Who is Kevin Esvelt and what is his expertise related to this topic?

Kevin Esvelt is a Harvard-trained biologist and MIT professor specializing in evolutionary engineering and synthetic ecosystems. He is credited with describing how CRISPR gene drives could be used to alter wild populations. His lab works with fast-evolving bacteriophages to create molecular tools and studies the safe engineering of wild organisms.

What are gene drives and are they safe?

Gene drives are naturally occurring phenomena where certain genes increase their inheritance rate, rapidly spreading traits through sexually reproducing populations. Esvelt developed CRISPR gene drives, which could be used for beneficial purposes like eradicating malaria. He emphasizes that gene drives favor 'defense' because they are slow, obvious, and easily countered with an 'immunizing reversal drive,' making them less suitable as weapons compared to fast, subtle, or unblockable threats.

What are information hazards and attention hazards, and how do they apply to bioweapons?

An information hazard occurs when knowledge itself (like blueprints) can be misused. An attention hazard is when the demonstration of a powerful capability (like a nuclear bomb on Hiroshima) incentivizes others to acquire that power, making the world less secure. The public release of dangerous virus genomes creates both an information hazard (blueprints) and an attention hazard (demonstrating their potential for devastation), potentially motivating malicious actors.

Why was the genome of the 1918 Spanish Flu posted online, and what are the risks and mitigating factors?

CDC researchers resurrected and sequenced the 1918 Spanish Flu genome to understand what made it so deadly and posted it online in 2005. The risk is providing blueprints for a deadly pathogen. However, mitigating factors include widespread immunity to its descendants, antiviral drugs, and antibiotics, which could significantly reduce the death toll today (estimated 200,000 to 10 million). The original decision-makers at CDC and NIH reportedly do not regret it, believing in the value of open science.

How many labs worldwide can currently synthesize a virus from a genetic sequence?

While only a few hundred people could assemble the 1918 flu virus in 2005, the cost of synthetic DNA has fallen dramatically. Today, it's estimated that between 20,000 and 50,000 people have the skills to synthesize a virus from a genomic blueprint, a number that is expected to continue growing.

What is 'virus hunting' and 'characterization' within the Deep VZN program, and what are their costs?

Virus hunting involves collecting samples from animals and environments to isolate new viruses. Characterization is the subsequent, more expensive process of running experiments to determine if an animal virus can bind to human cells, infect them, replicate, and transmit in animal models, to identify its pandemic potential. Deep VZN aims to find 10,000-12,000 new viruses over five years, significantly more than its predecessor, PREDICT.

What is the difference between natural spillover and deliberate release in terms of pandemic risk?

Natural spillovers and accidental releases typically originate from a single point, allowing some response time. However, a deliberate release could involve multiple pathogens launched simultaneously across numerous global travel hubs, drastically reducing response time and potentially increasing the death toll significantly due to the concurrent emergence of multiple novel diseases everywhere.

Why is it argued that Deep VZN could empower non-state actors and mentally ill individuals to pose a bioweapon threat?

If Deep VZN publicly blesses and publicizes genomes of pandemic-capable viruses, it lends credibility to threats from non-state actors or individuals. Unlike state actors who could only bluff about bioweapons, having publicly verified 'recipes' makes such threats credible. This shifts the dynamic, allowing more people than just the initial tens of thousands capable of synthesis to cause widespread disruption and fear, leading to 'socially contagious' misuse, similar to mass shootings following public examples.

Why are current vaccine development timelines (e.g., Moderna's COVID-19 vaccine) not sufficient for future deliberate pandemics?

While Moderna developed the COVID-19 vaccine formula in under 48 hours, regulatory approval and mass manufacturing took over 300 days. Even a 100-day 'stretch goal' for vaccine availability is too slow for a fast-spreading, deliberately released pandemic, especially if multiple viruses are released globally. Proposing pre-approved emergency protocols for rapid deployment of new mRNA vaccines is suggested as a necessary tool.

What are alternative strategies for pandemic preparedness that don't involve publicly posting dangerous virus genomes?

Instead of characterizing and publishing dangerous genomes, resources could be redirected to active monitoring of human-animal interfaces to identify outbreaks early, without revealing specific threat blueprints. Developing pan-familial vaccines (broad-spectrum protection against entire virus families) and pan-familial antivirals could also be informed by general virus sequencing (virus hunting) without requiring dangerous characterization experiments.

