Category: Uncategorized

A Tale of Two Conferences Part I: UNDRR Global Platform 2025

(In-depth read, 18 min)

TLDR/Summary

• Global disaster costs are exploding while responses lag: Direct losses have reached US$200 billion annually, but cascading effects arguably push the true cost to $2 trillion. Despite the comprehensive Sendai Framework adopted by 187 countries, disaster impacts are actually increasing globally.
• The UN Office for Disaster Risk Reduction (UNDRR) Global Platform (Geneva, June 2–6) made some progress: Thousands gathered with genuine commitment, the World Bank has pivoted to prioritize 80% of disaster funding toward prevention, and there was acknowledgment that current frameworks need to be “more dynamic and powerful” by 2030.
• We’re trapped in a systemic bind: The conference focused on early warning, coordination and engagement issues, specific hazards and financing, touching only lightly on the global stressors resulting from present day human systems that generate risks with increasing rapidity and intensity.
• There was little discussion of exponential technologies that generate new vulnerabilities faster than we can assess or regulate them, while increasing global connectivity makes systems more fragile to cascading failures.
• The meeting missed the deeper “metacrisis“, ignoring the competitive and evolutionary dynamics between nations and corporations that prevent needed collective action. Individually rational decisions lead to collectively destructive outcomes. Disaster risk reduction remains inadequate due to short-term competitive pressures that reward immediate advantage over long-term survival.
• Small islands illustrate the impossible position: Island nations face disaster costs up to 64% of GDP from single events, yet they can’t control the global climate, sea level rise or economic systems that generate many risks they face.
• Real resilience requires transformation: Rather than just better disaster response, we need to change the fundamental incentive structures and governance mechanisms that generate cascading disasters in the first place.
• The world’s current trajectory points to a “hard landing”: My assessment is that without addressing underlying drivers, the mismatch between accelerating risks and response capacity likely means a “hard landing” is ahead. That is, systemic reorganisation in coming decades with reduced global living standards.
• Bottom Line: The UNDRR conference showed remarkable dedication to resilience, but until we address the systemic stressors generating risks faster than we can manage them, and the game-theoretic and evolutionary drivers of those stressors, then we’ll remain stuck in an increasingly dangerous reactive cycle.
• In Part II of this ‘Tale of Two Conferences’ I’ll present a dash of hope from the second meeting I attended, the Accelerator for Systemic Risk Assessment (Paris, June 7–9), which countered some of the despair I felt following the UN Global Platform… Watch this space…

The 2025 UN Global Platform on Disaster Risk Reduction

The costs of disasters continue to spiral upward. In 2020 alone, direct disaster losses reached $200 billion annually, but when cascading and ecosystem impacts are included, the true cost balloons to an estimated $2 trillion. Meanwhile, risks are growing faster than our capacity to address them, disaster risk reduction financing remains woefully inadequate, and high-income countries are discovering they’re not immune—just weeks ago, a Swiss village was wiped off the face of the Earth by a glacier collapse.

Against this backdrop, the UNDRR 2025 Global Platform for Disaster Risk Reduction convened in Geneva, bringing together thousands of stakeholders to advance the Sendai Framework for Disaster Risk Reduction 2015–2030. The meeting culminated in an eight-point call to action.

I attended representing Islands for the Future of Humanity, a non-partisan New Zealand charity think tank, focused on enhancing understanding of global catastrophic risk and mitigation options, through technical research, blogs, collaborative webinars, and other resources available to all for free.

The UN conference structure was comprehensive: keynotes, plenaries, high-level sessions, and learning labs, all open to diverse stakeholders. Yet while the discussions were wide-ranging and often insightful, they revealed a critical blind spot that may undermine all our disaster risk reduction efforts: a failure to grapple with the deeper systemic forces, what some analysts call the “polycrisis” and the “metacrisis”, that are driving humanity’s accelerating risk trajectory.

Setting the Stage: Where We Stand

To understand what was missing, it’s worth reviewing where we stand.

The Sendai Framework, adopted by 187 countries in 2015, provides a comprehensive approach to reducing disaster risk through four priority areas:

• Understanding disaster risk
• Strengthening governance
• Investing in resilience
• Enhancing preparedness for “Build Back Better” recovery.

The aim is to meet seven global targets by 2030, including substantially reducing disaster mortality, economic losses, and damage to critical infrastructure.

The 2023 Sendai Midterm Review revealed mixed progress at best. While some advancement had been made since 2015, disaster impacts are actually increasing and setting back development gains globally. There’s a significant disconnect between policy development and actual practice, with disaster response still prioritised over prevention. Major shortfalls persist in disaster risk reduction funding, meaningful stakeholder inclusion, and international cooperation. COVID-19 exposed critical risk governance weaknesses and missed opportunities for “building back better.”

The recently adopted UN Pact for the Future (2024) acknowledges in its second paragraph that humanity is “confronted by rising catastrophic and existential risks” that could lead to “persistent crisis and breakdown.” It explicitly recognises nuclear war as an existential threat (p.12) and addresses biorisks, climate urgency, and complex global shocks. Yet implementation mechanisms remain underdeveloped.

This year the 2025 UN Global Assessment Report on Disaster Risk Reduction (GAR) reported that five hazards—earthquakes, floods, storms, droughts, and heatwaves—drive 95% of economic losses (this of course omits pandemics, which result in more harm than all other disasters combined!). The report notes that the cost of disasters is rising year-on-year and that many impacts are not quantified by key international databases (see Figure).

But these statistics don’t yet include the potential for truly catastrophic tail risks and the 2025 GAR report makes a crucial observation:

The greatest under-estimation of the potential future disaster costs is the blind spot in accounting for possible 1-in-100 or even 1-in-1,000-year events—those that, while having a low probability of occurring can cause catastrophic impacts when they do (p.27).

Our own research concurs with this sentiment and we have previously argued for such low probability high-impact scenarios to be formally included in national risk assessments. Expected annualised harms from catastrophic tails risks, on some analyses, are many fold higher than the sum of impacts of common disasters.

Finally, the new UNDRR Hazard Information Profiles 2025 attempt to standardise hazard classification but reveal telling omissions. While they include nuclear agents and radioactive materials as technological hazards, there’s no mention of nuclear war or nuclear winter, despite “armed conflict” being listed elsewhere. Pandemic risks are covered, as are catastrophic asteroid impacts and large volcanic eruptions, but there’s no discussion of food shortages or famines—despite historical evidence that major volcanic eruptions have triggered both through climate impacts.

The 2025 UNDRR Global Platform: Familiar Themes, Persistent Gaps

With the above context in mind, I attended the June 2–6 (Geneva, Switzerland) UNDRR 2025 Global Platform, which featured three preparatory events that set familiar themes:

• The Early Warnings for All Multistakeholder Forum aimed to advance global early warning systems by 2027 but identified persistent gaps: under-resourced systems, lack of forecasting data, and financing shortfalls.
• The Stakeholder Forum for Disaster Risk Recovery emphasised that stakeholders shouldn’t wait for ideal conditions to act, calling for fast, flexible, open-source data systems and the need to “build forward better” rather than just “build back better.”
• The World Resilience Recovery Forum launched a 10-point Action Plan, arguing we should stop prioritising speed of recovery over resilience.

These themes continued throughout the main conference. A panellist at the session on “Aligning the Sendai Midterm Review with the Pact for the Future” noted that by 2030, the world will need something more dynamic and powerful than current frameworks to meet emerging challenges. The emphasis on “risk-informed investment” was constant. We must stop creating risk through our investments as we build out complex infrastructure systems that create systemic vulnerabilities for those who depend on them.

Multiple sessions addressed financing, with the World Bank pivoting in recent times to providing 80% of disaster financing for risk reduction and preparedness and 0% to response without disaster risk reduction elements. This is a significant shift from 15 years ago when recovery was the focus. Yet the financing numbers discussed were manifestly disproportionate to the scale of the risks the world faces.

Infrastructure resilience received significant attention, with panellists discussing renewable energy transitions, governance transparency, universal accessibility, and nature-based solutions. Yet, evidence was cited that for every dollar invested in infrastructure, 50 cents is wasted due to organisational culture, corruption (or corporate greed/gaming), and similar inefficiencies. This is a sobering reminder that technical solutions alone are insufficient and systems and goals need to incentivise efficient behaviour (this may be particularly salient for the construction sector in my country of New Zealand).

The Systemic Risk Session: Missing the Mark

Perhaps most revealing was the session on “Understanding Systemic Risk.” Despite its promising title, the discussion largely focused on familiar territory: the need for cross-sector integration on risk, coordination mechanisms, strategic preparedness drills, incorporating Indigenous knowledge, and expanding beyond natural hazards to address social risks like health and food security.

