We should count ourselves fortunate. Today, we live in what astronomers Greg Laughlin and Fred Adams called, in their 1999 book The Five Ages of the Universe, the Stelliferous Era. It’s a period in which the Universe is ablaze with the light of stars collected into galaxies. More stars are being born all the time, even as other stars perish, giving their heavy elements back to the cosmos with which it can build new stars, planets and life. Large galaxies can contain hundreds of billions of stars, and there’s an estimated two trillion galaxies in the visible universe alone.

Nothing lasts forever though. Eventually, in a 100 trillion or so years’ time, the Universe will have exhausted all of its star-forming material. The last star will be born, and from thereon the Universe will face a slow death as gradually each and every star burns out. The Universe will go dark – well it would, if a darkness had not already descended upon it. Long before the last star is born, life in the Universe will already be in deep trouble. And life’s enemy will be dark energy.

The energy of the Big Bang, 13.77 billion years ago, spurred the Universe into action, pushing it to expand. Astronomers figured that this expansion would have begun running out of oomph by now, and should have started slowing down. When, in the late 1990s, astronomers found that the expansion was speeding up instead, the shock to cosmology was palpable. A mysterious force called dark energy was ascribed to this accelerated expansion. Nobody knows what dark energy is exactly, but the name seems to fit.

Our best model describes it as the ‘cosmological constant’. Back before we knew the Universe was expanding, Einstein’s equations implied that the Universe should be changing with time, so he dreamed up the concept of the cosmological constant that would counteract gravity and maintain the cosmos in a steady state. When Edwin Hubble discovered in 1929 that the Universe is expanding, Einstein jettisoned his cosmological constant, calling it is his “biggest blunder.”

With the discovery of dark energy, scientists revisited Einstein’s old idea, twisted it around a bit and applied it to the accelerated expansion. It’s kind of a latent energy associated with space, though the details are all hand-wavy right now. The point is, the more that space expands the more space there is, and the more space there is the more dark energy there is, which expands space even more, and so on and so forth. In a Universe governed by the cosmological constant, expansion of space ultimately becomes exponential. 

In a finite Universe, light has only had a fixed amount of time to travel across space. There are galaxies so far away that their light just hasn’t had time to reach us yet. This marks the cosmic horizon, which is currently set about 46.4 billion light years away in real terms and is still growing with cosmic expansion.

Yet, while time is working to enlarge the horizon and reveal more distant galaxies, dark energy is working against it, dragging those galaxies, and every other galaxy that is not gravitationally bound to our Milky Way Galaxy, away from us. Oxford’s Toby Ord, in his 2021 paper The Edges of Our Universe, calculated that there’s a practical maximum size to the cosmic horizon. If a photon was released close in time to the Big Bang, the farthest galaxy it could reach is 62.9 billion light years away. After that, dark energy would have pulled any more distant galaxies so far away that the photon would not be able to reach them. For photons setting off now, that distance is even shorter – for a photon setting off from Earth right now, the most distant galaxy it could reach before dark energy pulls everything too far apart is 16.5 billion light years away. Ord describes a spherical volume of this radius, centered on Earth, as the ‘affectable universe’ – everything beyond cannot be causally affected by events happening in the cosmos today.

In about 100–150 billion years time every galaxy to which we are not strongly gravitationally bound will have been pulled away over the horizon. All that will be left will be our Local Group of galaxies – the Milky Way and Andromeda would have long since merged with each other and perhaps also with the Triangulum Galaxy and the few dozen smaller dwarf galaxies that are our neighbors. Beyond this super-galaxy the observable Universe, once illuminated by trillions of galaxies, will be dark. Everything else will have receded out of sight.

It’s against this backdrop of runaway cosmic expansion that life must find a way to survive.

Life needs energy, and stars are the Universe’s power stations. Without the stars, life will struggle to survive into the deep future, but dark energy is intent on stealing the stars from us. It’s incumbent upon life to take the stars back.

This audacious idea forms the basis of a 2017 paper from Fermilab astrophysicist Dan Hooper, brilliantly titled Life Versus Dark Energy.

“There will be a point somewhere between roughly this era and 100–150 billion years into the future where a civilization will look up into the sky and notice that everything is moving away from us, and they are going to have to think about what they are going to do about it,” Hooper tells Supercluster. “And if they have the technological means to do it, then they will send out their emissaries to try and grab everything they can before it disappears.”

It would be the great star heist. Billions, perhaps trillions of stars could be stolen from their galaxies and moved physically through space back to the core of civilisation. There, they would form a dense concentration of stars that can not only power life into the deepest future, but whose combined gravity can also rebuff the divisive influence of dark energy. As the cosmos around this core of civilization is ripped away from us to disappear into the night, a line will have been drawn. The super-galaxy that our Local Group merges into will become a new haven, a ‘Fortress Galaxy’ whose gravitational ties will be like armor, warding off the expansion.

