We Can't Reach Andromeda, But Andromeda Is Coming to Us

The Andromeda Galaxy is so far away that its light takes about 2.5 million years to reach us. Point a probe like Voyager 1, the fastest one now leaving our solar system at Andromeda, and it would need about 44 billion years to arrive. That is more than three times the age of the universe. Bigger rocket has barely dented that number, so physicists have spent decades on the alternatives. Light sails, ships that carry whole cities, bubbles of bent space, tunnels through it. Each one runs into its own wall, and the most likely way we ever meet Andromeda needs no ship at all. On a clear autumn night, away from city lights, you can find Andromeda without a telescope. It looks like a faint, stretched smudge near the Great Square of Pegasus. That smudge is one of the most distant things the human eye can see. The light entering your eye from it left around the time our ancestors were making some of the first stone tools, about 2.6 million years ago. And it comes from roughly a trillion stars, several times as many as ours. People have been looking at it for more than a thousand years. The Persian astronomer Al Sufi recorded it in 964 as a small cloud. For centuries after him, many astronomers assumed it was a cloud of gas inside our own galaxy. That changed in October 1923. Working at the 100-inch Hooker telescope on Mount Wilson in California, Edwin Hubble spotted a star in Andromeda that brightened and dimmed on a regular rhythm. On the glass plate, he crossed out the N he had written for nova and wrote VAR for variable. Stars of that kind, called Cepheids, revealed their true brightness through their rhythm, and that gave him the distance. His estimate came out around a million light years, far beyond the edge of our galaxy. Today's figure is two and a half times his. To get a feel for that gap, shrink everything, make the Milky Way a dinner plate 10 inches across. Andromeda becomes a serving platter about 15 inches wide, sitting roughly 21 feet away, about as wide as a two-car garage. Two galaxies, a few strides apart. It almost sounds reachable. Now look at what else is on that model. A single light year is about a thirteenth the width of a human hair. The Sun and everything that orbits it out to the farthest comets would be about the size of one red blood cell sitting on a plate. Andromeda is close only compared with the size of galaxies. For anything humans have built, the space between those two plates is effectively endless. A passenger jet cruises at around 550 miles an hour. A flight from New York to Los Angeles takes you about five or six hours and, to most of us, that feels fast. The fastest machine ever built is NASA's Parker Solar Probe. On its closest passes by the Sun, it reaches about 430,000 miles per hour. Parker could cover that same coast-to-coast flight in about 20 seconds. Light makes Parker look parked. A beam of light covers about 186,282 miles every second, roughly 1,560 times faster than Parker at its very best. So how long would you wait for a Parker class probe to reach Andromeda? On that record speed the whole way, and the trip lasts about 3.9 billion years. Almost as long as Earth itself has existed. Parker only hits that speed while diving toward the Sun, though. The long-distance champion is Voyager 1, which is leaving the solar system at about 38,000 miles per hour. At Voyager's pace, the crossing stretches to the 44 billion years we started with. More fuel helps far less than you might expect. Every extra ton of fuel is more weight that the rest of the fuel has to push, so the tanks grow much faster than the speed does. The Saturn V that sent astronauts to the Moon shows the problem. At liftoff, it weighed about 6.2 million pounds. Only about 100,000 pounds of that was sent on toward the moon, less than 2%. Nearly everything else was fuel, burned mostly to lift more fuel. So why not build a far better engine, one powerful enough to push a ship almost to the speed of light? On paper, physics offers a bonus for trying. Einstein's relativity says that time aboard a fast-moving ship runs slower than time on Earth, and the closer you get to light speed, the bigger the effect. Suppose a ship reaches 99.99999999% of light speed. Its clocks would tick about 70,000 times slower than clocks back home, so every year aboard would match about 70,000 years on Earth. The crew would feel the trip to Andromeda take about 35 years. They could leave as young adults and arrive in middle age. On Earth, about 2.5 million years would pass. Everyone who waved goodbye, every country, language and city the crew knew, would be gone long before the news of their arrival could make it home. If you like seeing the universe measured in dinner plates and road trips, subscribe to Insane Curiosity, because the next number on this list gets harder to believe. The bigger problem is the cost of getting up to speed. You may have heard that a ship gets heavier as it goes faster. Businesses today usually describe it differently. The ship's mass stays the same, but each extra mile per hour costs more energy than the last, and the bill climbs without limit as you approach the speed of light. How big is that bill? Run the numbers for a ship ship weighing just one ton, about as much as a small car. Reaching that speed would take about 10,000 years of all the energy humanity uses today, almost as long as people have been farming. That figure leaves out the fuel itself, and the braking at the other end. Light gets to move at that speed because it has no mass at all. Anything built from atoms, including you and me, has to pay the bill. So the engineers turn the problem around. If a heavy ship cannot go fast enough, could a ship that weighs almost nothing do better? There is a real design for exactly that, backed by serious money and a famous name, and on paper it reaches the nearest star in about 20 years. In 2016, the tech investor Yuri Milner and Stephen Hawking announced a project called Breakthrough Starshot. The plan strips the spacecraft down to almost nothing. The probe is a chip weighing a few grams, about as much as a couple of paper clips, attached