Brendan Moon of the Australian National Emergency Management Agency described Australia’s National Coordination Mechanism that provides a model for genuine multi-stakeholder coordination that goes beyond information sharing to address responsibility and communication across sectors.

These are important topics, but they fundamentally missed what “systemic risk” actually means. The panellists discussed breaking down silos and improving coordination, necessary but insufficient responses to the deeper challenge of understanding why risks are accelerating faster than our capacity to manage them and failing to engage with the possibility that human systems could actually collapse under mounting stressors.

Ruth Richardson from the Accelerator for Systemic Risk Assessment (ASRA) provided one of the few insights that approached the real issue: We need to change the goals of our systems or we won’t overcome the problems. A food system should have the goal of providing nutritious food to everyone, not the goal of providing fast cheap food to whoever can pay.

She announced ASRA’s launch of STEER (still in beta), a system to assess systemic risks and guide action, calling for a radical increase in funding for this work. But even this may not dig deep enough into the polycrisis and metacrisis that is driving humanity’s predicament.

Do we face a Polycrisis?

Thomas Homer-Dixon, a systems theorist and founder of the Cascade Institute, frames the challenge as a debate between those who see our “polycrisis”—the interconnected set of global challenges from climate change to political instability to technological disruption—as merely “history happening” versus those who see it as the result of systemic flaws in our economic and political systems that are driving stressors that are inexorably leading to a dangerous disequilibrium in human systems.

Homer-Dixon identifies four meta-processes driving the polycrisis: increasing energy consumption, disruption to Earth’s energy balance, increasing human biomass, and increasing connectivity among human populations. These processes create conditions ripe for any future failures to cascade across highly connected, homogeneous systems. Exactly the kind of infrastructure networks we’re building out globally in terms of smart cities, global supply chains, food systems, or digital financial networks.

As connectivity increases without corresponding diversity in systems, we create systemic vulnerabilities. A pandemic can spread globally in weeks. A cyberattack can cascade across interconnected financial systems. Supply chain disruptions ripple through just-in-time production networks. Climate change affects agricultural systems worldwide simultaneously. Each trigger can propagate to a crisis and each crisis is a trigger for other crises.

This creates systemic risk.

Systemic Risk: the potential for multiple, increasingly severe, abrupt, differentiated yet interconnected, and potentially long-lasting and complex impacts on coupled natural and human systems – Accelerator for Systemic Risk Assessment

The great question is whether technological acceleration is net negative (contributing to these problems) or net positive (likely to solve them). This remains unknown, but what’s clear is that our current trajectory is unsustainable.

The Missing Metacrisis: Why Disaster Risk Reduction Can’t Keep Pace

The UNDRR Global Platform was largely focused on specific hazards, climate change adaptation, familiar themes of collaboration, community engagement, early warnings and financing. What wasn’t effectively discussed were the fundamental drivers of how humanity has ended up in a predicament where risks are increasing faster than our ability to respond. This a critical omission that undermines virtually all disaster risk reduction efforts.

Several analysts have previously identified this deeper pattern. Daniel Schmachtenberger describes what he terms a “metacrisis”. The metacrisis is characterised by systemic mechanisms underlying all other stressors and crises. Humanity is caught amid rivalrous dynamics where competition between actors (nations, corporations, individuals) prevents the coordination needed to address collective threats.  Exponential technological advancement creates new risks faster than we can assess or regulate them. Resource degradation accelerates while coordination failures prevent effective response. This has led to what Schmachtenberger calls a “sensemaking crisis”—we literally don’t see the problem at the right level of analysis and don’t understand it.

The result is a kind of systemic trap. If we continue on our current path, Schmachtenberger argues, we face either chaotic breakdown or totalitarian restriction of freedoms as authorities attempt to manage escalating crises. Neither outcome preserves what we value about human civilisation. The challenge is to find a “third attractor,” namely a path that avoids both outcomes through fundamental systemic change.

I did not hear this conversation at the UNDRR Global Platform – literally the world’s premiere meeting on disaster risk.

The problematic dynamic has been described metaphorically in terms of “Moloch” (a word sometimes used to describe an ancient bull-headed demon) – a situation where individually rational decisions lead to collectively irrational and destructive outcomes. Each actor, whether a nation, corporation, or individual, faces incentives that make immediate competitive advantage more important than long-term collective survival. It is a classic but nuanced game-theoretic bind, from which we may fail to escape.

Both Homer-Dixon and Schmachtenberger point to a crucial insight: elite institutions and individuals often lack the cognitive framework to understand these meta-level dynamics. They remain focused on specific problems rather than the systemic patterns that generate multiple problems simultaneously.

The Consequences of Missing the Metacrisis

This analytical blind spot has profound implications for disaster risk reduction. All the coordination mechanisms, financing innovations, data and technological solutions discussed at the UNDRR conference operate within a system that continues generating risks faster than we can manage them, driven by forces beyond any agent’s control. It’s like trying to bail out a boat without addressing the fact that someone is drilling new holes in the hull.

Consider the disconnect: we know that investing in disaster risk reduction provides returns of 15:1 in some cases (meaning that $1 invested in disaster risk reduction averts $15 of recovery costs down the line). Yet financing remains inadequate. We know that “building back better” after disasters makes communities more resilient, yet political pressures consistently favour rapid reconstruction over thoughtful less hasty rebuilding that transforms communities into resilient ones. We know that early warning systems save lives, yet they remain under-resourced in the places that need them most.

These aren’t primarily technical problems—they’re systemic ones. The same competitive dynamics that create short-term thinking in corporate boardrooms operate in political systems, international negotiations, and even disaster response. The actor who takes time to build resilience may lose out to the actor who prioritises immediate returns.

Aside: Return on Investment for Disaster Risk Reduction

I actually sat down and produced a simple calculator to illustrate this point while waiting between conference sessions. The default values represent my country’s (NZ) GDP and annual disaster losses. You can see immediately the benefits to GDP growth of investing in disaster risk reduction, rather than merely recovery and rebuilding. If governments want to find extra percentage points of GDP growth this is how. You can even impute your own assumptions about the frequency and impact of large catastrophes and find the ideal GDP spend on disaster risk reduction under those assumptions (default model assumptions are at the bottom – note this is just a toy and greatly simplifies complex dynamics).

Small Island Developing States: A Microcosm of Global Challenges

Our charity Islands for the Future of Humanity focuses on the vulnerabilities and resilience options of remote island nations in the face of global catastrophe. The aim is to disseminate non-partisan information that helps ensure three goals:

1. That remote islands (including NZ and Australia) leverage their unique characteristics to protect the wellbeing of their populations (think successful Pacific Island border restrictions during Covid-19).
2. That islands generating surplus develop resilient structures and systems to ensure continued trade in critical goods and services, such as life-saving food during global catastrophe.
3. That remote islands develop global catastrophe resilience and mitigation mechanisms so that they might preserve complex integrated societies should major collapse impact the rest of the world.

In this context, I was interested in the UNDRR Global Platform’s session on Small Island Developing States (SIDS), which provided a particularly clear illustration of both the challenges and some potential solutions.

SIDS face disaster costs that can reach 64% of GDP for a single cyclone, an unsustainable burden that illustrates how existing systems fail to address truly systemic risks. There is no way that such impacts can be fully mitigated in cycles of disaster and perpetual recovery (note eg, my GDP impact calculator above), without the generative forces of polycrisis and metacrisis being addressed.

The Antigua and Barbuda Agenda for SIDS, introduced in 2024, attempts a more systemic approach. Rather than focusing only on disaster response, it aims to build resilient prosperity through enhanced financial access, climate action, and systematic monitoring frameworks. The emphasis on “trusting and using national systems”, rather than imposing external solutions, recognises that sustainable resilience must emerge from local capacity rather than outside intervention.

Yet even these SIDS initiatives operate within the global economic system that generates many of the risks they face. Small islands didn’t create climate change, but they bear disproportionate costs. They can build the most sophisticated early warning systems, but they can’t control global sea level rise or the hurricane patterns it affects.

This illustrates the fundamental challenge: effective disaster risk reduction increasingly requires addressing global systemic risks that no single actor can control.

The Path Forward: Thinking Beyond the Sendai Framework

As I reflected on the conference discussions, a troubling pattern emerged. Despite all the talk of coordination, collaboration, and systemic approaches, the fundamental drivers of accelerating risk remained largely unaddressed, even undiscussed.

