Existing mask designs

Stockpile volume and funding

Emergency response

“Not my responsibility”

Financing

Government incentives

Mask designs

Business model innovation

Design and manufacturing

Distribution

Training

Introduction

Cost-effectiveness

What does success look like?

Why now?

If-then

Short-term wins

Longer-term objectives

Thank you!

It’s 2035. A single, lonely virologist working at a BSL-4 laboratory has a mental breakdown. She reads the [Unabomber Manifesto](https://web.cs.ucdavis.edu/~rogaway/classes/188/materials/Industrial%20Society%20and%20Its%20Future.pdf), and concludes that 21st century civilization is evil and must be brought to an end. She steals samples from her lab, and uses the most powerful open-source LLMs, gLMs, her virology expertise and at-home equipment she’s purchased to enhance 10s of pandemic-class pathogens. She spends years creating them in secret. She tests them on human cells and ferrets. Once done, she infects herself with the 3 most viable, dangerous pathogens, and goes on an airport tour to London, Dubai, Singapore and New York. She’s initiated the most severe catastrophe known to mankind. 

This is concerningly plausible. And it’s only one threat model among many: we could fall prey to nation state bioweapons programmes, doomsday cults, or accidental lab leaks. 

What do we do in this scenario? How does humanity respond? 

Firstly, we need to know that novel pandemic-class pathogens are infecting people. We need a [pandemic early warning system](https://naobservatory.org/). 

Once we’ve detected the early infections, we could try [global containment](https://www.gatesnotes.com/meet-the-germ-team), but this is doomed to fail in this scenario given the widespread dispersal at airports. If the pathogens are highly transmissible, they will have travelled to most countries on earth by the time we’re confident something is amiss, by which point the pathogens will be impossible to fully eradicate. They will have established themselves in our global community. 

So we close all borders, and institute severe lockdowns. We deploy the smartest minds towards the mission of producing vaccines, therapeutics, and passive defenses against infections. But it will take months, perhaps years, until we’ve got medical and non-pharmaceutical solutions that allow us to release the lockdown. If everyone’s locked down, who’s producing and delivering food? Who’s running the power plants to ensure we have energy to heat our homes and power our computers? Who’s keeping the water clean and the taps flowing?

A significant fraction of the population are _critical workers_, i.e. they cannot simply stop working or work from home. If the energy, water or food sectors stopped working, then the lockdown would collapse as people leave their homes to loot stores and escape the cities. Millions or billions would die from the ensuing infections, and human civilization would be set back by decades.

To convince critical workers to continue working, they need to be confident that they won’t be infected at work so they don’t endanger themselves or their families. They need pandemic-proof **personal protective equipment**, especially a respiratory face mask.

Today, the US Government has ~10 days worth of N95 masks stockpiled. N95s are only rated to block 95% of airborne particles, and most users don’t wear them properly. They’re better than nothing, but they do not provide adequate protection in this scenario. And the stockpile will be depleted in 10 days. 

If nothing changes and the aforementioned scenario plays out, we predict that billions will die and 21st century human civilization will collapse. 

To avoid this horrendous scenario, governments should stockpile **reusable respirators**, especially **elastomeric half-mask respirators** (EHMRs), to protect critical workers and maintain basic societal function while we develop medical and non-pharmaceutical interventions to prevent mass death. It’s critical that we stockpile these masks _before_ a pandemic, because they need to be deployed immediately in the event of an outbreak.


COVID-19 cost the world >$20 trillion, and killed >20M people. The deployment of multiple enhanced pathogens with a 10%+ infection fatality rate could kill 100Ms of people and collapse our civilization. 

Building a PPE stockpile for critical workers in the US would cost between $1-20B over 20 years. That’s $50M to $1B/year — 0.0007% and 0.014% of the US Government’s $7T in federal spending each year.

The US military spends $1B every 10 hours.


For PPE stockpiles, success means having enough high-quality PPE that critical workers trust and know how to wear, such that they safely continue going to work. 

We believe that reusable respirators (RR) are the best option for PPE stockpiles, as their lifetime cost is cheaper and their effectiveness is higher than disposable masks like surgical masks or N95s. 

Around 15% of the population are critical for preventing societal collapse. The US would need ~50M RRs stockpiled, to be rapidly distributed in the event of an outbreak. 

The stockpile could be bootstrapped by philanthropy, but will ultimately need government funding. Or a private company could develop innovative business models for PPE stockpiling, with critical workers or organisations as customers.

The RRs would need to be designed to fit a diverse population - current masks don’t fit women or non-white ethnic groups well. Their shelf life should be extended to the maximum effective duration (20 years?), to reduce the overall stockpile costs. Critical workers need training in how to don/doff and decontaminate the RRs, ideally before a pandemic. The government should also stockpile a small number of PAPRs for those who can’t use RRs, due to beards or religious headgear. 

In the event of a pandemic being declared, RRs should be deployed to critical workers within a few days, ideally within 24 hours. The stockpiles should be decentralised and pre-positioned near the critical workers to avoid a single point of failure. 


COVID demonstrated how vulnerable we are to pandemics, but governments have done almost nothing since to protect their citizens from the next pandemic. It also demonstrated that scaling up domestic PPE production takes many months, and countries will restrict the export of critical products during emergencies.

AI companies are building technologies which aid the design and creation of pandemic-class pathogens. Every AI model is more capable and dangerous than the last. 

The world is increasingly unstable, with multiple regional conflicts and growing hegemonic tensions between the US and China. 

The technology for building effective, pandemic-proof PPE exists today. There are no major technological breakthroughs required.


PPE Stockpiles

What does success look like?