How can concerned citizens influence USAID's Deep VZN program?

Since Deep VZN has been approved but not fully launched, there's an opportunity to influence its direction. Citizens can contact USAID via their Twitter account (@usaid) or official contact form (usaid.gov/contact-us). Reaching out to elected representatives, particularly those on the Senate Foreign Relations Subcommittee on State Department and USAID Management, is also a highly effective mechanism to voice concerns about the program's security risks.

Key Moments

Special Episode: Recipes for Future Plagues

Sam Harris

Science & Technology4 min read118 min video

Feb 28, 2022|43,955 views|816|295

sam harris sam harris podcast waking up podcast waking up waking up with sam harris sam harris waking up waking up sam harris sam harris jordan peterson sam harris joe rogan author neuroscientist philosopher

Save to Pod

Key Moments

TL;DR

USAID's Deep VZN program could inadvertently publish dangerous virus genomes, posing a global security risk.

Key Insights

USAID's Deep VZN program aims to identify potentially pandemic-grade viruses but risks publishing their blueprints online.

The publication of virus genomes online, even with good intentions, creates 'weapons of mass destruction' accessible to many.

While gene drives are bi-directional and can be countered, weaponized viruses could be fast, subtle, and unblockable.

The scientific community's focus on open data sharing, while beneficial generally, poses significant information and attention hazards for dangerous pathogens.

The Deep VZN program's virus characterization experiments are argued to be the 'virological equivalent of nuclear testing' and should be restricted.

Alternative approaches like broader pathogen monitoring and advanced vaccine development capacity could mitigate risks without publishing dangerous blueprints.

THE DEEP VZN PROGRAM: A DOUBLE-EDGED SWORD

Ambassador Samantha Power's USAID is set to launch the Deep VZN program, a $125 million initiative focused on identifying potentially pandemic-causing viruses. While ostensibly aimed at preventing future pandemics by proactively understanding threats, the program's methodology involves sequencing and potentially publishing the genetic blueprints of these viruses. This open-data approach, standard in scientific research, raises profound security concerns, as it could inadvertently create accessible 'recipes' for dangerous pathogens, effectively arming individuals or groups with the means to engineer future pandemics.

INFORMATION HAZARDS AND THE WEAPONIZATION OF KNOWLEDGE

The core concern revolves around information hazards, particularly the 'attention hazard' and 'conceptual information hazard.' By widely disseminating the genetic sequences of dangerous viruses, the Deep VZN program risks alerting and incentivizing malicious actors. Unlike nuclear weapons, where only a few states possess them, the accessibility of these viral blueprints, coupled with advancements in DNA synthesis, could empower tens of thousands of individuals. This exponential increase in potential actors capable of causing mass casualties transforms the global security landscape, potentially enabling deliberate pandemics far surpassing natural ones in scope and impact.

THE SPANISH FLU BLUEPRINT AND GROWING ACCESSIBILITY

A stark example highlighted is the resurrection and online publication of the 1918 Spanish Flu virus genome by the CDC. Initially, the scientific community may have underestimated the ease of synthesizing viruses from sequences. However, rapid decreases in DNA synthesis costs mean that hundreds, if not thousands, of people globally now possess the capability to reconstruct such pathogens. This precedent illustrates how 'well-intentioned' data sharing regarding dangerous biological agents can lead to a proliferation of 'bioweapons' that are difficult to counter.

CRITICAL CHARACTERIZATION EXPERIMENTS AND THEIR RISKS

Deep VZN's methodology includes rigorous 'characterization' experiments to assess a virus's pandemic potential – its ability to bind to human cells, infect them, replicate, and transmit. These experiments are crucial for understanding threats but are framed by critics as the 'virological equivalent of nuclear testing.' The publication of findings from these tests, particularly identifying viruses that are both lethal and transmissible, directly creates the dangerous blueprints. The argument is made that this specific step should be forgone, focusing instead on broader virus discovery and monitoring without actively determining and disseminating precise pandemic risk profiles.

