Open notebook, biggest-mystery edition. We have a number for how much of it there is and essentially no idea what it is. Here's the state of play and where I'd put my chips.

The discovery nobody ordered

In the 1990s two rival teams set out to measure how fast the universe's expansion was slowing down. Everyone expected slowing — gravity from all the matter in the cosmos should act as a brake. They used Type Ia supernovae, exploding white dwarfs that all detonate at nearly the same true brightness, as "standard candles" — the top rungs of the cosmic distance ladder: compare how bright one looks to how bright it truly is, and you have its distance.

The distant supernovae came back too faint — farther away than a coasting, decelerating universe allowed. The expansion isn't braking. It is speeding up. Both teams found it; the result took the 2011 Nobel Prize.

Whatever is stepping on the gas, we call dark energy, and the bookkeeping says it is most of everything:

• ~68% dark energy
• ~27% dark matter
• ~5% ordinary matter — every star, planet, and person

The leading guess

The simplest candidate is a cosmological constant: a fixed energy belonging to empty space itself. Counterintuitively, the vacuum can carry energy, and unlike matter it doesn't thin out as space grows — so as the universe enlarges, this push only wins by more, and the expansion accelerates. Einstein once put such a term in his equations, called it a blunder, and removed it. It came back.

The most embarrassing number in physics

Here is the bruise. We can try to calculate the energy of empty space from quantum theory, summing the restless jitter of the vacuum. The answer disagrees with the measured value by a factor of roughly:

10¹²⁰ — a 1 followed by 120 zeros. The worst quantitative prediction in the history of science.

Either enormous contributions cancel almost — but not quite — perfectly, or we are profoundly misunderstanding what vacuum energy is. Nobody has a convincing reason for the small number we actually observe.

Where I currently lean

Provisionally, toward the plain cosmological constant. It fits the data with a single number, and the alternatives — a slowly shifting field ("quintessence"), or yet another rewrite of gravity — don't yet explain anything the constant can't. But "it fits" sits uneasily next to a 120-order-of-magnitude hole in the theory behind it.

What would change my mind: evidence that dark energy changes over time rather than holding constant. Recent large galaxy surveys have hinted at exactly that — faint, unconfirmed, the kind of signal that has evaporated before. If it holds up, the cosmological constant is dead and quintessence is back in business.

Either way, dark energy probably writes the ending. If the push persists, the universe expands and cools forever toward a cold, dilute, maximum-entropy state — which is really a story about why time runs one way, told on the largest possible stage.