“The more you gravitationally bind it as tightly as possible, the more that matter can resist the effects of dark energy,” says Hooper.

So, first things first. How exactly do you move a star?

Hooper admits to being vague about how this could be accomplished.

“It’s difficult to imagine what technology would likely look like for these hyper technologically advanced civilizations,” he admits. “What I think we can say is that whatever is possible within the confines of the laws of physics, I’m going to imagine that eventually we’ll find a way to do that.”

“Does a star produce enough energy to get that star moving?” asks Hooper rhetorically. “In some cases the answer is yes, you can outrun the horizon with the energy that a star produces.”

There’s a few plausible ways to move stars, and they all belong to a class of technology known as ‘stellar engines’. One is a Shkadov thruster, named after the Soviet scientist Leonid Shkadov who first came up with the idea in 1987. He proposed that a civilization could build a huge, hemispherical screen that sits to one side of the star. This screen would be enormous, a megastructure as wide on its long axis as the star itself, and it would reflect stellar radiation back towards the star only on that side, so that stellar radiation only escapes into space on the other side of the star. This would result in an asymmetrical force that pushes the star in the direction of the screen. In other words, the star becomes a thruster. And because the screen is perfectly balanced between the stellar radiation trying to push it away and gravity holding onto it, it can remain stable over long time scales.

The only trouble is that Shkadov thrusters are slow. After a billion years, a Shkadov thruster would have moved a Sun-like star about 34,000 light years, which is fine if stars are just being rearranged in a galaxy, but too sluggish if we’re trying to bring them in from tens, perhaps hundreds of millions of light years away. So we need to find a faster way.

That faster way could be a ‘star tug’, which was devised by Alexander Svoronos of Yale University in 2020. A star tug consists of a megastructure that mines material from the star, converting it into propellant to produce thrust. Because the megastructure and star are gravitationally bound, it ‘tugs’ the star along with it. Svoronos reckons that, in ideal conditions, a Sun-like star can be accelerated to about a quarter of the speed of light within a timescale of a few million years. 

Hooper proposes that the best stars with which to build the Fortress Galaxy are those with masses between that of the Sun and a fifth of the Sun’s mass. For one thing, these smaller stars are commonplace; surveys show that three-quarters of all stars are small red dwarfs, so there will be many more of this type of star available. Their lower mass makes them, in principle, easier to move. And lower mass stars live longer lives; red dwarfs with less than a quarter of our Sun’s mass can be expected to survive for up to 100 trillion years, thanks to how they are able to access their full store of hydrogen, unlike more massive stars, and burn it more slowly without having to resist gravity trying to collapse them as strongly as in massive stars, which in some cases only live a few million years before going supernova.

Yet a star tug eats away at a star’s mass, consuming it for fuel, so this method might not work for low-mass stars as they don’t have enough mass to power a long intergalactic journey. So we’d need another plan for them. A star tug could work for higher-mass stars though. The stellar engine would whittle them down, perhaps to the point of them having the mass of a red dwarf. Then we wouldn’t just be transporting stars, we’d be engineering them too.

Maybe a technologically advanced civilization would also want to build their own new stars, using a Bussard ramscoop to sweep up hydrogen, concentrating it into a cloud that can be forced to collapse and give birth to new stars that are optimized for maximum energy extraction.

So far, we’ve talked vaguely about a future civilization. Perhaps that will be us, if we can survive our dangerous youth. Or maybe it will be someone else, and just maybe, they have already begun the work. Hooper suggests that if there is an alien civilization already out there stealing the stars, we might find spectroscopic evidence for it in the form of galaxies that have a spectrum where the redder light from the cooler, less massive stars is absent, leaving only more massive starlight in the spectrum. Since massive stars don’t have long lifetimes on cosmological scales, they may have gone too. There have actually been cases of galaxies that have been found to be almost devoid of stars and made up almost entirely of dark matter and gas. Astronomers call them ‘low surface brightness’ galaxies. One standout case is a galaxy catalogued as J0613+52, which is located 270 million light years away and which has no visible stars – it is only detectable by the radio emission from the hydrogen gas that fills the galaxy. The likelihood is that J0613+52 and other low surface brightness galaxies exist because of some quirk of physics preventing them from forming stars – J0613+52 contains huge amounts of gas that has gone unused. However, a galaxy that has had its stars stolen would look like a low surface brightness galaxy. 

If we are alone in the Universe, then the objective is clear: we build the Fortress Galaxy for ourselves. If we’re not alone in the Universe, things become more complicated because different civilizations, even if they exist in different galaxies, would be competing over the same stars. One could imagine interstellar, or even intergalactic, wars as civilizations fight over the stars themselves. Any civilizations that can’t keep up with the mad star rush would risk falling by the wayside.

