The apocalyptic posits the end of this world as the end of the world, the only possible world; what lies beyond the apocalyptic horizon is not another world, but the end of all possible worlds, of all worlds as worlds.

The actual extinction of humanity, indeed of all life on this planet, is both a possibility (in that it possibly might happen within my lifetime, or within a generation or so; indeed, it possibly might happen this evening - we certainly have the megatonnage to bring it about) and a certainty (in that one day, even if meteors or microbes don’t do for us all first, the sun itself is certain to explode).

Even this does not mean the end of all possible worlds, although it may well mean the end of all worlds in which humanity as such will be able to appear. Badiou draws on the “fossil argument” of Quentin Meillassoux to indicate that the world of the archi-fossil was both a world, and a world in which the human did not appear, thereby establishing the independence of worlds in general from human habitation. (Dinosaurs, strangely, are appearing in our world all the time; my living room floor has frequently been littered with countless plastic apatosauruses and velociraptors).

The end of the universe is rather less certain, not because the fate of the universe in hangs in any sort of a balance, but because we do not have a definite concept of the universe that would allow us to plot with complete certainty its ends and its beginnings. The universe may just be not the kind of thing that can have either an end or a beginning as such. Time and space might just fuzz causelessly and probabilistically into being. Furthermore, the universe itself might go on existing indefinitely after everything in it has succumbed to entropy. Whatever happens to the universe in the (very) long run is in any case unlikely to be our problem.