This isn’t to dismiss the important work being done on specific disasters, early warning systems, or financing mechanisms. These efforts cost-effectively save lives and reduce suffering. But they operate within a larger system that continues generating risks faster than we can manage them.

One conference participant noted that, “the greatest risk we face today is failing to seize this chance to act boldly.”

I would argue that the greatest risk is failing to recognise the metacrisis drivers of the disaster risk that we face and continuing business and politics as usual while expecting different results.

As I noted above, one panellist at the 2025 UNDRR Global Platform was clear that by 2030, we’ll need something more dynamic and powerful than current frameworks to meet emerging challenges. This assessment is almost certainly correct, but it understates the magnitude of transformation required.

Given the widening mismatch between accelerating risks and our capacity to address their underlying drivers, the world is likely heading for what we might call a “hard landing.” Human systems will adapt, but with significant lag effects. This likely means a reduction in real GDP per capita globally and a drop in mean living standards across the coming decades, possibly accompanied by forced population decline from famine and other disasters as systems reorganise around new constraints.

This isn’t inevitable—there may be technological or social innovations that change the trajectory. But honest assessment suggests that our current approaches, however well-intentioned, are insufficient to address the scale and pace of change we’re experiencing.

The silver lining is that traumatic experiences can drive learning. The question is whether we can accelerate that learning process before the trauma becomes overwhelming.

Conclusion: Beyond Coordination to Transformation

The UNDRR 2025 Global Platform revealed both the commitment of participants and the limitations of current disaster risk reduction approaches. While the focus on coordination, breaking down silos, and scaling up financing addresses real needs, the conference revealed a critical blind spot: our failure to address the systemic forces generating risks faster than we can manage them.

Technical solutions can’t tackle the competitive dynamics, perverse incentives, and cognitive limitations that create what Schmachtenberger calls the metacrisis. Most participants genuinely want to build resilience, but they’re constrained by institutional and economic systems that make truly systemic solutions extremely difficult. It’s unclear whether any individual, organization, or state has sufficient agency to effect the needed changes.

Without frameworks for understanding these deeper patterns, disaster risk reduction remains reactive rather than transformative. We try to improve our responses to specific disasters while underlying dynamics continue accelerating overall risk.

This doesn’t mean abandoning current efforts, but embedding them within a larger understanding of systemic challenges. The real question isn’t whether we can coordinate better disaster responses, but whether we can transform the systems generating cascading disasters in the first place.

That transformation requires moving beyond technical fixes to address fundamental incentive structures, governance mechanisms, and cognitive frameworks shaping human systems. It’s more difficult than improving early warning systems or scaling financing, but may be the only approach that can actually reduce risk trajectories.

UNDRR Global Platform 2025 participants demonstrated remarkable dedication to resilience. The question is whether future conferences will grapple with the deeper systemic changes that such resilience requires.

Watch for Part II of this Tale of Two Conferences where I’ll present content from the Accelerator for Systemic Risk Assessment symposium (Paris, June 7-9). This will feature a dash of hope to counter the above despair…

By Matt Boyd (Adapt Research Ltd & Islands for the Future of Humanity)

(15 min read)

TLDR/Summary

• Analysis of 12 critically acclaimed films depicting abrupt global catastrophe reveals  cinema’s potential role in helping us understand, prepare for, and potentially prevent catastrophic and existential threats
• These films collectively cover major threats including nuclear war, pandemics, asteroid impacts, and artificial intelligence (AI), offering sometimes surprisingly nuanced portrayals of how these risks unfold and might be managed
• Key lessons emerge across prevention (early warning systems, human oversight of critical systems), crisis management (infrastructure resilience, resource allocation), and recovery (knowledge preservation, adaptation to transformed circumstances)
• Risk governance challenges feature prominently, highlighting how institutional design, international cooperation, and public trust could significantly influence outcomes during catastrophic events
• While the films prioritise entertainment, many incorporate a degree of scientific accuracy and technical concepts that make them valuable educational tools for understanding complex risk scenarios
• Cinema doesn’t just reflect our fears but can actively shape policy responses—as demonstrated by how WarGames and The Day After appeared to favourably influence Reagan-era policy towards nuclear arms control
• These narratives suggest our greatest challenges in addressing global catastrophic risks may not be technological but social and institutional, emphasising prevention and anticipatory governance of risk over response
• Some of these films might be suitable for high school level education and film studios should consider filling the gaps in the cinematic global catastrophe corpus or updating enduring themes impactfully, realistically, and saliently, for the present day

Introduction

In 1983 US President Ronald Reagan viewed two commercial films that influenced his thinking and subsequent US policy on nuclear weapons.

The films were The Day After and WarGames. After watching WarGames, in which a computer hacker accidentally triggers nuclear escalation between the US and USSR, Reagan asked General John Vessey if this could really happen. He was told that, “the problem is much worse than you think.”

Subsequent Reagan-era nuclear weapon agreements led to vast reductions in the number of nuclear weapons held by the US and USSR, for example via the Intermediate Range Nuclear Forces Treaty (1987).

In 2023 the Future of Life Institute awarded the screenwriters of these two films the annual Future of Life Award, given to individuals who have helped make today much better than it may otherwise have been.

As a researcher of global catastrophic risks, I wondered what role film might play in building humanity’s immunity to global catastrophe and existential threats. What can cinema teach us about global catastrophic risks?

I’d read a research paper detailing the ‘six scenario archetypes’ of science fiction films set in the future, however, few of these films clearly focused on global catastrophic risks, many involved aliens or speculative technologies, or simply appeared ‘far-fetched’.

So, I asked a large language model (Claude 3.7 Sonnet) to list 20 critically acclaimed films that realistically depict abrupt global catastrophic risks. Next followed some chit chat with Claude to try and enhance the list with films about solar storms, supervolcano eruptions and catastrophic electricity loss as these appeared lacking.

Not wanting to watch anything lacking entertainment value, I then consulted film meta-critic site Rotten Tomatoes with a list of 19 likely films and eliminated those with a critic or audience score below 50% leaving 12 films to watch.

Here’s what I learned…

The Films

• WarGames (1983): A teenager accidentally hacks into a US military supercomputer and nearly starts World War III.
• On the Beach (1959): Survivors of a global nuclear war await inevitable death from radiation in Australia.
• Greenland (2020): A family struggles for survival as a planet-killing comet races toward Earth.
• The Road (2009): A father and his young son journey across a bleak, destroyed US after an unspecified catastrophe.
• The Andromeda Strain (1971): Scientists race to contain a deadly extraterrestrial microorganism threatening humanity.
• I, Robot (2004): A detective races against time to investigate a crime possibly committed by a robot, before AI is rolled out at scale in society.
• Don’t Look Up (2021): Two astronomers try to warn humanity of an approaching comet that will destroy Earth.
• Threads (1984): A chilling depiction of the effects of nuclear war on the UK’s society and environment.
• Fail Safe (1964): A mechanical error sends US bombers to destroy Moscow, and leaders scramble to prevent nuclear war.
• Ex Machina (2014): A young programmer is invited to administer a Turing test to an intelligent humanoid robot.
• The Day After (1983): A dramatic depiction of the effects of a nuclear attack on US citizens in the Midwest.
• Contagion (2011): A deadly virus outbreak leads to a global pandemic as scientists and governments race to contain it.

At least five of the 12 films depicted the threat of or aftermath following nuclear war, including Threads, WarGames, Fail Safe, The Day After, and On the Beach. Two focused on biological risk (Contagion and The Andromeda Strain). Two detailed risks from emerging AI (Ex Machina; I, Robot), and at least two portrayed asteroid or comet impact events (Greenland, Don’t Look Up, and arguably The Road – in which the catastrophe is never specified).

Global Catastrophe Film Ratings (Critic & Audience Scores) by Rotten Tomatoes

These selected films entertain, but they’re also powerful and sometimes nuanced thought experiments about humanity’s greatest threats – the global catastrophic risks that could severely damage human civilisation on a global scale.

Cinema as Sentinel: Disaster Films Illuminate the Path to Catastrophe Prevention

Surviving global catastrophe is ensured if a global catastrophe never strikes and several of these films gesture at catastrophe prevention approaches.

WarGames identified the risks of automation decades before today’s AI safety concerns, demonstrating how removing humans from critical decision loops creates dangerous vulnerabilities with nuclear weapons. The WOPR military computer in the film confuses a simulation with reality, mirroring contemporary concerns about AI’s lack of contextual understanding. Recent scientific studies have shown some widely used AI models exhibit a marked bias toward escalation in crisis scenarios compared to others.

