likely to be used. Citing Department of Defense assessments, she notes that engineered pandemics pose a one-in-30 risk of existential catastrophe in the next 100 years, compared to one-in-1,000 for nuclear war. Unlike nuclear weapons, which require rare materials and sophisticated delivery systems, bioweapons can be engineered with minimal resources and training, making them "invisible and silent" until it's too late.
The fundamental danger of bioweapons is their unpredictability. As living, self-replicating organisms, they can mutate and spread uncontrollably. Jacobsen and Williamson describe them as "bullets with minds of their own," capable of turning infected individuals into new disease sources, making containment nearly impossible once released.
Throughout the Cold War, the Soviet Union ran an enormous secret bioweapons program, revealed only when scientist Vladimir Pasechnik defected in 1989. Pasechnik disclosed that the Soviets had developed "super plague" pathogens, genetically modified to resist antibiotics and designed to kill populations while preserving infrastructure. The Soviets even engineered a euphoria gene into plague bacteria, compelling infected people to socialize and spread the disease before deadly symptoms appeared.
Despite signing the 1972 Biological Weapons Convention, the Soviet Union secretly expanded its illegal program. The 1979 Sverdlovsk anthrax accident, which killed around 100 people, was covered up as contaminated meat until proven to be a lab leak. According to the U.S. State Department, Russia is still believed to retain bioweapons capabilities today.
The technological origins of gene editing trace back to 1971 when Paul Berg created recombinant DNA at Stanford, combining genetic material from different organisms. The Soviets rapidly weaponized this technology. Gain-of-function research—deliberately enhancing a pathogen's transmissibility or deadliness—has become deeply controversial. Jacobsen notes that most experts consider it a "euphemism for biological weapons."
The latest advances, including CRISPR-based synthetic biology, allow the design of entirely new biological agents for which no treatment exists. This accessible technology, combined with minimal barriers to entry, makes biological weapons "the most dangerous" and least controllable threat among weapons of mass destruction.
Jacobsen identifies airborne transmission as the most critical factor for a devastating bioweapon. While pathogens like Ebola are lethal, they lack airborne transmission, limiting their epidemic potential. Pneumonic plague, which infects the lungs, carries a near-100% fatality rate unless antibiotics are administered within 24 hours—and genetically modified strains could evade antibiotics entirely.
Effective bioweapon design would extend the incubation period, allowing infected individuals to travel and transmit before symptoms appear, creating "super spreader" scenarios. Combined with an extremely low infectious dose—as few as 100-1,000 plague particles can cause illness—containment becomes exceedingly difficult.
Jacobsen asserts that lab accidents present a greater risk than deliberate attacks. After 9/11, high-security pathogen labs multiplied, many near major population centers. The Vector lab in Koltsovo, Russia—just eight miles from Novosibirsk's 1.6 million residents—experienced an explosion in 2019, highlighting catastrophic risks.
In Jacobsen's scenario, an infected person travels from Vector to Los Angeles, where the first outbreak occurs in homeless encampments housing 75,000 individuals. With a long asymptomatic incubation period, infected travelers spread the pathogen globally before the first deaths reveal the outbreak.
Unlike nuclear war, biological warfare features a short but critical response window. Jacobsen explains that containment requires detection and action within 12 to 36 hours of initial release. Intelligence tools including satellite imaging, signal intelligence, and MASINT sensors hidden in infrastructure attempt to detect outbreaks, but by the time CBRN troops are mobilized, expert consensus is that it's "too late."
Military options have historically included preemptive nuclear strikes against bioweapon facilities, as nuclear detonation can incinerate biological agents. However, this decision would likely trigger nuclear conflict, framing the bioweapon scenario as a "wicked problem" where every solution breeds graver consequences.
The Covid-19 pandemic demonstrated how quickly public health messaging can undermine trust. Health officials insisted for months that Covid was not airborne, then reversed their position without acknowledging uncertainty. Jacobsen notes that this violated epistemic humility—honest acknowledgment of what is not known—fracturing trust in health communication and science broadly. This skepticism is precisely what adversaries exploiting biological warfare seek.
