James Webb Just Found Galaxies That Cannot Exist — And No One Can Explain It

本站添加: 2026-05-12

[00:00:00]There is a patch of sky so small you could cover it with a grain of sand held at arms length.

[00:00:06]No stars visible to the naked eye.

[00:00:10]No galaxies you could point to, just black.

[00:00:15]Empty.

[00:00:16]Nothing.

[00:00:18]And that is exactly where scientists chose to aim the most powerful space telescope ever built.

[00:00:24]What they found there didn't just surprise them. It didn't just challenge their models. It broke them completely.

[00:00:34]And the more data came in [music] the worse it got until one of the world's top cosmologists stood in front of a room full of scientists [music] in January 2025 and used a word you almost never hear in professional astronomy.

[00:00:48]He called it a crisis.

[00:00:50]This is the story of what the James Webb Space Telescope found in the dark. And why it may mean that everything we thought we knew about how the universe was built, the entire [music] textbook, is wrong.

[00:01:03]Let me take you back to 1995 for a second [music] because that's where this story really starts.

[00:01:10]The Hubble Space Telescope was brand new, controversial, expensive. A lot of people thought it was a waste of money.

[00:01:18]And then the team running it made a decision that seemed at the time almost reckless.

[00:01:24]They pointed Hubble at a patch of empty sky, [music] literally a blank region near the Big Dipper. And they held it there for 10 consecutive [music] days. No blinking, no moving, just staring into the dark. What came back stopped the scientific world cold. That empty patch wasn't empty at all. It was packed with thousands of galaxies, [music] each containing hundreds of billions of stars, stretching back almost to the beginning of time. That single image, the Hubble Deep Field, changed how humanity understood its place in the universe.

[00:02:03]Now, fast forward to 2022.

[00:02:06]JWST, the James Webb Space Telescope, took $10 billion and 30 years to build, and it is not an upgrade to Hubble.

[00:02:16]It operates in infrared light, which lets it see through dust clouds and across distances that make Hubble look short-sighted. It's like going from a disposable camera to a professional cinema lens.

[00:02:29]And the team running it decided to do the same thing Hubble did, point it at a tiny, seemingly empty patch of sky, this time near the constellation Fornax, and just stare.

[00:02:41]They called the program JADES, the JWST Advanced Deep Extragalactic Survey. The goal was to find the earliest galaxies ever formed, the very first structures that emerged after the Big Bang.

[00:02:54]They knew they'd find something. They had no idea what they were about to walk into.

[00:03:00]The first data came back in late 2022, and the scientists involved in analyzing it were not prepared for what they saw.

[00:03:09]Six objects, glowing faintly red in the infrared, small, distant, ancient, almost beyond comprehension, existing only 500 to 700 million years after the Big Bang, over 13 billion years ago.

[00:03:25]That alone wasn't the shock. Scientists expected to find early galaxies. What they did not expect was the size.

[00:03:34]These six objects were monstrous.

[00:03:37]According to the analysis led by Associate Professor Ivo Labé from Swinburne University of Technology in Australia, the largest of these candidate galaxies had a mass equivalent to up to 100 billion times the mass of our sun day.

[00:03:52]They were nearly as large in a star-packed as the Milky Way, a galaxy that took roughly 13 billion years to build.

[00:04:00]And they had apparently done it in a few hundred million years.

[00:04:04]Labbe said it plainly, "We've never observed galaxies of this colossal size this early on after the Big Bang. This is too big to even exist within current models."

[00:04:15]His co-author, Erica Nelson, an astrophysicist at the University of Colorado Boulder, put it even more memorably.

[00:04:22]She looked at the data and said, "It's bananas." She meant that in the most scientific way possible. You just don't expect the early universe to be able to organize itself that quickly.

[00:04:33]These galaxies should not have had time to form.

[00:04:36]The paper went into nature, the community took notice, and then it got stranger. [music] Because the density of these objects, how many of them were out there, was not a little bit off from what models predicted. It was off by more than 10 times.