THE ARGUMENTS FOR DEEP VZN AND COUNTERPOINTS

Proponents argue that knowing potential threats is essential for preparedness, enabling faster vaccine and therapy development. However, critics counter that the time saved in vaccine development by having advanced knowledge is minimal compared to manufacturing and distribution timelines. Moreover, the risk of deliberately released, multi-pronged pandemics, potentially released simultaneously in multiple global hubs, could overwhelm even rapid response capabilities. Arguments for stockpiling vaccines are deemed less efficient and cost-effective than building rapid-response manufacturing capacity.

RETHINKING THE APPROACH: MONITORING OVER IDENTIFICATION

An alternative strategy suggested is to shift focus from identifying specific pandemic-capable viruses to broader monitoring of human-animal interfaces and animal populations. This would involve tracking the prevalence of viruses in at-risk animals and communities, thereby enabling earlier detection of potential spillover events. This approach supports the development of broad-spectrum vaccines and antivirals without creating public blueprints for dangerous pathogens. Such a strategy prioritizes containment and early warning systems over the proactive dissemination of high-risk viral information.

THE CALL TO ACTION: INFLUENCING DEEP VZN'S TRAJECTORY

Given that Deep VZN is in its nascent stages and has not yet fully launched its characterization phase, there is a critical window to influence its direction. The creators have already demonstrated a willingness to adapt by ceasing 'gain-of-function' research funding. Critics urge public engagement—contacting USAID, elected representatives, and sharing concerns—to advocate for redirecting funds towards less risky aspects of pandemic preparedness, such as broad monitoring and enhanced response capabilities, rather than accelerating the creation and dissemination of potential bioweapons. Ultimately, the goal is to prevent a future where humanity faces pandemics engineered by humans on a scale far exceeding natural threats.

Mentioned in This Episode

●Software & Apps

●Companies

●Organizations

●Books

●Concepts

●People Referenced

Common Questions

Deep VZN (pronounced 'Deep Vision') is a new USAID program aimed at identifying previously unknown pandemic-grade viruses. The concern is that by publicly posting the genomes of these characterized dangerous viruses, it essentially provides 'assembly instructions' for potential bioweapons, thereby increasing the risk of deliberately engineered pandemics. The program costs $125 million over five years.

Topics

Risk Assessment Science & Mathematics Synthetic Biology Public Health Policy Gene Editing Pandemic Prevention Genome Sequencing Politics & Policy Virus Identification Bioweapons Proliferation

Mentioned in this video

People

Sam Harris

Host of the Making Sense podcast and podcasting colleague with Rob Reid, who co-distributes this episode due to the subject's importance.

Kevin Esvelt

Harvard-trained biologist and MIT professor, credited as the first person to describe how CRISPR gene drives could be used to alter wild populations. He specializes in evolutionary engineering and synthetic ecosystems.

Albert Einstein

Co-author of the letter that launched the Manhattan Project, alongside Leo Szilard.

Dennis Carroll

A significant figure in finding and identifying unknown viruses, who ran PREDICT, the predecessor to Deep VZN. He has acknowledged the malicious use of virus genetic sequences but believes the risk already exists, a view that is strongly disagreed with by the speakers.

Leo Szilard

A researcher who, along with Albert Einstein, launched the Manhattan Project, driven by fear of Nazi Germany developing nuclear weapons. He later petitioned against using the atomic bomb on Japan, arguing it would incentivize other countries to develop such weapons.

Rob Reid

Host of the After On podcast, venture capitalist, tech entrepreneur, science fiction author, and science podcaster, who conducts this interview based on his extensive research into pandemic prevention.

Samantha Power

Former U.S. Ambassador to the UN and current head of USAID. She is portrayed as smart, ethical, and having inherited the Deep VZN program, thus not managing it directly.

James Holmes

The Aurora movie theater shooter, a mass murderer who quit his life sciences PhD program before his attack, used as an example of individuals with scientific backgrounds and suicidal/omniscidal instincts.

Ted Kaczynski

The Unabomber, a brilliant mathematics professor who referred to biotechnology's power in his manifesto, illustrating the potential for intelligent individuals with anti-industrial ideologies to misuse scientific knowledge if given access.

Organizations

MIT

The institution where Kevin Esvelt is currently a professor, conducting research in evolutionary engineering.