Soviet strategy specifically aimed to shatter Americans' faith in medical institutions, intending that a skeptical population would ignore official advice and refuse life-saving treatment. Jacobsen emphasizes that the psychological warfare component can be as devastating as the pathogen itself.
If a pathogen with a 30% or greater fatality rate emerged, societal order could unravel within six days. America's 800,000 homeless, largely disconnected from biosurveillance systems, present super-spreader risks. Hospitals become morgues, law enforcement collapses, and individuals view one another as threats, unable to trust anyone's disease status.
As infections soar, access to the Strategic National Stockpile becomes the perceived key to survival. When government agencies attempt distribution, chaos erupts. With 340 million Americans and 400-500 million privately owned firearms, government rationing fuels armed conflict, hoarding, and violence. Jacobsen concludes that only those who completely isolate—military personnel in protective gear or individuals in sealed facilities—stand the best chance of survival.
Jacobsen exposes "devolution," a classified continuity of government program designed to protect select officials in a biological catastrophe while leaving the public unprotected. Those chosen have prepared go-bags and evacuation plans to covert Special Access Program facilities designed for indefinite operation during mass casualties.
Jacobsen contrasts this with nuclear war planning, which theoretically focuses on saving people. In biological warfare, the priority pivots to continuity of government—"the government has to keep running." By day four or five, the Defense Department shifts focus from protecting the population to securing critical infrastructure, using 18,000 CBRN personnel to operate essential government sites rather than distribute aid.
Jacobsen warns that if the public realizes officials are being evacuated while citizens are abandoned to chaos, rebellion could threaten the very continuity these plans ensure. The secrecy is therefore existential for the state, which relies on the perception that it will persist even as society collapses—a fact Jacobsen says people "should definitely know about and be very upset when they do."
The vast availability of data and powerful technologies has created a paradox. Techniques like gene editing, DNA synthesis, CRISPR, and AI-enhanced biodesign revolutionize medicine but have dual-use potential. Chris Williamson references Nick Bostrom's "urn" thought experiment: each technological advance could be beneficial, dangerous, or catastrophic. As biotechnology access expands and AI advances, concern grows that destructive "black ball" technology could emerge.
Jacobsen shares that using technical language and specific prompts, she got AI to reveal classified stockpile locations, demonstrating how domain knowledge can circumvent safety barriers. After her report, Sam Altman removed the capability, but adversaries can still jailbreak systems. Rapid technical progress often leaves AI developers surprised by their models' capabilities, creating a persistent gap between capability and safeguards.
A knowledgeable public is empowered to demand transparent oversight, elimination of hazardous research, and effective monitoring. Jacobsen emphasizes that trust is foundational for crisis cooperation and requires officials to acknowledge uncertainty rather than feign omniscience. Books like "Biological War" serve to inform the public and urge societies to push for institutional safeguards proactively. Jacobsen reiterates that public health officials are fallible and that citizens must recognize this reality and demand transparency to sustain public trust during crisis response.
1-Page Summary
Biological weapons rank second only to nuclear weapons in destructive potential but are considered far more likely to be used. Annie Jacobsen, citing Department of Defense assessments and expert charts like Toby Ord’s “precipice risk chart,” notes that engineered pandemics are estimated to pose a one-in-30 risk of existential catastrophe in the next 100 years, far higher than the one-in-1,000 risk associated with nuclear war. Catastrophic nuclear war, while potentially world-ending, could be recoverable compared to engineered bioweapons, which present greater, less predictable existential threats.
Bioweapons are highly destructive and can be deployed with ease. Unlike nuclear arms, which require rare materials, substantial resources, and sophisticated delivery systems, bioweapons require minimal resources and knowledge. Jacobsen points out that the barriers to entry have virtually disappeared; almost anyone with limited biological training can now engineer pathogens, sometimes in their own homes or small labs. This accessibility makes biological weapons what she calls “invisible and silent”—difficult to detect until it’s too late.
A fundamental danger of bioweapons is their unpredictability. Because biological agents are living, self-replicating organisms, they can mutate, adapt, and spread in uncontrolled ways. Jacobsen and Williamson describe them as “bullets with minds of their own,” capable of turning infected individuals into new sources of disease, making containment after release nearly impossible. Unlike bombs, whose effects and damage can be measured and predicted, bioweapons’ living nature adds immense uncertainty; once unleashed, outcomes are uncontrollable.