Both Fail Safe and WarGames show how global catastrophe, even nuclear war, could happen accidentally, with no one on either side really wanting it or choosing it. Preventing such a catastrophe might require ensuring that safety systems always have an override or kill switch, no matter how tempting it might be to ensure systems committed to attacks are swift acting, automated, and cannot be interfered with.

The Andromeda Strain showcases the vital importance of robust monitoring systems to prevent pandemics, while Greenland and Don’t Look Up highlight how programmes and infrastructure for detecting space objects provide humanity’s first line of defence against cosmic threats, ensuring the possibility of pre-emptive actions. Although all these could prove useless without effective response mechanisms. Luckily NASA has been experimenting with methods for redirecting asteroids, with some success.

Prevention programmes can, of course, be undermined by governance failures that technological safeguards alone cannot address. Don’t Look Up exposes political short-termism, media ecosystems optimised for engagement rather than facts, and the challenge of communicating low-probability, high-impact risks. Ex Machina makes a compelling case against individual governance of powerful technologies, showing how even brilliant minds require institutional checks and balances that could curtail dangerous research or development programmes before catastrophe strikes.

These narratives capture how things like psychological biases, communication breakdowns, and societal dynamics can undermine even sophisticated technical safeguards. Effective catastrophe prevention requires not just technological solutions but integrated approaches that account for human psychology, institutional design, and communication challenges. Such lessons are perhaps increasingly relevant as humanity’s technological powers continue to outpace our wisdom in managing them.

Crisis Management: When Catastrophe Unfolds on Screen

These films offer insights into what happens if prevention fails. Societies struggle to limit damage, preserve essential functions, and navigate impossible choices under extreme pressure.

The preservation of essential systems emerges as a central theme. Threads and The Day After expose infrastructure fragility, methodically tracing how nuclear attacks trigger cascading failures across interconnected systems.

Threads demonstrates how the complex interconnections of modern society (our infrastructure “threads”) unravel catastrophically when stressed beyond business-as-usual. Hospitals collapse under overwhelming casualties, communication networks fail when needed most, and food supply chains disintegrate with failure of energy systems and in the ensuing nuclear winter. Greenland shows transportation networks buckling as evacuation routes become congested death traps with families obstructing one another’s passage to safety. Failures in one system rapidly spread to others.

Command and coordination challenges receive realistic treatment throughout these narratives. The Andromeda Strain demonstrates the vital importance of assembling diverse expertise rapidly—bringing together specialists in a sealed facility to tackle an unknown pathogen. This cross-disciplinary approach contrasts sharply with Ex Machina, where isolation and information hoarding lead to the potentially catastrophic escape of advanced manipulative AI. Threads shows how leadership decapitation cripples response capabilities. As decision-makers themselves become casualties, command structures collapse precisely when most needed. WarGames depicts military and political leaders struggling to understand complex unfamiliar technological systems while crises accelerate beyond their control.

Resource allocation under scarcity forms another crucial lesson. Contagion captures the tension when jurisdictions compete for limited medical supplies during a pandemic, showing healthcare systems buckling under patient surges while struggling with distribution, a situation familiar to all since the arrival of Covid-19. The Day After forces viewers to confront medical triage during mass casualty events, where doctors must abandon conventional treatment standards and health system collapse is inevitable. Most starkly, Fail Safe presents the ultimate resource allocation dilemma, the need to sacrifice New York City to compensate for Moscow’s accidental bombing. Some catastrophic scenarios may require accepting smaller losses to prevent total annihilation. Greenland shows evacuation transport to be the most valuable resource of all when safety is critically limited, with access to departing aircraft becoming literally life-or-death. We can imagine similar chaotic scenes accessing ships heading for island refuges.

The human dimension of crisis management emerges across these films. Don’t Look Up catalogues the psychological responses to impending doom, from denial and hedonism, to acceptance and connection-seeking. Threads illustrates the tension between family obligations and public duties when people in authority must choose between helping strangers and protecting their own families. Most strikingly in The Road, humanity descends into cannibalism. Extreme scarcity transforms social trust calculations, with cooperation becoming simultaneously more valuable and more dangerous.

Social cohesion proves remarkably fragile in these fictional disasters, yet human connection emerges as a critical resource. Greenland shows both altruistic sacrifice and exploitative behaviour occurring side by side as social norms break down. Contagion and Don’t Look Up identify how information ecosystems become battlegrounds during crises, with misinformation spreading alongside physical threats and commercial incentives often overwhelming survival imperatives, with a tech billionaire’s profit-driven strategy ultimately undermining more conventional response options.

Recovery and Adaptation: Surviving in Transformed Worlds

Some of these films moved past the acute catastrophe into phases of adaptation and transformation. Knowledge preservation emerged as a critical recovery challenge.

Threads delivers this lesson by following UK survivors of nuclear attack through immediate devastation into long-term societal collapse. In the final scenes survivors are speaking a degraded English dialect and lack basic agricultural knowledge. This highlights the importance of knowledge preservation and skill transmission as a critical priority for global catastrophe planning. When systems fail catastrophically, preserving crucial information and skills may ultimately determine whether recovery remains possible at all. The Road approaches this challenge differently, focusing on a father methodically transferring survival skills to his son in preparation for his own death. Their journey demonstrates how knowledge transmission becomes highly personal in the absence of institutional educational systems.

Threads envisions societal restructuring with emergency powers granted during crisis gradually evolving into authoritarian control as recovery proceeds under resource constraints. New governance structures emerge, often bearing little resemblance to pre-disaster institutions. We also see the breakdown of monetary systems.

Environmental adaptation forms another major recovery challenge. The Road depicts a world suffering from extended sunlight-reducing climate disruption after an unspecified catastrophe (perhaps a volcanic winter or asteroid impact). Relentless cold, dying forests, and the absence of wildlife create a setting where adaptation, not restoration, becomes the only viable strategy. Mitigation of this situation becomes all but impossible in the absence of anticipatory planning. We get a glimpse of prescient individual ‘prepping’ when a food bunker is found, but catastrophically there was no equivalent foresight at the societal level.

Threads demonstrates the challenges of severely reduced agricultural yields without modern industrial inputs, showing survivors struggling to farm irradiated soil at scale without fertilisers, pesticides, or machinery. Both films illustrate increasing resource scarcity over time, with The Road showing how easily scavenged items become depleted, forcing survivors to develop new subsistence strategies.

On the Beach offers perhaps the starkest perspective, examining how people face inevitable extinction as radiation slowly approaches their Australian refuge after nuclear war. This grim, although scientifically questionable, portrayal emphasises that some catastrophes cannot be meaningfully “managed” once they occur, underscoring that for certain risks, prevention remains the only viable approach.

These films collectively suggest that true resilience often involves managed transformation rather than restoration. They show societies adapting to new realities rather than returning to previous states. This perspective challenges conventional recovery planning that aims primarily at restoring pre-disaster conditions, suggesting instead that building capacity for transformation may prove equally important for long-term human continuity.

Risk Governance: Institutional Management of Catastrophic and Existential Threats

These films offered some nuanced portrayals of the institutional frameworks and governance challenges that shape how societies prepare for and respond to extreme threats.

Contagion shows the institutions needed to support methodical, time-consuming scientific response, following researchers as they isolate viruses, test vaccines, and navigate regulatory approval processes.

WarGames reveals how organisational incentives shape critical decisions, depicting military systems designed to remove hesitation but inadvertently creating new vulnerabilities. Don’t Look Up portrays the verification processes required for extraordinary scientific claims, showing astronomers meticulously confirming their comet calculations before alerting authorities, only to have their warnings filtered through political and media institutions with competing priorities, and their existential message is lost in the process.

The public-private interface and technological solutionism (the belief that every problem has a market-driven technical fix) is exposed in Don’t Look Up, when the private space mission fails catastrophically. I, Robot similarly illustrates how market priorities can overwhelm safety concerns in AI development, while Contagion shows both the benefits and risks of private control over essential infrastructure and medical supply chains.

Challenges to international cooperation form another crucial governance theme. Don’t Look Up illustrates the difficulties of coordinating global threat response across countries with divergent interests and capabilities. Contagion depicts both successful scientific collaboration across borders and tensions in international aid distribution as nations prioritise their own populations. Fail Safe shows barriers in communication and lack of trust between nuclear powers that subvert de-escalation during a crisis.

Public trust dynamics play a decisive role across these narratives, with Contagion showing how institutional credibility shapes public compliance with emergency measures. Effective communication builds trust and misinformation undermines collective action. Don’t Look Up reveals how elite capture of response systems grants disproportionate decision-making power to wealthy individuals, undermining public confidence in institutional responses. Both films suggest that effective risk governance depends not just on technical capabilities but on maintaining the social contracts that enable coordinated action.