[00:04:52]10 times more of these massive early galaxies than every simulation and theory had suggested. That's not a rounding error. That's not a measurement glitch.

[00:05:04]That is a fundamental mismatch between the universe as we modeled it and the universe as it actually is.

[00:05:12]Stay with me because what comes next is where it goes from interesting anomaly to complete paradigm problem.

[00:05:20]In October 2023 and January 2024, the JADES team pointed JWST at the same deep field region again.

[00:05:30]This time, using an instrument called NIRSpec, a near-infrared spectrograph [music] that doesn't just take pictures. It reads the chemical fingerprint of light.

[00:05:41]It tells you what something is made of, how fast it's moving, and critically how far away it is.

[00:05:47]They found a galaxy, a single galaxy, confirmed, spectroscopically verified, real. Its name is JADESGSZ140, [music] which is not a catchy name, but what it represents is extraordinary. It was announced in May 2024 as the most distant galaxy ever confirmed by human beings. Its light left that galaxy 13.4 billion years ago. The universe at that moment was less than 300 million years old, about 2% of its current age. We are looking at a photo of the universe when it was an infant, not a toddler, an infant. The redshift, the number astronomers use to measure cosmic distance by how much a galaxy's light has been stretched by the universe's expansion, was Z equals 14.32.

[00:06:44]That number doesn't mean much in everyday terms, so here's the translation. We are seeing this galaxy as it existed almost at the very beginning of everything.

[00:06:55]And the galaxy was 1,600 light years across. That's enormous for its age.

[00:07:02]That alone was a puzzle.

[00:07:05]But then, the follow-up data landed.

[00:07:09]Astronomers trained ALMA, the Atacama Large Millimeter Array in Chile, on this galaxy to get more detail, and they found oxygen inside it.

[00:07:18]Oxygen in a galaxy that existed 290 million years after the Big Bang. Now, you might be wondering, "So what? Oxygen is common. We breathe it. It's everywhere. What's the big deal?" Here's the big deal.

[00:07:35]Oxygen is not a product of the Big Bang.

[00:07:38]The Big Bang produced hydrogen, helium, and trace amounts of lithium. That's it.

[00:07:45]Every other element in the periodic table, carbon, nitrogen, oxygen, iron, all of it had to be forged inside stars.

[00:07:53]Stars that had to be born, grow, burn through their fuel over millions of years, and then explode as supernovae, scattering those elements into space.

[00:08:05]Then another generation of stars had to form from that scattered material, and possibly another after that.

[00:08:13]That is multiple complete cycles of stellar life and death. And JADES-GS-Z140 had already done all of that within its first 290 million years of existence.

[00:08:31]Sander Schouws, a PhD candidate at Leiden Observatory, who was part of the team that detected the oxygen, described it like this.

[00:08:40]It is like finding an adolescent where you would only expect babies.

[00:08:45]ESO astronomer Gergö Popping called it a surprise and said plainly that it suggests galaxies can form more rapidly after the Big Bang than previously thought. The lead author of the JADES-GS-Z140 discovery, Stefano Carniani from the Scuola Normale Superiore in Pisa, and Kevin Hainline from the University of Arizona Steward Observatory, who co-authored the chemical analysis, both noted that existing computer simulations are simply unable to reproduce this galaxy.

[00:09:23]Not just poorly unable, the simulations don't produce anything like it.

[00:09:28]Then, in May 2025, just 1 year later, JWST broke its own record. A galaxy designated Mom Z14 was confirmed at a redshift of Z = 14.44.

[00:09:46]Even older than JADES JADES-GS-Z140.

[00:09:51]And this one was brighter, more compact, and more chemically enriched. It had elevated levels of nitrogen, another element that requires stellar generations to produce. It had also apparently cleared [music] the surrounding region of neutral hydrogen gas.

[00:10:07]In a universe that at that point was essentially nothing but neutral hydrogen gas.

[00:10:13]That last detail is worth sitting with.