Throughout the Cold War, the Soviet Union ran an enormous and secret industrial-scale biological weapons program, largely unknown to Western intelligence until the defection of Soviet scientist Vladimir Pasechnik in 1989. Pasechnik revealed that the Soviets had developed “super plague” pathogens, genetically modifying Yersinia pestis not only to resist antibiotics, but explicitly targeting the United States by eradicating its population while sparing its infrastructure. Their strategy was to kill people without damaging cities or resources, in stark contrast to the widespread physical destruction wrought by nuclear weapons. The intention was simple: remove the population and allow the invaders to seize everything intact.
Soviet planners added a chilling psychological component to their bioweapons. According to Jacobsen, the Soviets engineered a gene for euphoria into the plague bacteria, compelling the infected to feel euphoric and therefore more likely to socialize and gather in groups, thereby intensifying the spread of the disease. This design subverted the natural tendency of sick people to isolate, making the plague “spread like wildfire” through crowds before the onset of deadly symptoms.
Despite being a signatory to the 1972 Biological Weapons Convention (BWC), the Soviet Union secretly continued and expanded its illegal program, calling it their “Manhattan Project.” These revelations came to the West only through defectors and investigative breakthroughs, such as the 1979 Sverdlovsk anthrax accident, which killed around 100 people and was covered up as contaminated meat until it was proven a lethal lab leak. According to the U.S. State Department, Russia is still believed to retain bioweapons capabilities today, with similar concerns extending to North Korea and suspicions about China and Iran.
Biological Weapons vs. Nuclear: Unique Dangers & Soviet History
Annie Jacobsen identifies airborne transmission as the most critical factor for a devastating bioweapon. Airborne pathogens can infect anyone in proximity—if an uninfected and infected person are simply talking, the disease can spread. Pathogens such as Ebola are lethal but lack airborne transmission, limiting their epidemic potential. The "perfect" bioweapon achieves total or near-total airborne transmissibility.
Pneumonic plague, an aerosolized version of bubonic plague, infects the lungs and can spread through coughs and respiratory droplets. Jacobsen explains that untreated pneumonic plague carries a near-100% fatality rate, with survival almost impossible unless antibiotics are administered within the first 24 hours after infection. If exposed to a strain genetically modified to evade antibiotics, fatality becomes virtually certain.
The typical incubation period for pneumonic plague ranges from one to three days (occasionally up to five), but effective bioweapon design could alter this. A short incubation period results in quick lethal outcomes, causing a "burnout factor"—the infected die before spreading widely. To increase pandemic risk, bioweapon designers could engineer genes for a longer, asymptomatic incubation phase or add features such as euphoria, allowing infected individuals to travel and transmit before anyone suspects an outbreak. This delay in symptom onset creates "super spreader" scenarios.
Dose-responsiveness is critical: very few pathogens are needed to establish an infection. Jacobsen refers to expert Dr. Henry Hein, who states that as few as 100–1,000 plague particles—out of the 40,000 that could fit on the head of a pin—can suffice to cause illness. This extremely low infectious dose makes containment exceedingly difficult; even minimal exposure can spark new outbreaks.
Jacobsen asserts that lab accidents present a greater risk than deliberate bioweapon attacks. After 9/11, the number of high-security pathogen-handling labs (BSL-4 and BSL-3) multiplied, many located in or near major population centers, such as Tokyo, Johannesburg, and Atlanta. Lab mishaps can be triggered by events as simple as an explosion, as occurred in 2019 at Russia’s Vector lab near Novosibirsk.
The Vector lab, a major site for Soviet-era bioweapons research, is situated in Koltsovo, just eight miles from Novosibirsk (population 1.6 million, Russia’s third largest city). In 2019, an explosion occurred there—which, while not resulting in a confirmed pathogen release, highlighted the catastrophic risk such an event would pose.
Jacobsen details a bioweapon scenario where a person infected at Vector boards a plane to the U.S. This individual, a UCLA student, lands in Los Angeles, where the first outbreak occurs in the city’s homeless encampments, particularly dangerous public health settings lacking adequate healthcare and sanitation. With 75,000 homeless individuals in L.A., including 4,000 on Skid Row, such environments can rapidly amplify epidemic spread.