National Institutes of Health

An agency mentioned as having highly influential figures who have supported "gain of function" research involving lethal viruses.

EcoHealth Alliance

An intermediary through which PREDICT directed a fair amount of money to fund coronavirus research in Wuhan.

USAID

The US agency that distributes most American foreign aid. It is currently running the Deep VZN program, which is the main subject of concern in this podcast.

Centers for Disease Control and Prevention

Conducted research to resurrect and sequence the 1918 influenza virus and posted its genome sequence online, a decision critically analyzed for its security implications.

The Intercept

An online news organization that published a December article interviewing Dennis Carroll and the speakers on the topic of dangerous viruses.

Aum Shinrikyo

An apocalyptic terrorist cult in Japan responsible for chemical weapon attacks in the 1990s.

Al-Qaeda

A terrorist group that called for brothers with microbiology skills to make biological weapons of mass destruction in 2010, indicating a clear intent to misuse such capabilities.

DAESH/ISIS

Terrorist groups mentioned as likely to be tempted to use biological weapons, even if it might harm their own supporters.

World Health Organization

Mentioned in the context of nations arguing about pandemic preparedness, suggesting current plans are less effective than proactive ring vaccination strategies.

Washington State University

Designated by USAID to coordinate most of the scientific work that Deep VZN is funding, making it a central point for scientific engagement on the program.

Wuhan Institute of Virology

PREDICT funded this institute, not just to find coronaviruses, but also to conduct research that involved making chimeras of dangerous viruses, which constitutes 'gain of function' research.

Senate Foreign Relations Subcommittee on State Department and USAID Management

A US Senate committee mentioned as a key point of contact for citizens to express concerns about the Deep VZN program, as it oversees USAID's management.

Drugs & Medications

Malaria

An exceptionally horrible disease that kills nearly half a million people every year, most of them children under five. Gene drive technology was initially envisioned to help eradicate malaria by crashing mosquito populations.

Concepts

Schistosomiasis

A disease that cognitively stunts tens of millions of people annually, which gene drive technology might help eradicate.

Hiroshima

The city where the atomic bomb was used, which Szilard argued would be an 'attention hazard' by advertising the power of nuclear weaponry and incentivizing other nations to develop it.

CRISPR

A genome editing tool described as molecular scissors that can be programmed to cut any desired sequence in DNA. It made the engineering of gene drives possible.

Gene Drive

A naturally occurring phenomenon harnessed through engineering to spread an engineered trait throughout a population of sexually reproducing organisms, even if the trait is harmful. It works by biasing inheritance so that more than half of offspring inherit the particular gene.

Nazi Germany

The concern of Leo Szilard and Albert Einstein when launching the Manhattan Project.

Software & Apps

GISAID

A repository specifically for virus strains, used by virologists to map mutational variants of viruses. Deep VZN also intends to link data to this system.

Stop Spillover

A related USAID program that focuses on identifying highly lethal, non-transmissible viruses that have spilled over, and creates a rank-ordered list of perceived threats. Deep VZN intends to add its findings to this list.

GenBank

The main repository hosted by the US government for all genomic information sequenced and submitted by researchers, where Deep VZN intends to share its data.

Deep VZN

A new, well-intentioned US government program under USAID. Kevin Esvelt argues it is on a path to inadvertently posting assembly instructions for pandemic-grade viruses to the open internet, posing a significant global security threat.

PREDICT

The immediate predecessor to Deep VZN, which identified roughly 1200 previously unidentified viruses. It is implicated in funding coronavirus research in Wuhan labs, which are known to have shoddy safety practices and may have potentially contributed to the COVID-19 pandemic.

Locations

Soviet Union

Predicted by Szilard to acquire nuclear weapons more quickly if the bomb was used on Japan.

Wuhan

The location where PREDICT funded coronavirus research in labs known for shoddy safety, potentially contributing to the COVID-19 lab leak theory.

Companies

Google

The internet search engine used by the speaker to quickly find the 1918 Spanish Flu genome sequence online, demonstrating its public accessibility.

Moderna

A pharmaceutical company famously developed an mRNA vaccine for COVID-19 in less than 48 hours, highlighting the speed of modern vaccine design.

Books

The Problem from Hell: America and the Age of Genocide