Global Catastrophic Risks
When the world stops turning
Description
In 2008, the philosopher Nick Bostrom, who ran the Future of Humanity Institute at Oxford, brought together a couple of dozen specialists — astrophysicists, epidemiologists, economists, weapons experts — and asked them a question most people spend their lives avoiding. Not "how might I die?" but "how might we all die?" The result was a thick, sober volume called Global Catastrophic Risks, edited with the philosopher Milan Ćirković. It was not a doomsday pamphlet and not science fiction. It was an attempt to do something the disparate fields of danger had never done together: sit at the same table and compare notes.
The framing was deliberately wide. A global catastrophic risk, in the book's definition, is any event with the potential to inflict death and destruction on a planetary scale — enough to derail civilization, or in the worst cases, to end it. Some of these threats are ancient. Plagues have emptied cities before; wars and murderous ideologies have darkened entire centuries. What Bostrom wanted to add was the unsettling news that the twentieth and twenty-first centuries have introduced a new category of danger — one built by human hands, and growing faster than our wisdom about how to handle it.
The book's quiet provocation is that we are systematically bad at thinking about this. We obsess over risks that frighten us and shrug at the ones that could actually finish us. A single plane crash dominates the news; a slow-building threat to the species barely registers. Bostrom's collaborators set out to correct the imbalance — not to terrify, but to get the arithmetic of survival right.
The question we’re asking : What separates a bad day for humanity from a threat to its very existence, and why do we keep misjudging the difference?What we’ll see : How a philosopher gathered the world's dangers into one room and asked what it would take for the world to stop turning.
Table of contents
01Chapter 1 — Why the biggest risks are the ones we underweight
Bostrom opens with a distinction that turns out to do most of the work in the book. Not all catastrophes are equal. There is the local disaster, terrible but bounded. There is the global catastrophe, which knocks civilization off its feet. And then there is the category he cares about most: the existential risk, one that would either wipe out humanity or permanently lock it into a diminished state from which it could never recover. The difference between a catastrophe that kills most people and one that kills everyone is, in his accounting, not a matter of degree. It is a difference in kind, because the second closes the door forever.
This is where our instincts fail us. We are built by evolution and habit to react to threats we can picture — a fire, a predator, a car swerving into our lane. We are poorly equipped to feel the weight of a risk that is enormous but abstract, diffuse, and slow. Bostrom points to the cognitive biases that distort the picture: we discount the future heavily, we assume tomorrow will resemble yesterday, and we treat events without precedent as effectively impossible. Nothing on this scale has happened in living memory, so the mind quietly files it under "won't happen."
02Chapter 2 — The old killers wearing new faces
Before the book reaches its newer dangers, it pays respect to the ancient ones, because they set the baseline. Disease is the oldest global catastrophe we know. The Black Death of the fourteenth century killed something on the order of a third of Europe's population; the 1918 influenza pandemic killed tens of millions in a single year, more than the war that preceded it. These were not metaphors for catastrophe. They were the real thing, and the contributors treat them as evidence that a species-scale plague is not a hypothetical but a recurrence.
What has changed is the terrain the pathogens travel across. A crowded, densely connected planet — global air travel, megacities, industrial agriculture pressing against wild habitats — is close to an ideal medium for a fast-moving disease. A virus that would once have burned out in a single valley can now circle the globe before anyone names it. The book, written well over a decade before COVID-19, reads in retrospect as a patient warning that the systems designed to move goods and people are equally efficient at moving contagion.
03Chapter 3 — The dangers we are building ourselves
The heart of the book is its argument that the most serious risks are increasingly of our own manufacture. Nature has its catalog of horrors, but they are, in the long run, roughly constant — the asteroid is no more likely this century than last. The human-made risks are the ones on a steepening curve, rising in step with our technological reach. This is the shift Bostrom most wants readers to absorb: for the first time, the species has acquired the power to end itself.
Nuclear weapons were the first proof. Within a few years of splitting the atom, humanity built enough warheads to devastate every major city on Earth, and the specter of nuclear winter — soot from burning cities blotting out the sun and collapsing agriculture worldwide — turned a regional exchange into a plausible global one. The arsenals shrank after the Cold War but never disappeared, and the knowledge cannot be unlearned. The contributors treat this not as history but as a standing condition.
04Chapter 4 — The one experiment we don't get to rerun
Step back from the individual threats and the book's deeper argument comes into focus. Bostrom's real subject is not asteroids or algorithms; it is the peculiar structure of a risk you can only run once. For most dangers, humanity learns by surviving them. We suffer a disaster, bury the dead, study what went wrong, and build defenses for next time. That entire method — trial, error, correction — depends on there being a next time. Existential risk removes the second attempt. There is no lesson to be learned from an extinction, because no one is left to learn it.
This is why he treats the survival of humanity as a category apart, not merely one large stake among others. He frames the potential of a long human future — the sheer scale of lives, thought, and value that could still lie ahead — as something so vast that even a small reduction in its probability of loss outweighs almost any nearer concern. It is a coldly logical case, and it lands as an uncomfortable one, because it asks us to care intensely about people who do not yet exist and a threat that has never occurred.
05Conclusion
The book ends where it began, with a plea for proportion. Bostrom is not forecasting doom; he is objecting to the way we file the largest questions under "not now." The dangers he catalogs — plague and war, asteroids and warheads, engineered pathogens and thinking machines — do not share a cause, but they share a shape. Each could, under the wrong conditions, remove the future entirely. And each is easier to ignore than to face, precisely because it has not yet arrived.