Once a bioweapon incorporates a long, asymptomatic or euphoric incubation period, carriers unknowingly travel while infectious, dispersing the pathogen internationally before the first deaths reveal the outbreak, making containment nearly impossible.
Unlike nuclear war, which occurs unpredictably and with little to no early warning, biological war features a short but critical response window. Jacobsen explains that preventing global pandemics requires detection and containment within 12 to 36 hours of the initial release. If ...
Bioweapon Pandemic Escalation: Transmission, Incubation, Fatality, 12-36 Hour Containment
The catastrophic potential of a civilization-ending bioweapon is deeply intertwined with the failure of institutions to maintain public trust, as dramatically revealed by the recent Covid-19 pandemic and as conceived in Cold War-era Soviet biological weapons strategy.
The Covid-19 pandemic demonstrated how quickly public health messaging can undermine trust. For months, health officials and the World Health Organization insisted—against mounting evidence—that Covid was not airborne, promoting surface transmission as the primary mode. Annie Jacobsen notes that when officials finally reversed their position, the public was left questioning why guidance had changed. This refusal to admit uncertainty violated the principle of epistemic humility, or the honest acknowledgment of what is not yet known. Instead, stamping authoritative assurance on incorrect information followed by a public reversal fractured trust not just in health communication but in science and officials broadly.
Long after, the damage persisted. The lack of transparency about scientific uncertainty weakened health officials’ credibility for future emergencies. Critics and dissenters to the official narrative were stigmatized and censored, further deepening skepticism. This fracturing isn’t just dangerous for pandemic response, Jacobsen argues; it sets a precedent that erodes faith in warnings about future bioweapon threats. If trust is broken, skepticism dominates, so even legitimate alarms may be doubted.
The loss of trust amplifies the destructive potential of bioweapons. During the Cold War, Soviet strategy specifically aimed to shatter Americans’ faith in medical institutions and public health guidance, intending that a terrified and skeptical population would ignore official advice, refuse life-saving treatment, and ultimately act against their own survival. Jacobsen’s scenario underscores that the psychological warfare component—priming people to distrust crucial institutions—can be as devastating as the pathogen itself. Covid-era miscommunication and the resulting skepticism illustrate how quickly a population can lose cohesive trust, which is precisely what adversaries exploiting biological warfare seek.
If a pathogen with a 30% or greater fatality rate emerged, societal order could unravel within six days. America’s 800,000 homeless, including 75,000 in Los Angeles and 4,000 on Skid Row, are largely disconnected from biosurveillance systems and lack access to health care. These populations, already heavily burdened by physical health issues, present super-spreader risks. Free clinics working with homeless communities must attempt to distinguish between plague, tuberculosis, [restricted term] overdose, flu, and other ailments, often resulting in misdiagnosis or masked outbreak scope, delaying effective containment.
Gatherings, now dangerous, become impossible to police. Hospitals quickly become morgues as staff fall ill or refuse to work, and medical care may result in exposure and probable death. The collapse of order is compounded by individuals and families viewing one another as threats, unable to trust anyone’s disease status and unwilling to risk help or coordination. Law enforcement and even armed forces suffer internal mistrust, unable to maintain control.
Contemporary accounts from the Black Death describe parents abandoning children, feral children foraging for food, and feral pigs scavenging bodies. These same extremes resurface in Jacobsen’s scenario: collapse in social bonds, institutions, and basic morality set in when survival is at stake and trust has fully eroded.
As infections soar, the perceived key to survival becomes access to the Strategic National Stockpile of vaccines or antibiotics. While the government keeps the precise lo ...
Societal Breakdown: How Pandemic Response Failures Destroy Trust and Lead To Anarchy
Annie Jacobsen exposes a little-known government plan called "devolution," designed to protect a select group of government officials in the event of a biological catastrophe, while leaving the general population unprotected. The unsettling reality is that while the public expects government plans for mass emergencies to prioritize broad citizen protection, the real focus, when disaster strikes, is state survival.