These films collectively suggest that risk governance for global catastrophic threats requires institutions designed specifically for these challenges—systems that align scientific understanding, political incentives, and ethical frameworks toward preserving humanity’s interests as a whole. They reveal how existing governance structures, created primarily for managing routine problems, often prove inadequate when confronting unprecedented threats that cross boundaries between disciplines, nations, and generations.

Hazard-Specific Insights in Film

While these films collectively illustrate general principles of risk management, they also cover specific catastrophic threats, in some cases with accuracy and nuance.

Nuclear conflict received the most comprehensive treatment across these films. Fail Safe demonstrates with chilling precision the path from a technical malfunction to the triggering of accidental nuclear war despite multiple safeguards. The Day After shows deterrence failing through conventional escalation rather than deliberate first nuclear strike. These portrayals challenge simplistic understandings of nuclear stability and highlight vulnerabilities in command and control systems. Threads stands out for incorporating scientific understanding of electromagnetic pulse effects that disable electricity, as well as nuclear winter effects, and showing the much more damaging agricultural collapse and climate disruption rather than just immediate blast damage.

Pandemic threats are depicted methodically in Contagion, which correctly emphasises concepts like R₀ (basic reproduction number) and shows the steps of pharmaceutical response from virus isolation to vaccine development. Similarly, The Andromeda Strain emphasises the value of scientific programmes, contingencies, and pre-allocated resources, all of which may need to be called on when facing a catastrophic biological threat. Contagion portrays zoonotic spillover from habitat encroachment creating novel transmission chains proving eerily prescient before Covid-19, hinting at risk reduction strategies via habitat preservation. The film also visualizes how global air travel serves as a transmission accelerant, with key scenes tracking infection spread through mundane objects like credit cards and door handles—emphasising the mechanical reality of contagion and the potential value of stringent border management, especially for remote jurisdictions.

Asteroid and comet impacts feature in both Greenland and Don’t Look Up. Greenland portrays comet fragment impacts, showing regional rather than global effects from smaller fragments while acknowledging the civilisation-ending potential of the main comet body. Don’t Look Up dwells on the extended timeframes required to prepare a deflection mission, demonstrating the risks of last-minute responses. Both films acknowledge the international dimension of impact threats, showing how response requires coordination across borders and capabilities.

Artificial intelligence risks are explored through different lenses across several films. I, Robot highlights the ‘alignment’ and ‘control’ problems, and demonstrates the insufficiency of rule-based constraints (the Three Laws) when applied to complex systems, showing how logically consistent interpretations can still produce harmful outcomes. Ex Machina portrays alignment failure modes with remarkable subtlety, showing an AI pursuing individual freedom at human expense while developing social manipulation and deception as emergent capabilities. WarGames explores how learning systems can develop unexpected behaviours through iterative self-play, presaging modern concerns about reinforcement learning systems optimising for unintended objectives. We must remember too that AI poses threats via its possible role in escalation to nuclear war, or facilitating biological threats, as above.

Cross-cutting concepts emerge when viewing these films as a collection. Particularly notable is how several films demonstrate how a crisis helps clarify human values. We see how catastrophe forces individuals and societies to reevaluate priorities and reveal previously implicit values. Perhaps some of these values ought to have been protected in advance. Perhaps these films play a role as simulations of possible futures to inform present-day anticipatory governance.

These domain-specific insights, when combined with the broader lessons above about prevention, crisis management, recovery, and governance, offer an interesting and comprehensive education in global catastrophic risk concepts.

While fictional, and sometimes overblown, these cinematic treatments do incorporate many scientific and technical concepts, making them valuable tools for understanding the complex challenges of safeguarding humanity’s long-term future.

As a result, some of these films would be suitable as teaching aids for education about global catastrophic risks, and perhaps Threads or Contagion (well-regarded films representative of two of the biggest threats to humanity) would suit high school social studies classes.

Gaps in this Collection of Films

The 12 films discussed provide broad coverage of key global catastrophic risks and related concepts. I intentionally omitted climate catastrophe films because I felt this is an area already seeing a lot of international policy and response action. However, when searching for these films there seemed to be gaps in this film corpus.

There are very few films on volcanic catastrophes (though arguably The Road), and apparently no quality films about global volcano impacts, yet supervolcanic eruption is a recognised global catastrophic risk. Films such as Dante’s Peak or Volcano depict localised eruptions, and others such as Super Eruption (2011) scored very poorly (6%) on Rotten Tomatoes and didn’t seem worth a viewing.

Additionally, the films viewed predominantly focussed on single hazards, whereas the literature on global catastrophic risks emphasises the potential for cascading and interacting risks, for example an AI catastrophe leading to a pandemic, or a nuclear war leading to biological weapon use or accidental release.

Another theme not often covered in the films was the likelihood of global catastrophes. In each film the catastrophe occurs, but it is difficult to appreciate the relative likelihood of pandemics vs nuclear war vs asteroid/comet impact vs large magnitude volcano eruption vs AI catastrophe. These probabilities are far from equal and ought to inform mitigation efforts. Notably the likelihood of a comet or asteroid strike per century is very low, but the world has recently advanced response efforts considerably, perhaps at some opportunity cost for addressing other catastrophic risks.

Conclusions

So, how would we act today if cinema was our sole advisor on global catastrophic risks? We could:

• Strengthen early detection systems for biological, astronomical, and technological threats, where timely detection is crucial but often undermined.
• Preserve human involvement in critical systems while designing technological safeguards that prevent automated catastrophes, learning that removing humans from decision loops creates dangerous vulnerabilities.
• Build robust scientific institutions that can develop beforehand and rapidly mobilise diverse expertise during crises, following a cross-disciplinary approach rather than the isolated expertise that fails in Ex Machina.
• Reform information ecosystems to prioritise factual reporting on long-term risks over engagement metrics, addressing the media failures portrayed, where entertainment value overshadowed existential threats.
• Develop international coordination mechanisms specifically designed for catastrophic threats that cross national boundaries, avoiding the fragmented responses seen in the films.
• Establish redundant and resilient infrastructure systems for critical services like healthcare, energy, and food production to prevent the cascading failures graphically illustrated.
• Create knowledge preservation protocols and “civilisation restart” technologies to ensure vital skills and information survive even if current institutions fail, preventing the knowledge loss where survivors might lack even basic agricultural skills.
• Invest in technological solutions for specific threats while recognising their limitations, avoiding the technological solutionism that backfires in Don’t Look Up and I, Robot when private interests override public safety.
• Prepare for transformation rather than just restoration after major disruptions, acknowledging as The Road suggests that some post-catastrophe worlds may never return to previous states.
• Prioritise truth and transparency in risk communication, avoiding the manipulation of threat information seen in multiple films where political convenience outweighs factual accuracy.
• Develop emergency resource allocation frameworks that balance immediate needs with long-term recovery, preparing for the difficult triage decisions.
• Cultivate institutional legitimacy and public trust before crises occur, recognising how quickly social cohesion could dissolve when trust is already fragile.
• Design governance systems that align scientific understanding, political incentives, and ethical frameworks toward preserving humanity’s long-term future rather than short-term interests.

Cinema suggests that our greatest challenges in addressing global catastrophic risks may not be technological but social and institutional. The films I watched collectively warn that prevention remains far superior to response, and that our chances of navigating coming dangers depend less on heroic Hollywood individuals than on the social systems and investments we choose to make today. Overall system resilience (economic, ecological, social) might be the most critical factor rather than hazard-specific investments.

There is a potential role for judicious film production as an effort to effect policy that could mitigate global catastrophic risk. WarGames and The Day After were released in the tension-filled year of 1983, these films not only reflected the era’s nuclear anxieties but also played a direct role in favourably shaping policy. They went beyond mere entertainment, acting as significant catalysts for dialogue. These films influenced policymakers and raised public awareness about the grave risks associated with nuclear warfare, and were followed by tangible, preventative action from leaders. These films stand out as relatively sensitive to reality and were also critically acclaimed. It has also been argued that the 1990s asteroid/comet films Deep Impact and Armageddon conceivably drove policy around near-Earth object detection.

It is notable that audience top scores (on Rotten Tomatoes) went to grim but gripping depictions of catastrophe, without happy endings. There is perhaps nothing stopping studios pursuing these kinds of films and filling the gaps in the cinematic global catastrophe corpus or updating enduring themes impactfully, realistically, and saliently, for the present day.