to a sail no wider than about 13 feet and far thinner than a sheet of paper. There is no engine on board. A huge array of lasers on the ground, spread over about a square kilometer, would fire a beam with up to 100 gigawatts of power at the sail. That's roughly the output of 100 large nuclear power plants focused onto one tiny mirror. The push of light is gentle, but the sail is so light that it adds up fast. In about 10 minutes, pushed at roughly 10,000 times the pull of Earth's gravity, the probe would reach a fifth of the speed of light. At that speed, it would cover the distance to the Moon in about 6 seconds and reach the Alpha Centauri system in 20 to 30 years. Its first photos would need about 4 more years to travel back. On our dinner plate model, that whole flight to the nearest stars would cover about the width of a single speck of dust. Now aim the same probe at Andromeda instead. At that same speed, the trip takes about 12.5 million years. And the probe would be unlikely to survive. Space holds a thin haze of dust. At that speed, a single grain thinner than a human hair would hit with more than ten times the energy of a 100-mile-per-hour fastball. Over millions of years, the sail would be worn away long before it arrived. The project itself has gone quiet. In 2025, its director described it as on hold. If a machine cannot survive the trip, maybe people can, as long as they never expect to arrive themselves. That is the idea behind the Generation Ship, a giant vessel carrying a self-sustaining community. Thousands of people live on board, grow their food, recycle their water and air, and raise children who take over the ship. If you were born on it, you would live and die somewhere in the dark between the galaxies, and so would your grandchildren. We have tested the first step of that on Earth. In September 1991, eight people sealed themselves inside Biosphere 2, a glass-covered world in the Arizona desert, planning to live there for two years. Within months, the oxygen in the air dropped from about 21% to around 14%. The crew suffered headaches and fatigue, and managers had to pump oxygen in from outside. That was two years, with Earth right outside the door. Give a generation ship a generous speed of 1% of light speed, about 6.7 million miles per hour, already far beyond anything we can build. The crossing to Andromeda would still take about 250 million years. Longer than the time since the first dinosaurs appeared. At 25 years per generation, that comes to about 10 million generations. All of recorded history fits into roughly 200. The ship would also have to keep working the whole time. Voyager, built to last, should run for around 60 years before its power runs out. A generation ship would need to last about 4 million times longer. Every plan so far tries to push a ship through space. The next idea leaves the ship alone and moves the space around it. In 1994, Miguel Alcubierre, a Mexican physicist finishing his PhD at Cardiff University in Wales, described a kind of warp drive allowed by Einstein's equations. It would squeeze space in front of a ship and stretch it out behind, while the ship rides inside a calm bubble that is carried along. You've probably seen the same trick at an airport. Step onto a moving walkway and stand still. Your legs are doing nothing, yet you glide past everyone else because the floor itself is moving. In Alcubierre's design, space plays the part of the floor. Inside its bubble, the ship never outruns light. So Einstein's speed limit holds right where it is. The catch is the fuel. The design needs something called negative energy, energy below that of perfectly empty space. Every kind of energy we use, from gasoline to sunlight to the battery in your phone, is positive. Physics does allow small amounts of the strange kind. Put two metal plates extremely close together in a vacuum, and they get pushed toward each other, because the space between them holds slightly less energy than empty space outside. The effect is named after Hendrik Casimir, and in 1997 Steven Lamoureux measured it in the lab. It is real, and it is unimaginably far from the amounts a warp drive would require. In 2021, two physicists, Alexey Bobrik and Gianni Martire, showed that a warp bubble made from ordinary positive energy is possible in principle. Their version comes with a limit that matters here. It can only move slower than light. That leaves one more shortcut, the boldest of all. Skip the space in between entirely. And in 2022, headlines around the world announced that physicists had built one inside a computer chip. That shortcut is a wormhole. The physicist John Wheeler gave it the name in 1957. But the idea is older. In 1935, Einstein and his colleague Nathan Rosen showed that his equations allow a kind of bridge linking two distant regions of space. A wormhole would be a tunnel joining two distant places, so that a crossing of millions of light years could shrink to a short hop from one mouth of the tunnel to the other. Keeping it open is the hard part. In 1962, Wheeler and Robert Fuller showed that the Einstein-Rosen bridge would pinch shut too quickly for light itself to get through. In 1988, Kip Thorne, who later advised the film Interstellar, and Michael Morris worked out what it would take to hold one open. A lining of exotic matter, once again with negative energy. No wormhole has ever been observed anywhere. As for the 2022 headlines, that experiment ran on nine qubits of Google's Sycamore quantum processor. It simulated a mathematical model that behaves in some ways like a wormhole. Critics, including the computer scientist Scott Aronson, pointed out that no actual tunnel through space was created, and nothing traveled anywhere faster than light. So every road to Andromeda ends at one of three walls. Too slow, too fragile, or waiting on a kind of energy no one can make. Andromeda, however, is already on its way. Most galaxies are moving away from us as the universe expands. Andromeda is one of the few exceptions. In 1912, at Lowell Observatory in Arizona, the astronomer Vesto Slipher found its light shifted slightly toward the blue end of the spectrum, the sign of