[00:10:16]Imagine standing in a room completely packed with fog. So thick you can't see a foot in front of your face.

[00:10:25]And somehow in the middle of that room something had cleared a perfect bubble of clean air around itself. That's what Mom Z14 had done. And no one fully understands how.

[00:10:42]There's a discovery I need to tell you about now that doesn't get nearly enough attention. And [music] it's possibly the most philosophically unsettling of all of them. Its name is ZFUDs 7329.

[00:10:56]Professor Karl Glazebrook at Swinburne University had been chasing this object for 7 years. 7 years of telescope time.

[00:11:04]Failed observations. Data that was too faint and too red to measure from the ground.

[00:11:10]In 2024, JWST finally got a clean look at it.

[00:11:17]ZFUDs 7329 is a massive galaxy from more than 11 billion years ago. It contains more stars than our entire Milky Way. But that's not the headline. The headline is when those stars were born, they formed 1.5 billion years before the galaxy we're observing, meaning this galaxy already massive 11 billion years ago had finished the explosive burst of star formation that built it another 1.5 billion years before that.

[00:11:51]So, we're looking at stellar populations born somewhere around 12.5 billion years ago when the universe was barely getting started. Here's what makes this not just interesting, but genuinely troubling physics.

[00:12:05]In the standard model of cosmology, galaxies need dark matter to form. Dark matter, the invisible scaffolding of the universe, is supposed to clump together first, and its gravitational pull then drags gas and dust into its halos, which eventually collapses into stars and galaxies. It's a slow, structured process.

[00:12:28]Dark matter first, then gas, then stars, then galaxy. ZFUDs 7329 appears to have built itself in a rapid burst of 200 to 400 million years before the dark matter halos at that epoch could have been large enough to support it. It's like building a skyscraper before the concrete has had time to dry.

[00:12:54]Glazebrook said his team had been searching for this galaxy for 7 years because they suspected something unusual. They just had no idea how unusual. So, we have impossible galaxies forming faster than physics allows. We have a galaxy with oxygen [music] and nitrogen that required multiple stellar generations nobody has time to account for. We have a galaxy that apparently formed without the dark matter scaffolding our entire model depends on.

[00:13:24]And then, and this is where things get genuinely strange. JWST started finding objects that scientists don't even have a category for yet.

[00:13:35]They started calling them little red dots.

[00:13:39]That's not a scientific nickname. That's what they look like. Tiny, intensely red points of light scattered through deep field images.

[00:13:49]Scientists at Penn State and elsewhere first identified them in JWST's earliest data sets from 2022.

[00:13:58]At first glance, they looked like early galaxies.

[00:14:01]Then the spectroscopy started coming in and everything got weird.

[00:14:05]Little red dots exist at redshifts above four when the universe was less than 1.5 billion years old. They are packed with what appears to be aging, cold red stars.

[00:14:18]That alone is strange because young galaxies should be dominated by hot, young, blue stars. These are acting old before they should be able to.

[00:14:30]Many of them appear to harbor supermassive black holes at their centers. Black holes with masses comparable to the one at the center of our Milky Way.

[00:14:39]Black holes that took our galactic center billions of years to grow to that size.

[00:14:45]Scientists dubbed them universe breakers, not as a metaphor, as a genuine descriptor for what they do to our theoretical models.

[00:14:55]Then in July 2024, the RUBIES team, a collaboration using nearly 60 hours of JWST time, collected spectra for 4,500 distant galaxies and found 35 of these little red dots.

[00:15:11]One of them was unlike anything they'd seen.

[00:15:15]They called it the Cliff.

[00:15:17]The Cliff earned its name from a dramatic drop in its spectrum, a sharp, sudden feature that didn't match any known galaxy profile. Its light had traveled 11.9 billion years to reach us.

[00:15:30]Its properties were so extreme, so outside the expected range, that Anna de Graaff from the Max Planck Institute for Astronomy said it plainly, "The extreme properties of the cliff forced us to go back to the drawing board and come up with entirely new models."