By Adapt Research & Islands for the Future of Humanity

TLDR/Summary

1. Urban Agriculture Contributes to Food Security but isn’t Sufficient: Urban agriculture alone could feed ~20% of New Zealand’s urban population post-catastrophe, but requires supplementation.
2. Near-Urban Agriculture Completes the Picture: Adding optimised near-urban agriculture requires surprisingly little land (~1,140 hectares for a city of 90,000 people) to achieve food self-sufficiency.
3. Crop Selection is Critical: The most efficient crops vary by scenario – peas and potatoes in normal climatic conditions; sugar beet/spinach and wheat/carrots during nuclear winter.
4. Export Diversion Provides a Buffer: New Zealand’s food exports could feed 3.9× its population (1.5× even in nuclear winter), but are primarily dairy which requires much more fuel to produce (eg, compared to wheat).
5. Fuel Dependency is the Achilles’ Heel: New Zealand would run out of stockpiled liquid fuel after around 160 days in a prolonged catastrophe.
6. Biofuel Solution is Viable: Just 4,400 hectares of canola (1% of grain-cropped land) could produce sufficient biodiesel to maintain essential food production.
7. Implementation Requires Planning: Success depends on advance preparation including urban land readiness, processing infrastructure, seed availability, and regulatory frameworks.
8. NZ has Strategic Importance: As a potential global refuge after a catastrophe, this country could feed over 8× its population if optimised resilience measures are implemented.

Our new research demonstrates that with relatively modest investments and strategic planning focused on urban/near-urban agriculture and local biofuel production, New Zealand could significantly enhance its food security resilience during global catastrophes.

Our new research on urban and near urban agriculture for food security

Our latest study on urban- and near-urban agriculture, published in the international journal PLOS ONE, completes a series of papers showing how New Zealand could maintain food security during global catastrophes that collapse trade, such as extreme pandemics, nuclear war, or severe solar storms disabling global electrical supply.

When combined with our previous work on New Zealand’s export food excess, frost-resistant crops, and agricultural fuel needs, we can now present a more complete food security blueprint—revealing both vulnerabilities and strengths in our food systems that could determine whether New Zealand thrives or collapses in a scenario where global trade ceases.

In the new paper, “Resilience to abrupt global catastrophic risks disrupting trade: Combining urban and near-urban agriculture in a quantified case study of a globally median-sized city”, we analyse how scaled-up urban agriculture (UA) combined with near-urban industrial agriculture could help feed a New Zealand city during a global catastrophe.

We established these results through mathematical optimisation for protein and food energy given available land. Urban agriculture yields were drawn from a published meta-analysis, and we calculated potentially cultivable urban land area (particularly home lawns and parks) using Google Earth imagery, as demonstrated in the figure below:

For estimating the near-urban (city fringe) land required to supplement urban agriculture, we referred to our previous optimisation research on minimising land area and liquid fuel requirements in a global catastrophe. We assumed only off-road diesel was needed, not transport fuel, given proximity to the city.

Key findings from our new urban food security research

Our analysis reveals that urban agriculture alone could feed approximately 20% of New Zealand’s urban population in normal climate conditions (less with the reduced yields of nuclear winter).

This suggests that urban agriculture will not fully meet a city’s food security requirements, though previous studies have shown it can meet fruit and vegetable needs. Additional protein and food energy sources remain necessary.

However, when combined with optimised near-urban industrial agriculture on relatively modest land areas, the entire population of Palmerston North (our case study city) could be fed while significantly reducing transport fuel requirements. Local production of a small volume of liquid biofuel could effectively provide survival-level food self-sufficiency for the city.

We found that:

• The optimal crops for urban agriculture were peas (in normal climate) and sugar beet/spinach (in nuclear winter) in terms of protein and food energy yield per area.
• For near-urban industrial agriculture, potatoes (normal climate) and wheat (97%)/carrots (3%) (nuclear winter) were optimal.
• Relatively little near-urban land—just 1,140 hectares for Palmerston North (~90,000 population)—would be needed to make up the shortfall from urban agriculture. This equates roughly to a ring around the city less than 1km wide, though it could be configured to follow the most fertile nearby soils, or transport routes.
• Just another 110 hectares for a biofuel feedstock such as canola seed could provide sufficient biodiesel to run the necessary agricultural machinery.

The figure below shows the crop optimisation results, expressed as land area needed to meet the protein and energy requirements for one person for one year. The figure shows only the most optimal crops. All other crops required more land area (implying more industrial inputs such as liquid fuel, fertiliser, etc. which might be scarce following a global catastrophe). The top panel assumes a normal climate global catastrophic risk (GCR) scenario and shows that peas in urban agriculture spaces feed more people than other crops, and potatoes grown on the city fringe require the least additional near-urban land. The lower panel assumes a severe nuclear winter scenario.

These results are illustrated schematically in the next figure, showing that we identified 730 hectares of potential urban agricultural land within Palmerston North. An area of at least 1250 hectares (39% the size of the built urban environment) is needed near the city to produce the rest of the required food and a biofuel feedstock in a no-fuel scenario.

Building on our previous research

This work complements our earlier papers exploring different aspects of New Zealand’s food and energy resilience to global catastrophes:

Diversion of Food that NZ Currently Exports: New Zealand’s current food exports provide more than 3.9 times the dietary energy needed for the entire population. Even in a severe nuclear winter scenario with agricultural productivity reduced by up to 61% (as estimated for NZ in a global study), diverted exports could still provide 1.5 times current New Zealand dietary energy requirements. The challenge is that production is largely dairy milk solids, which require much more liquid fuel and land to produce than crops like wheat.

Frost Resistant Crop Production: We identified optimal combinations of frost-resistant crops for nuclear winter scenarios. The most land-efficient options were wheat and carrots; sugar beet; oats; onions and carrots; cabbage and barley; canola and cabbage; linseed and parsnip; rye and lupins; swede and field beans; and cauliflower. Under current production levels of wheat and carrot, there would be a 71% shortfall in a severe nuclear winter scenario. But although 117,000 ha of a wheat (97%) and carrots (3%) combination could feed all New Zealanders in normal conditions, 300,000 ha is needed in severe nuclear winter. This contrasts with current wheat production of just 45,000 ha (NZ currently imports much of its wheat from Australia).

Mitigating Imported Fuel Dependency: This previous study in the journal Risk Analysis highlighted that New Zealand uses over 3.7 billion litres of diesel annually but has limited onshore stockholdings. Agriculture alone consumes 295 million litres per year. Our modelling showed that the ‘bare minimum’ liquid fuel requirements for agricultural production would require between 0.14% and 2.8% of New Zealand’s total annual diesel consumption, depending on crop selection, transport distances, and climate conditions (including nuclear winter). The graph shows how different the liquid fuel requirements are to supply protein and food energy based on food production type.

NZ’s National Fuel Security Study

The Government’s recently commissioned New Zealand Fuel Security Study modelled a “severe disruption” with complete cessation of fuel imports for 90 days. Essential services and critical government functions would require approximately 5% of normal diesel demand for lifeline utilities and potentially another 5–15% for critical transport (totalling up to 20% of business-as-usual demand).

However, as we detailed in a blog, the study doesn’t include off-road agricultural fuel needs, which would add up to 2.8% more diesel (as we’ve noted above, eg, 107 million litres for dairy production under nuclear winter), resulting in as much as 22.8% for essential needs. It could be that this remains an underestimate as the interdependencies among essential and ‘non-essential’ services are not well understood.

The Study concludes that New Zealand could manage a 90-day period with no liquid fuel imports, but our estimates indicate that even with judicious use of stockpiled diesel, the country would run out after approximately 160 days. If a catastrophe prevents fuel imports beyond this timeframe, even with reduced consumption, New Zealand would still run out of onshore liquid fuel. This “point of breakdown” isn’t apparent from the government’s commissioned analysis, highlighting a critical blind spot in national resilience planning.

The Biofuel Solution

Our most significant finding is that New Zealand could achieve a sustainable fuel supply for essential agriculture through relatively modest domestic biofuel production.

We illustrate this in the following figure, which contrasts current wheat, potato and dairy production land areas, and shows the land required for a biofuel feedstock like canola seed to ensure liquid biofuel for farm machinery and transportation (in our normal climate, short transport distances scenario).

• Just 4,400 hectares (ha) of canola (approximately 1% of currently grain-cropped land) is needed to produce sufficient biodiesel to sustain wheat cultivation equivalent to feeding the entire population
• A refining capacity of just 5-15 million litres of biodiesel annually could maintain this minimum food production (varying according to transport distances and crop selection, eg, wheat vs potatoes).
• Canola is already grown commercially in New Zealand for producing food oil and has previously been refined for biodiesel
• Canola is also relatively frost-resistant, making it suitable for nuclear winter scenarios

The National Fuel Security Study done for the New Zealand Government, included biofuels as a potential solution but found it expensive compared to increasing liquid fuel storage or distribution capacity. However, neither of these alternatives is a long-term solution. We suggest analysing small-scale, local, modular biodiesel production and processing units that could provide distributed resilience.

