To the ancients, exploding stars were bad news. To astronomer Adam Riess, poring over data from a telescope in Chile, it looked like supernovas were still cursed. He and his colleagues were measuring the brightness and distance of supernovas in order to figure out the little matter of whether the universe would end in fire or in ice. Would it halt its expansion and collapse back on itself in a gnab gib (that's the reverse of the big bang, and passes for humor among astronomers) or expand forever, its light and warmth fading into eternal cold and darkness? But when Riess calculated how much mass the universe must have in order to account for the supernova data, he got a negative number. A nonsense answer—negative mass—"was the first hint that something was wrong," he says.

Ten years on, astronomers know what was wrong: their assumption that the universe is decelerating. By measuring supernovas, they hoped to discover whether the deceleration was large, portending a fiery fate, or small, implying an icy one. Either way, they assumed, there was some deceleration. The absurdity of negative mass blew that assumption out of the sky. The universe is not decelerating; it is accelerating. A mysterious energy is pushing apart galaxies and space itself, like a crazed toddler with bulging cheeks and infinite lung capacity blowing up the cosmic balloon faster and faster. Ice is our fate. This "dark energy" is so pervasive that it makes up 76 percent of all the mass energy in the universe. Atoms in the likes of us make up a piddling 4 percent.

As astronomers mark the 10th anniversary of the discovery of dark energy this fall, they have a new appreciation for how much Copernican humility remains in order. For dark energy revealed that the matter in planets, stars, ourselves and everything we hold dear is a cosmic afterthought, the fallen scraps from the main fabric of the cosmos. And efforts to explain what dark energy is—where it comes from, what generates it—are showing that the craziest ideas in astronomy pale next to what nature can concoct.

The best bet for the source of dark energy is the sea of subatomic particles that pop into and out of existence in "empty" space. In 1917, Einstein proposed something like this, only to reject it as his greatest blunder. (The resurrection of this "cosmological constant" led the mother of one astronomer, Robert Kirshner of Harvard, involved in the discovery of dark energy, to ask her son, "So, do you think you're smarter than Einstein?") But Einstein might have been right the first time. The evanescent particles give space a stretchiness, pushing it apart. There's only one problem: when you calculate how much energy the roiling sea of particles would create, the answer is 100,000 [insert 51 more zeroes] times more than the dark energy. "That's a large discrepancy even in astronomy," Mario Livio of the Hubble space telescope institute said at a workshop on dark energy last week. If Einstein's cosmological constant truly is the source of dark energy, then something else cancels out all but a smidgen of the energy from the popping particles. That something else is anyone's guess. Worse, the precision of the required cancellation—erase the ink on every magazine ever printed except for exactly one comma, here—strains credulity.

Unless "the" universe is actually only one of many universes. Cosmologists are seriously entertaining that possibility. In the big bang that started our universe, little bits of space might have pinched off, said physicist Steven Weinberg of the University of Texas. Each pocket universe could have different features. In one—ours—the improbable cancellation of most of the cosmological constant would have occurred, leaving just enough to explain the dark energy. As for why we happen to be living in that improbable universe, that's easy: only this feature allows galaxies to form and life to emerge.

If the explanation of dark energy does lie in a multiplicity of universes, more humility will be in order. It will mean that there is no first-principles explanation for why the cosmological constant is just the right size to account for dark energy. In a multitude of universes, some have some properties and others have others, but there is no deeper explanation.