While I was at WesterCon last weekend I did a number of panel discussions. And one of the more interesting ones was on the topic of anthropogenic world-ending events.
So here are my post-convention thoughts, and a question.
News flash: nuclear weapons are not likely to cause human extinction. There's certainly no doubt that a large-scale nuclear war with the arsenals that existed on both sides at the peak of the cold war (say, circa 1988) would have been a thing of horror and disastrous for civilization on a global scale; even nations untouched by bombs, fallout, or climactic changes triggered by continent-sized fires would have suffered grievous damage from disrupted supply chains, loss of access to resources, and the total collapse of the global trade system. But it's hard to see a path leading to a total death toll in excess of 90% of the human population, or rendering it impossible to retain a 19th century engineering culture (steam traction, electricity production from hydropower or local fossil fuels, iron and steel production) at least in isolated locations after the event.
About the only path through which warfare could lead to total human extinction would have been through the use of modern biological warfare technology. For example, it's possible to modify poxviruses such as mousepox (or smallpox) by inserting the gene for IL-4 and turn them into a ferociously lethal airborn plague that triggers a cytokine storm in its victims, with close to 100% lethality. This was once, and recently, the domain of high-level vaccine research and biowarfare labs; but the tools that enabled the genomics revolution also permit the relatively easy development of such a "biobomb in the basement" weapon—the only thing going in our favour is that nobody would be crazy enough to deploy a species-extinction weapon on their own kind. Or would they?. (See also; apocalyptic eschatology may be one of those things we don't get to survive the 21st century and keep.)
But then we get to other doomsday scenarios. I'm going to take anthropogenic climate change as a given at this point and bring up the Green Sky hypothesis as a possible "oops, didn't mean to do that" scenario for the end of the world. It's a model for how the Cenomanian-turonian extinction happened, roughly 110 MYa ago. As temperatures rise, the deeper waters of the ocean become oxygen-depleted. Gradually the upper border of the anoxic level rises until it's close to the top; below it, anaerobic bacteria and archaea survive by switching to the sulfate reduction cycle. This is something that happens in stagnant ponds; but there's some evidence that entire oceans are overwhelmed by this sort of bacterial bloom when conditions are right, leading to catastrophic consequences for life. Sulfate-cycle metabolism releases hydrogen sulfide, H2S, rather than carbon dioxide, and H2S is rather more toxic than hydrogen cyanide; so as the atmosphere is depleted of its breathable oxygen by the shutdown of oceanic photosynthesis, the bacterial bloom produces poison in its place. For a while the stagnant bottom layers will be stable ... but eventually the H2S-saturated waters will begin to circulate, releasing the dissolved gas, in a disastrous outburst similar to the outgassing of Lake Nyos in Cameroon, where in 1986 the volcanic lake emitted a large burp of carbon dioxide, asphyxiating nearly 2000 people in villages nearby. (As with the anaerobic ocean hypothesis, Lake Nyos was gassified from the bottom up; like a bottle of soda water, it finally out-gassed and bubbled up when conditions became right for a limnic eruption.
The reason for focussing on the oceanic disaster is its scale. To get to it, we first need to produce a Canfield ocean—a stagnant one, where the Thermohaline circulation ceases and the bottom and upper reaches of the oceans are no longer mixing. In the past, this usually arose as a side-effect of continental drift. But we're currently seeing unprecedented discharges of melting water from the Greenland and Antarctic ice caps which, being extremely cold, have a significant impact on the existing deep oceanic currents—just as we're subjecting large chunks of the upper oceanic biosphere to unprecedented stress and over-fishing induced extinction is causing a large-scale repopulation event or two (for truly, right now the future seems to belong to the tentacled ones!).
Anyway. Forget nuclear weapons: they're nothing like destructive enough! And it's doubtful that a small group of fanatics wouldd be able to deflect a big enough asteroid to kill everyone, with any technology that's on the horizon—certainly not without providing enough lead time for national space agencies to intervene. (As most asteroids seem to be gravel piles rather than solid objects, and tend to explode at high altitude under the stress of entering the atmosphere, "dropping rocks" from orbit is a somewhat problematic doomsday weapon.)
Biological weapons like weaponized IL-4 augmented poxviruses are a possible doomsday scenario. And we could do it to ourselves accidentally, albeit over a thousand year time scale, if we fuck up the thermohaline circulation beyond all hope of remediation—or if someone utterly crazy figures out a shortcut to cause a sulfide bacteria bloom in the deep ocean.
Short of handing out lethal technological magic wands (and I include Drexler-level nanotech in this category: genetically engineered bacteria are quite bad enough!), what other paths to we-did-it-to-ourselves human extinction can you see that might be viable within the next century?