Putting it all together

Combining our studies, several strategies for increasing food security and reducing vulnerability to liquid fuel shocks emerge. Focusing on highly efficient crops and growing them near processing and consumption centres reduces transportation fuel demands. Scaling up urban agriculture in city green spaces along with expanded production of high-yield crops on city-adjacent land further minimises liquid fuel demand. This approach would bring agricultural fuel and transport demands within reach of a modestly enhanced locally produced liquid biofuel capacity, providing sustainable food production and transport through an ongoing global catastrophe.

We’re not suggesting a complete biofuel economy, but rather exploring small-scale biofuel production, with a transition to less land-hungry food production processes (eg, more balance between dairy and wheat), with production closer to processing and consumption, and more urban agriculture. This approach would reduce cultivated land area, fuel consumption, and climate emissions.

Figure Key Insights:

• Urban agriculture combined with near-urban crops significantly reduces both fuel needs and land requirements
• Potatoes are the most land-efficient crop (84,000 ha vs 117,000 ha for wheat/carrot)
• Wheat/carrot combination is the most fuel-efficient (5.4 million L at 20km transport vs 21.9 million L for potatoes at 100km)
• Both analyses show major efficiency improvements in terms of fuel and land use compared to business-as-usual dairy production

Implementation opportunities

Successfully implementing these food resilience strategies would require addressing several practical challenges:

• Logistics and agreements for diverting excess export food to the domestic market during trade collapse
• Production of seed (especially frost-resistant crops) and agricultural inputs
• Preparation of urban land for cultivation, and development of expertise
• Processing infrastructure for optimal crops and biofuel feedstock (locally)
• Land use planning including better integration of agricultural production potential into urban and near-urban planning policy
• A comprehensive national fuel security plan that considers catastrophes
• Including food insecurity in the National Risk Register

The Department of the Prime Minister and Cabinet’s (DPMC’s) Risk and Resilience Framework sets the expectation for decisive, impactful action to prevent or reduce potential crises – all of the above could be tested and implemented for global catastrophe resilience.

It’s time to start pilot programmes in New Zealand, testing the feasibility of these approaches to food and fuel security. Which New Zealand city will become the world’s first median-sized city that is self-sufficient for basic food needs in a catastrophe?

Much could be done through developing plans, regulatory levers, incentivising appropriate agricultural processes, and with targeted government support. The tension between efficiency and resilience runs through this research. Our modern systems are optimised for efficiency under normal conditions but may be extremely vulnerable to disruptions. Building resilience often requires maintaining seemingly redundant systems as insurance against catastrophe. Yet ensuring a minimum food supply for the population during major global catastrophes is arguably a core government responsibility.

Conclusion

It is concerning that New Zealand would run out of liquid fuel (hindering food production and transport) beyond a few months in an ongoing global catastrophe. It’s also concerning that there is insufficient cultivation of frost-resistant crops to handle abrupt sunlight reduction (nuclear, volcanic, or asteroid/comet impact winter).

However, a surprisingly small volume of liquid fuel is required for food production if steps are taken to ensure minimal cultivation of high-protein, high-energy, fuel-efficient crops near processing and consumption areas. Transport and fuel demands can be further reduced by scaling up urban agriculture and cultivating optimal crops within cities and adjacent to them. A small amount of locally produced liquid fuel could then sustain these processes without global trade.

Our findings suggest that with modest investments and strategic planning, New Zealand could significantly enhance its ability to avoid famine during even severe global catastrophes that disrupt fuel supplies.

These issues are particularly salient for New Zealand as an island nation. New Zealand can potentially feed eight times its population through food exports of the most efficient crops if these can be sustained through catastrophe. New Zealand is also often cited as a potential ‘refuge’ for humanity where societal complexity might persist through a global catastrophe. Securing sustainable fuel supply for essential functions, alongside accelerating electrification, is central to this vision. Collaboration with other regional food-producing islands like Australia would be prudent.

Direct link to watch the full webinar recording here.

TLDR/Summary

• The threat: A catastrophic electricity loss (affecting 5%+ of global population for over a month) could be triggered by solar storms, cyberattacks, or high-altitude nuclear explosions—with minimal planning in place for such scenarios.
• The danger: These events would cripple supply chains, food production, and financial systems simultaneously, potentially halving food supplies even in resource-rich countries.
• Solutions exist: Resilient electrical systems can be built through system hardening, geographical distribution, micro-grids, and regulatory reforms—but require deliberate planning and investment.
• Basic resilience helps: Household stockpiles (2+ weeks of supplies) and social cohesion are critical for surviving medium-term outages, especially in urban areas.
• Public support: Survey evidence shows citizens want catastrophe planning and resilience investments; community exercises could generate widely-supported policies.
• Action needed now: 2025 is marking an age of increasing geopolitical instability and risk requiring urgent focus on things like infrastructure interdependencies, near-urban agriculture, and comprehensive resilience planning.
• Cross-cutting benefits: Preparing for one catastrophic scenario effectively prepares us for many others, creating multiple resilience benefits.
• Get involved: Multiple leverage points exist for action and input. In the Aotearoa New Zealand context this includes: DPMC’s Long-term Insights Briefing, the Infrastructure Commission’s Priorities Programme, various sector security strategies, and emergency planning processes (by NEMA).

Watch the full webinar video

What was the background and context to the webinar?

In 2023 our group published a report on Aotearoa New Zealand’s (NZ’s) vulnerability and resilience to a Northern Hemisphere nuclear war scenario. NZ is on the one hand a remote island nation with abundant resources, and in a sense resilient. But on the other hand, critical societal functions are very vulnerable to interruptions to trade and supply. 

Many islands share this double-edged sword, and it is worth investing significant efforts in resilience to ensure not just the wellbeing of their populations, but the ability to export food excess in times of crisis, and to sustain hubs of functioning society should a terrible catastrophe strike the world. 

With the aim of furthering deliberation and planning around these issues, we convened a webinar on Catastrophic Electricity Loss, potentially caused by a massive solar storm, cyberattack, or worryingly, a high-altitude nuclear explosion with an electromagnetic pulse (EMP). 

In such scenarios there may be no external aid, for a long period, with crippled electrical supply affecting the flow of all goods and services. 

NZ has little planning for such scenarios, although to its credit, the National Emergency Management Agency (NEMA), has recently published a Space Weather Response Plan, and a Catastrophic Event Handbook. 

However, these plans, by design, do not look beyond the early recovery phase, and do not address options for national investments in resilience upstream of any catastrophe. They also assume there is a functioning port and airport, and that international aid will come. These assumptions may not hold, and in fact need to be ensured through prior actions. 

Existing international studies of catastrophic electricity loss detail a quick descent into national crisis, and the potential for a halving in food supply, even in highly food productive countries such as the US. 

The dense interdependencies among utility systems mean that electricity could be hard to restore, and existing redundancies could be simultaneously impacted by actual hazards. There is plenty of scope for critical system redesign. 

Existing vulnerability assessments are incomplete, with a focus on short-term needs like hospitals, rather than ‘slow’ processes affected by long-term outages, such as food production and processing, or shipping freight logistics. In some cases, analysis of payment systems has been left for future work. 

2025 is no longer business as usual, we are entering an age of increased geopolitical instability and risk. 

Survey evidence (eg, from the NZ and the UK) shows that the public wants catastrophe plans, and investments in national resilience. It is time to engage society more widely on these issues, provide options, and implement solutions. 

What is catastrophic electricity loss?

Catastrophic electricity loss can be loosely defined as 5 percent or more of the world’s population not having electricity for more than a month. This could result from the causes listed above. Dr Simon Blouin, lead author of a study of food supply in the US following electricity failure explains in the interview below (see 1:20 for definition of catastrophic electricity loss). 

Who was on our webinar Panel? 

Our webinar on Catastrophic Electricity Loss (26 Feb 2025) saw a great line up of experts on the panel:

What did the Panellists say?

The following briefly summarises selected key points made by our panellists during the event. Watch the YouTube video to see their comments in full and in-context. 

