Last year, cosmologists working on the Dark Energy Spectroscopic Instrument (DESI) reported hints that the mysterious dark energy thought to be driving the expansion of the universe may be weakening over time. If these startling findings prove correct, then dark energy cannot be a cosmological constant – a fixed term in our equations that represents the energy of empty space – after all. When this bombshell hit, most of the buzz focused on what that means for the standard model of cosmology, known as lambda-CDM, our best attempt to explain the evolution of universe.

If the results firm up, we may finally have the clues required to build a better theory. Already, researchers are busy trying to rethink dark energy, and possibly dark matter and gravity, too.

But if the strength of dark energy really does diminish over cosmic time, the implications could run far wider and deeper. Wider, in the sense that it could provide fresh impetus for proponents of alternative cosmologies that change our understanding of the fate of the universe. And deeper, because it might even be telling us something profound about the deepest structure of space-time. “There certainly are very, very interesting possibilities for changing a lot of physics,” says Eric Linder, a physicist and cosmologist at the University of California, Berkeley.

According to lambda-CDM, in its first moments, the universe underwent a split-second spell of exponential expansion. Known as inflation, this explanation seems to provide a reason for why the universe is so smooth, flat and homogenous on its largest scales. But inflation has its critics, most prominent among them Paul Steinhardt, a physicist at Princeton University. “Inflation doesn’t work,” he says bluntly, adding that it requires unlikely initial conditions, is too flexible and leads to a multiverse scenario that many find implausible.

A cyclic universe

Steinhardt has long made the case for an alternative hypothesis known as the cyclic universe, in which the universe endlessly expands, contracts and bounces back. To make such models work, however, dark energy has to evolve.

“It must be some kind of decaying dark energy that stops accelerating the expansion of the universe, starts decelerating it and then eventually causes contraction, leading to a bounce and a new cycle,” says Steinhardt. The first part of that at least – that the acceleration of expansion is slowing – is precisely what we seem to be seeing with the DESI data.

This isn’t to say that the DESI results provide evidence for cyclic cosmologies. We may yet find systemic errors in the measurements and analysis, and it is entirely possible that dark energy weakens without ever producing a contraction or a bounce. If hints of decaying dark energy do firm up, however, that would lend credence to Steinhardt’s long-standing argument. “I tend to be very conservative and very patient,” he says. “What I would say, however, is that now the game is afoot.”

The same could be said for another controversial idea that has received a shot in the arm from the DESI results. Broadly speaking, string theory says that everything is ultimately made of vanishingly tiny strings, compactified into hidden extra dimensions, whose vibrations manifest as the various particles and forces we discern. It rose to prominence in the 1980s because it seemed to offer a route towards a theory of quantum gravity, reconciling quantum theory and general relativity into what some call a theory of everything.

But string theorists have long struggled to construct models of the universe with a small, positive cosmological constant. In a series of papers published in 2018 and 2019, theoretical physicist Cumrun Vafa at Harvard University and his colleagues built on a set of proposals known as the Swampland conjectures, which aim to distinguish theories of particles, forces and space-time that can arise from a consistent theory of quantum gravity from those that cannot. Using this framework, they suggested that dark energy can’t be a cosmological constant but must instead be a kind of field – similar to the one thought to have driven inflation – whose energy changes over time.

At the time, such a proposal conflicted with the long-held belief that dark energy stayed the same over cosmic time. “People were saying: ‘String theory is ruled out because dark energy is a constant,’” says Vafa.

Hidden dimensions

But he and his colleagues persisted. In 2022, they proposed a model in which space-time has a large hidden extra dimension, possibly as large as a micrometre, the size of which gradually changes over cosmic time. As the geometry of this dimension changes, the amount of energy in the universe we observe changes, too. The researchers argued that this would show up as a dark energy that slowly weakens. “There’s nothing exotic [here] from the perspective of string theory,” says Vafa. “The extra dimension is changing, and both dark energy and dark matter are responding to it.”

It is easy to see why the DESI results are intriguing for string theorists: Vafa and his colleagues had predicted dark energy should be gradually weakening, and now that seems to be what we are seeing. Indeed, when Vafa and his team analysed the DESI data combined with other cosmological datasets in 2025, they found their model fits far better than lambda-CDM and about as well as the best conventional models that allow dark energy to evolve. The difference here, he says, is that their model includes a physical explanation for what we are seeing. “This is why I’m so excited,” he says. “It’s very satisfying.”

To be clear, the DESI results don’t offer concrete evidence for string theory. For starters, the extent to which they prefer evolving dark energy over a cosmological constant still depends on which other cosmological datasets they are combined with. What’s more, non-stringy models that don’t invoke hidden extra dimensions fit the existing data equally well.

But if we assume for a moment that the DESI data holds up and the statistical significance grows to discovery level, evidence of weakening would not only remove an empirical obstacle to string theory, it would also weaken the argument that string theory doesn’t offer testable predictions. “We came up with this model years ago,” says Vafa. “Now they’re observing it, and it looks exactly like what we expected.”

To make good on the notion that this might provide observational evidence in support of string theory, however, theorists like Vafa would have to build a sharper model that makes more precise predictions, distinct from non-stringy alternatives, and show that it fits the full range of cosmological data better than other options. Intriguingly, the framework already hints at additional testable signatures, including departures from the standard picture of how dark matter evolves and deviations from general relativity at micrometre scales.

Some cosmologists are unconvinced the DESI results have any bearing on fundamental physics at all, even if they do firm up. “Dark energy operates on certain scales, and that is what we can talk about,” says Pedro Ferreira, a cosmologist and astrophysicist at the University of Oxford. “[When it comes to] what happens at quantum levels, I don’t think we can go there.”

But others are open to the possibility that these hints could have ripples well beyond cosmology, not least because they might give us a first glimpse into the deep quantum structure of space-time. “What Cumrun Vafa has come up with, it’s the most interesting thing I’ve seen,” says Mike Turner, a cosmologist at the University of Chicago in Illinois. “This is where cosmology and particle physics come together. We’re digging at really fundamental things, so the knock-on effects can be tremendous.”