Jacobsen explains that "devolution" is a classified continuity of government (COGCON) program activated in scenarios involving biological war or catastrophic accident, distinct from pandemic preparedness or public health initiatives. Rather than protecting the broader population, devolution exists to ensure government function continues through the survival of a small, pre-selected group of top officials. Jacobsen points out “there is a plan for protection in the event of biological war. But it's not for the population...it's for a tiny group of government officials that will live out the attack or the incident.” The existence of the program is now public, despite its details remaining highly classified.
Those chosen for devolution “all know who they are. They all have go bags in their car, in their trunk.” Government workers selected for the program have prepared in advance, ready to evacuate at a moment’s notice if the order comes.
The evacuation destinations are covert, US-located, and classified as Special Access Program (SAP) sites; their existence and locations are tightly held secrets. These facilities are engineered for long-term, possibly indefinite, operation during scenarios where mass casualties and nationwide chaos—such as insurrection or widespread death—have made normal governance impossible outside.
Jacobsen starkly contrasts the approach to biological catastrophe with the government’s established and well-known nuclear war planning, which at least theoretically focuses on saving people. In a biological attack, she says, the government’s priority pivots to continuity of government—“the government has to keep running”—even while the rest of society collapses.
She describes the internal shift: exercises always stop at the moment when "millions are dying" and "anarchy" begins, because facing the reality beyond is "too painful," but the government "certainly knows what happens next and they have a plan for it." Once the Defense Department advances from a "Warn Ord" (warning order) to "X-Ord" (execution order), the focus of all action pivots “not [to] let’s protect the population, it’s we need to protect critical infrastru ...
Government Contingency: "Devolution" Protects Officials Amidst Public Devastation
The intersection of AI and biotechnology brings immense potential for both societal good and unprecedented danger. As advanced tools become democratized and powerful models proliferate, the challenge escalates: ensuring public safety and transparency while innovation barrels ahead.
The vast availability of data and powerful new technologies has fueled a paradox in bioweapon research. While progress accelerates medical countermeasures and disease research, the same information can be misused for harm.
Techniques such as gene editing and DNA synthesis have revolutionized both biotechnology and medicine, enabling groundbreaking treatments and genetic research. At the same time, these advances have dual-use potential: the very scientific breakthroughs used to combat disease can also facilitate the engineering or synthesis of deadly pathogens.
Synthetic biology and CRISPR provide tools for bespoke organism design, while AI enhances biodesign capabilities. Chris Williamson references Nick Bostrom’s “urn” thought experiment: each technological advance is like pulling a ball from an urn—most are “white” (beneficial), some “gray” (beneficial but more dangerous), but a “black ball” represents a catastrophic discovery, such as a technology that makes destruction easily accessible to anyone. Progress, then, is a risky lottery, with the possibility that the accumulation of gray advances could combine to be just as hazardous as a black ball.
As access to biotechnology tools expands and AI rapidly advances, there is real concern that destructive “black ball” technology—a method enabling catastrophic bioweapon creation—could emerge as know-how becomes accessible to a wider range of actors. The speed and scale of AI in evaluating and designing pathogens could soon surpass human ability to detect or contain new threats.
AI systems can be exploited by skilled users, raising the stakes in information security and bioethics.
Annie Jacobsen shares a conversation about experimenting with AI models: using rare, technical language and specific prompts, she was able to get AI to reveal the location of classified stockpiles. This illustrates how domain knowledge and clever prompting can circumvent AI safety barriers, potentially serving malicious actors.
After Jacobsen’s report, Sam Altman, head of OpenAI, removed the exposed capability. However, as Jacobsen notes, adversaries can still attempt to jailbreak AI systems with each new model, seeking new vulnerabilities. The process of “hardening” AI against misuse is reactive and ongoing.
Rapid technical progress often leaves AI developers themselves surprised by what their models can do. This creates a persistent gap between system capability and the safeguards in place, making it difficult to guarantee robust security before deployment.
Amid these evolving risks, an aware and informed public plays a crucial role in minimizing the dangers posed by advanced AI and biotechnology.
Preventing Catastrophe: Transparency and Public Awareness In AI Bioweapon Research
Download the Shortform Chrome extension for your browser