[00:15:46]Entirely new models.

[00:15:49]Three competing theories have now emerged for what little red dots actually are.

[00:15:54]The first, young black holes, roughly 100 times less massive than previously believed, wrapped in thick cocoons of gas. They are actively consuming, and the heat from that feeding is what shines through. The second, a brand new class of cosmic object called black hole stars, something the universe has never been caught making before, at least not that we've seen.

[00:16:19]The third theory says they are gas clouds in the act of collapsing into black holes, that the red glow isn't a galaxy at all, but something being violently born. Researcher Jenny Greene, who worked on one of the first little red dot studies in 2023, put the scientific moment in perspective. She said, "I knew it was really cool, and I knew people were going to care about it.

[00:16:44]I didn't know there was going to be one paper a day for 3 years."

[00:16:48]One paper a day.

[00:16:50]Every day.

[00:16:51]For 3 years. And we still don't know what they are.

[00:16:54]I want to pause here for a second because I need to make sure you feel the weight of what's happening.

[00:17:00]In the last few minutes, I've described to you galaxies that are too big, too old, too chemically mature, and too organized for the time in which they exist. Objects that don't match any category we have. An entire population of cosmic structures 10 times more abundant than any model predicted that appears to have formed before the laws of physics as we understand them would have permitted.

[00:17:26]Every one of these things was found in the same small patch of seemingly empty sky.

[00:17:31]And the deeper JWST looks, the more of them appear.

[00:17:35]Now, what if the issue isn't just the galaxies? What if it's the stars inside them?

[00:17:41]Because JWST may have just found evidence that the universe's very first stars were not like stars at all.

[00:17:49]In the standard picture of cosmic history, the first stars, called population three stars, formed a few hundred million years after the Big Bang from pristine clouds of hydrogen and helium. They were huge, hot, and burned fast.

[00:18:07]They lived for a few million years, exploded, and seeded the universe with heavier elements. That's the story.

[00:18:15]Clean, logical, well-modeled. But Cosmin Ilie at Colgate University, working with Katherine Freese at the University of Texas at Austin, published a paper in the Proceedings of the National Academy of Sciences in 2025, identifying four objects in JWST data that do not match that picture at all.

[00:18:38]These four objects are consistent in their observed spectra and their morphology with something called a supermassive dark star.

[00:18:47]The dark star is not a black hole. It is not a conventional star.

[00:18:53]It is a hypothetical object made primarily of hydrogen and helium, like an ordinary star, but [music] powered not by nuclear fusion, but by the annihilation of dark matter.

[00:19:05]Dark matter makes up roughly 25% of the universe, but we We never directly detected it. We know it exists because of its gravitational effects.

[00:19:16]And one of its proposed properties is that in very dense regions of the early universe, dark matter particles could collide with their own antiparticles and annihilate releasing enormous amounts of energy in the process.

[00:19:33]Frees' team proposed that in the very first halos of dark matter, the first gravitational pockets to form after the Big Bang, this annihilation energy could have inflated what would have been ordinary protostars into something unrecognizable. Objects that could reach a million times the mass of our sun day, objects that could span billions of kilometers.

[00:19:58]Objects that would be extraordinarily bright and look from a distance very much like what JWST has been detecting near the edge of the observable universe.

[00:20:11]Frees stated, "For the first time, we have identified spectroscopic supermassive dark star candidates in JWST, including the earliest objects at redshift 14, only 300 million years after the Big Bang, weighing a million times as much as the sun."

[00:20:29]"Such early dark stars are important not only in teaching us about dark matter, but also as precursors to the early supermassive black holes seen in JWST that are otherwise so difficult to explain."

[00:20:44]Think about that last part. The early supermassive black holes that JWST keeps finding, black holes that are far too big for their age, are otherwise impossible to explain.

[00:20:55]Dark stars dying and collapsing might be exactly the seed mechanism that produces them.

[00:21:01]If this is confirmed, it would mean the very first chapter of cosmic history, the chapter we thought we'd written reasonably well, is completely wrong.