One would hope that calmer, more altruistic minds would prevail. Thus far, we have found no evidence of extraterrestrial life. Alien civilizations might be incredibly rare, less than one per galaxy. Once dark energy carries them away from us, beyond the cosmic horizon, then they’ll be gone for good and we would all be alone, at the mercy of a cold and dark cosmos. It would be better for civilizations to team up, to preserve what life they find in the cosmos across many galaxies, and bring ourselves all together at the hub of civilization that would be the Fortress Galaxy. It would be the biggest act of altruism, cooperation and engineering that the Universe would ever see, a project in which life itself would seek to defend itself against the unyielding expansion.

While there’s still plenty of time left on human timescales, waiting even a billion or two years to begin this endeavour might by too late, especially if someone else takes the initiative before us. 

“In the best case scenario we start doing this right away and we can get everything out to 50 mega-parsecs (about 163 million light years) transported, cosmologically speaking, to our local neighborhood,” says Hooper. “But the longer we wait, the more of this stuff will be unretrievable.”

For example, in 100 billion years the effective cosmic horizon will be where galaxies that are just 15 mega-parsecs (about 50 million light years) from us are now.

We could take advantage of Toby Ord’s ‘affectable universe’. Why spend time going out to fetch the stars back, if there is other life out there who might want to join us in the Fortress Galaxy endeavour and bring the stars to us? We could signal such vast distances by imprinting messages onto supernova light, encoding a call to arms in absorption lines in the spectra of supernovae. While being at the far edge of this affectable universe, 16.5 billion light years away, would not allow enough time for other beings to reach us, we could still call out to a wider area than if we were having to physically travel out there and bring the stars back ourselves.

A possible twist in the tail remains. Perhaps the Universe itself will give us more time.

Earlier in the article we talked about how dark energy is thought to be the cosmological constant, wherein the strength of dark energy remains the same per unit volume of space. However, there is an alternative theory that suggests dark energy could change strength over time. Scientists called this kind of dark energy ‘quintessence’, and it would be an energy field spanning the Universe with the potential to vary in strength with time and space. The specifics are even more hand-wavy than the cosmological constant, but the basic idea is that dark energy could slow down in the future. 

“The faster the Universe expands, the more stuff falls beyond the horizon,” says Hooper. “So if you want to give future civilizations the possibility to consume and use as much free energy as possible, which is essential to anything that I call life, then you want the Universe to not expand all that fast.”

There is a dark flip-side to this. Instead of slowing down, quintessence could speed dark energy up, and we’d lose the other galaxies faster. Worst case scenario: the expansion would tear the fabric of space-time apart in a ‘Big Rip’. Not even our Fortress Galaxy could survive that.

Most scientists would say that the bulk of the evidence so far is in favor of dark energy being the cosmological constant. However, initial results from the Dark Energy Spectroscopic Instrument (DESI) on the Mayall four-meter telescope at Kitt Peak National Observatory hint at some tentative evidence that the strength of dark energy could have been different in the past. Hooper says he and the scientific community remain skeptical at this stage, but they are “keeping their eye on it” until further observations by DESI are analyzed.

Let’s assume the Big Rip doesn’t happen. By 150 billion years into the future all the other galaxies beyond our Fortress Galaxy will have disappeared over the cosmic horizon. With just the stars that have been brought into the Fortress Galaxy, it could survive about 100 trillion years. If the inhabitants of the Fortress Galaxy have also stashed away lots of molecular hydrogen gas, the building material of stars, they could assemble new stars to replace those that have expired. How long they could keep this up for would depend on how much gas they have, but let’s say they have enough to replace every star in the Fortress Galaxy just once. That would mean that the Fortress Galaxy could survive 200 trillion years.

After that, things get really iffy. There will be stellar remnants – brown dwarfs, white dwarfs, neutron stars and black holes – to huddle around. Over unimaginably long timescales, the remnants of the Fortress Galaxy will begin to disperse, black holes would evaporate, perhaps even protons and atomic structure will decay, and the heat death will be the final nail in the coffin as the temperature of the Universe equalizes according to the Second Law of thermodynamics. The timescales involved would be beyond comprehension – for example, what the heck is a novemvigintillion? Apparently, it is 10^91 to 10^92 years, the timescale for a supermassive black hole to evaporate. How life could survive any of this is a problem for our deep-future descendants.

For now, we can barely imagine the advanced technology required to steal the stars, but that doesn’t mean we cant be thinking about it.

“The first thing to do is to try and learn all the laws of physics we can,” says Hooper. “In the long run, there’s no better investment for the future of humanity than investment in fundamental science.”

Who knows, through our studies of dark energy or the evolution of stars, or advanced engineering in space, a pathway to the next 100 billion years or so may become clearer. The next time that someone asks you what the point of astrophysics is, tell them that one day it could save the Universe.