something moving our way. Modern measurements show it closing in on our galaxy at about 250,000 miles an hour. That would carry it from New York to Los Angeles in about 36 seconds. Back on our model, the platter is sliding towards the dinner plate. Gravity is doing the work. The two are the biggest members of the local group, a family of several dozen galaxies, and their combined pull keeps drawing them together. For years, the open question was the sideways motion. A galaxy heading straight at us would collide. One drifting sideways might swing past. Measuring that drift from so far away is extraordinarily hard. In 2012, a team led by Roland van der Morel did it. They used Hubble to photograph the same small patches of Andromeda again and again, over five to seven years. And the result sounded final. The galaxies would meet head-on in about four billion years and merge over the two billion years after that. Now we've shown that a collision is inevitable, said one of the team, Parag Guha Tata. NASA released illustrations of what Earth's night sky would look like along the way. About 3.75 billion years from now, Andromeda's disk would fill much of the sky, and its gravity would already be bending the Milky Way out of shape. What would you make of that sky if anyone were still here to look up? That forecast became one of the best-known facts in astronomy. Then, in 2025, a team ran the numbers again with newer data, and the ending changed. The team was led by Till Savala of the University of Helsinki, and the study appeared in Nature Astronomy. They combined the latest measurements from Hubble and from Gaia, a European spacecraft that tracked nearly 2 billion objects in our galaxy before it was switched off in March of 2025. Then they ran about 100,000 simulations to cover every uncertainty in the data. The result was a coin toss. In about half of the simulations, the Milky Way and Andromeda merged within the next 10 billion years. In the rest, they never got close enough in that time. Often the two galaxies first swing past each other about half a million light years apart, and only fall back together much later. And the head-on crash in about 4 or 5 billion years, the one in all the illustrations, came out at only about 2%. Part of the change comes from two smaller galaxies. With only the two big galaxies in the model, they merged in a bit under half the runs. Adding the Triangulum Galaxy raised that to about About 2 and 3. The Large Magellanic Cloud, the Milky Way's biggest satellite, had the opposite effect. On its own, it cut the odds to about 1 in 3, because it tugs our galaxy sideways. With all four included, the odds land just above 50%. We've simply been able to explore a much larger space of possibilities, taking advantage of new data, Zavala said. In most of the scenarios where the galaxies do merge, it happens later than the old forecast, Almost around 8 to 10 billion years from now. Our galaxy may have another collision first. A 2019 study from Durham University found that this same satellite galaxy is on course to fall into the Milky Way in about 2.4 billion years. The key number is still Andromeda's sideways drift, which is very hard to measure from that distance. Gaia's remaining data releases, the last of them expected near the end of the decade, should sharpen it. And if the merger does happen, almost nothing would hit anything. Our galaxy has been through this before. Gaia data revealed the scattered stars of a smaller galaxy, named Gaia Enceladus, that merged with ours when our galaxy was only a few billion years old. You can still pick out its stars today by their orbits and their chemistry. Stars are tiny compared with the distances between them. Shrink the Sun to a ping-pong ball and the nearest star would be another ping-pong ball about 700 miles away, roughly as far as New York is from Chicago. Galaxies are crowded for their size, so they collide. Their stars are so spread out that two galaxies can pass right through each other while almost no stars touch. Gas clouds, on the other hand, would slam into each other and light up with new stars. Over a couple of billion years, the two spirals would settle into one giant, 5. Rounder Galaxy Astronomers already have a nickname for it, Milkomeda. Simulations suggest the Sun would probably end up much farther from the center than it is today, and there is a small chance it gets flung out of the merged galaxy altogether. If you could watch from a living Earth, you would see a sky full of glowing gas and young stars. That viewer will not exist. The Sun grows about 10% brighter every billion years. And within roughly a billion years, it should be hot enough to boil away Earth's oceans. Some of what you've just heard is solid, and some of it is still uncertain. Andromeda's distance and its approach toward us are well measured. Its sideways motion is the weak spot, and that is why the forecast moved from inevitable to a coin toss. The travel times assume a probe could hold a constant speed for millions of years, which no machine has done. The time dilation numbers ignore the fuel and the braking at the far end. Starshot is on hold, warp drives and wormholes exist only as equations. And the fate of the Sun in any merger is a probability, not a schedule. Back to the two plates on the garage floor, 21 feet apart. Right now, the gap between them is closing by about 68 miles every second. Since you started watching this video, the Milky Way and Andromeda have moved roughly 80,000 miles closer together, about a third of the way from the Earth to the Moon. By this time tomorrow, the gap will be about 6 million miles smaller. The next time you find that faint smudge near Pegasus, you will be looking at a galaxy slightly closer than the last time you saw it. We cannot build a ship that reaches Andromeda. Gravity is bringing the two galaxies toward each other anyway. And whether they finally collide or only pass close is a question the next generation of measurements will settle. And if that sounds tiny next to the distance between galaxies, wait until you see what a single light year looks like on a map of the United States. That's the next video!