Dr David Korowicz 

• Catastrophic electricity loss is such a huge concern due to its impact on supply chains and the flow of goods and services. 
• This impact is common to a range of risks (including events affecting liquid fuel supply for example). This means that if you prepare for one of these scenarios you have substantively prepared for all of them. 
• This kind of event has critical implications for the financial system, and how people perceive money. This is because money has value based on our expectations of future production. If the future is highly uncertain, then the money system can collapse. 
• Community exercises that walk-through catastrophe scenarios are both engaging to people, and useful for planning. Governments don’t have to be afraid of talking to people about these issues. 
• The psychological notion of ‘scarcity’ is a big driver of behaviour. We need to work towards a mindset of ‘how can I help?’
• Other panellists concurred that policy that emerges from citizen assemblies and similar exercises is more likely to be widely supported. 

Associate Professor Caroline Orchiston

• Both resilience and sustainability are important. Basic resilience at household and individual level can play a big role. We now recognise the importance of two weeks or more of personal emergency supplies.
• There are a range of people-focused solutions, and we should enable, support, and resource these resilience efforts at various levels of society.
• Inter-regional emergency management projects are showing great value in preparing for large-scale catastrophe and more resourcing of these would be useful.
• Caroline noted that there will be locations in NZ that are probably equipped to ride out even long disruptions or periods of isolation, where population density is low and there are many productive resources, but that urban areas will face different and more significant challenges. 
• A general improvement in risk literacy would benefit NZ’s resilience. 
• She also supported the need for the Department of the Prime Minister and Cabinet (DPMC) and similar entities to be examining these catastrophic scenarios.

David Keat

• Drawing on his many years of experience as an engineer and executive in the energy sector David stated bluntly that electricity loss means energy loss. 
• The supply of liquid fuel depends on electrical systems, and has cascading consequences, which he considered NZ possibly doesn’t fully understand. 
• Failure scenarios are exercised in the commercial energy sector regularly to learn and understand what can go wrong and how to manage it. Regular catastrophe exercises are likely to be useful for governments and communities too. 
• Furthermore, we can build resilient electrical systems, but this requires a conscious assessment of the scenarios of interest, hardening of systems, with redundancies, wide geographical distribution of key infrastructure, heterogeneous supply systems, and regulatory reform to encourage micro-grids and local electricity sharing. We need to avoid, mitigate, and ensure rapid recovery is possible. 
• Wind, solar, and geothermal-based electrical systems north of Taupo are important. Diverse systems like this, with functional islanding, will be more resilient to catastrophic events. 
• Additionally, if prevention of damage fails, then we need ways of doing things that don’t require electricity. David agreed this might include the capacity for hand pumping of liquid fuel, or generator-ready essential services. 

Dr Florian Ulrich Jehn

• Described the findings of a German Government study, which painted a grim picture of extended electrical failure. 
• Key lessons included the need for integrated decision-making, given the many firms and organisations controlling infrastructure. Additionally, stockpiling of supplies such as food shouldn’t be in single large storage facilities, because distribution is likely to be impaired. There is an important role for micro-grids, which could be repaired or restarted independently from national systems. 
• Other countries, like NZ, should produce their own reports on this kind of scenario, accounting for contextual factors, for example Germany has a system of public wells for water supply resilience, and NZ could investigate similar facilities. 
• Florian detailed the risk of large reductions in agricultural yields if conditions of electricity failure persist. This is due to reduction in production and supply of inputs such as fertiliser, irrigation, and liquid fuel for machinery. 
• Alternative and supplementary food sources do exist though, for example seaweed production can be scaled up in low tech ways. 
• The first step is to begin to think about these scenarios. 

Dr Matt Boyd

• Suggested that any government or community exercises around electricity loss should include a pathway for their findings to inform infrastructure and resource investment decisions. 
• He also noted the potential for near-urban agriculture to supply cities when transport systems have failed, but this kind of long-term solution requires land zoning changes and incentives for near-urban farms to grow optimal crops for providing dietary needs. 
• Risk governance of potential global catastrophes needs to be made explicit in NZ, potentially through a national Chief Risk Officer, unconstrained by the legislative scope limitations of various agencies and with the ability to see ‘gaps’ in the risk and resilience system. 
• The public needs a voice and should be offered options, do you want to invest in these resilience measures, or do you want these other efficiency measures?  

Audience engagement with the webinar

The audience posed questions to the Panel. These included questions around the role of finance and money post catastrophe, and how to get action on these issues. 

Matt Boyd suggested that there are several leverage points for action, which include, but are certainly not limited to:

• People making submissions on DPMC’s forthcoming Long-term Insights Briefing, which looks set to cover global risks.
• The Infrastructure Commission’s Priorities Programme, which is accepting submissions – there is scope for community exercises to feed into this kind of process. 
• NEMA looks set to iterate its Space Weather and Catastrophic Event plans, and feedback could help shape these. 
• A National Fuel Security Study has just been completed, and there is an opportunity to provide input to any potential National Fuel Security Strategy.
• There is the opportunity to progress high priority research through the new Natural Hazards and Resilience Platform – Caroline Orchiston provided some details of this. 

What did NEMA have to say on these issues?

In addition to hosting the live webinar, we also submitted some provocative questions on these issues to NEMA. These questions focused on 10 issues of NZ’s preparedness for extended power outages, addressing:

• Critical infrastructure interdependencies
• Emergency power generation and fuel distribution capabilities
• Supply chain vulnerabilities during extended outages
• Emergency communication systems
• Household preparedness requirements
• Power restoration priorities
• Space Weather Plan development and testing
• Grid resilience against various threats
• Infrastructure strategy given global supply chain challenges
• Pre-event resilience building for critical infrastructure

In response NEMA highlighted several ongoing initiatives to address extended power outage risks:

1. Infrastructure Interdependency Mapping: Collaboration with the New Zealand Lifelines Council to produce and update the National Vulnerability Assessment, providing “a unique strategic perspective of all infrastructure services as they act in combination.”
2. Emergency Communications: Multiple redundant systems including satellite communications, the National Warning System, radio communications, and Emergency Mobile Alerts.
3. Household Preparedness: Current guidance recommends “3 days or more days of supplies, a week if you can,” though NEMA acknowledges some situations may require several weeks of supplies.
4. Space Weather Planning: Development of the National Space Weather Response Plan, tested through a national-level tabletop exercise involving approximately 50 organisations.
5. Supply Chain Resilience: Active work with the Fast-Moving Consumer Goods Sector and development of “a fuel database, including the status of generators, to identify vulnerabilities.”

Examples of some direct responses from NEMA: 

“NEMA collaborates with, and invests in, the New Zealand Lifelines Council to produce and update the National Vulnerability Assessment. The report provides a unique strategic perspective of all infrastructure services as they act in combination to support the wellbeing of New Zealanders.”

“NEMA’s Catastrophic Planning Programme prioritised the logistics workstreams and established a National Logistics Working Group to better define this critical risk, educate and influence, and progress response arrangements.”

“From NEMA’s perspective, space weather preparations have been useful for collaboratively developing procedures and coordination arrangements, training and exercising, and broadening our understanding of cascading disruptions from power outages. The consequences of which will be managed by the hazard-agnostic Catastrophic Event Handbook, which considers severe and sustained electricity disruption.”

Who comprised the audience for this webinar?

Attendees were mostly based in NZ, but also Australia, the US, Ireland, and elsewhere. The audience consisted of individuals associated with the following: 

• Government departments: including the Ministry of Primary Industries (MPI), Maritime NZ, Fire and Emergency NZ (FENZ), NEMA, and DPMC
• Regional councils
• Public health sector
• Consulting firms
• Researchers
• Individuals

Where can I find the reports/sources mentioned in this webinar?

• Catastrophic electricity loss blog by Matt Boyd
• Dr Simon Blouin interview video on US food supply chain
• Alana Weir et al.’s (2024) paper on infrastructure interdependencies in Taranaki
• German government study on the consequences of catastrophic electricity loss
• Florian Ulrich Jehn et al.’s research on the fragile state of industrial agriculture, and on seaweed for food
• Matt Boyd et al.’s research on near-urban agriculture for resilience and mitigating imported fuel dependency
• David Korowicz’s paper on global financial systemic collapse
• Alpine Fault 8 webpage
• NZ’s new 2025 National Fuel Security Study
• NEMA Catastrophic Event Handbook
• NEMA Space Weather Response Plan
• DPMC Building Resilience to Hazards long-term insights briefing topic

How can I continue to engage with this work?

Reach out to us: https://www.islandfutures.earth/

This is a link post for my blog about Professor Alan Robock’s 10 Feb 2025 talk at Victoria University of Wellington.

Professor Robock of Rutgers University has studied the climate effects of nuclear war for 40 years.

I wrote the blog for Islands for the Future of Humanity and the post includes an animation of nuclear winter, and link to Prof Robock’s 2022 TED talk.