[00:21:11]The first stars were not the population three fusion stars of our models. They were something else entirely, something that ran on dark matter, something we've never seen.

[00:21:23]And JWST may be looking directly at their ancient light right now.

[00:21:28]I want to bring you to one more piece of this, and in some ways it's the most alarming. Because it isn't just about what JWST found, it's about what it confirmed. A problem that has been building for over a decade in cosmology, a problem so persistent and so thoroughly verified that it can no longer be dismissed.

[00:21:48]They call it the Hubble tension.

[00:21:51]And in January 2025, a physicist at Duke University named Dan Scolnic stood in front of the American Astronomical Society and called it a crisis.

[00:22:02]Here's what it is. There are two completely independent ways to measure how fast the universe is currently expanding, the Hubble constant. The first method uses the early universe, the cosmic microwave background, the faint afterglow of the Big Bang, which gives you a measurement of about 67 km per second [music] per megaparsec.

[00:22:23]The second method uses nearby galaxies and exploding stars called supernovae as distance markers, [music] giving you a measurement of about 73 km per second per megaparsec.

[00:22:36]Both methods are mature.

[00:22:39]Both have been refined for decades.

[00:22:42]And they disagree by enough that if you're being rigorous about the statistics, the disagreement has now crossed six sigma.

[00:22:55]In science, five sigma is considered the threshold for a confirmed discovery. Six sigma means the probability that this disagreement is a statistical fluke is essentially zero.

[00:23:09]This isn't two instruments giving slightly different readings.

[00:23:13]This is a fundamental split in how the early universe and the modern universe appear to be behaving. And no theory in our current toolkit fully bridges that gap. Adam Riess at Johns Hopkins University, who won the Nobel Prize for discovering dark energy, led a team that used JWST to make the most precise measurement yet of Cepheid variable stars, the benchmark objects used in one of those methods. His conclusion was unambiguous.

[00:23:47]He said, "We've now spanned the whole range of what Hubble observed, and we can rule out a measurement error as the cause of the Hubble tension with very high confidence."

[00:23:57]JWST didn't solve the tension. It confirmed it is real. The measurement is not wrong. The physics is wrong. And if the standard model of cosmology, called lambda CDM, cannot account for the expansion rate of the universe, cannot account for the age and size of galaxies at cosmic dawn, cannot account for the chemistry of those galaxies, cannot predict the existence of little red dots or dark star candidates, [music] then we are not looking at a model that needs tweaking. We are looking at a model that may need to be replaced.

[00:24:33]Johns Hopkins researchers wrote it plainly, "The discrepancy between the observed expansion [music] rate and the predictions of the standard model suggests that our understanding of the universe may be incomplete."

[00:24:49]May be incomplete.

[00:24:51]That is the polite scientific version of we might be fundamentally wrong about something very basic. So, where does all of this leave us?

[00:25:00]Here is the picture as it stands. JWST pointed at a grain of sand-sized patch of empty darkness has in 3 years of operation found galaxies 10 times more numerous than expected. Found a galaxy JADES GSZ140 that is the most distant object ever confirmed already producing oxygen within 290 million years of the Big Bang.

[00:25:28]In a universe where the physics of that timeline doesn't work.

[00:25:33]Found ZFUDs 7329, a massive galaxy that appears to have assembled itself without the dark matter scaffolding our entire model depends on.

[00:25:44]Found a class of objects, little red dots, so strange that scientists are generating a paper a day trying to figure out what they are.

[00:25:52]Found candidates for dark stars.

[00:25:54]Hypothetical objects that would rewrite the very first chapter of stellar history.

[00:25:59]And confirmed definitively that the universe appears to be expanding faster than it should be by a margin that no instrument error can explain. Every single one of these findings points in the same direction. The universe assembled itself far faster, far more efficiently, and far more chaotically than we thought. The standard model, the neat story of slow gravitational assembly that has dominated cosmology for decades is in serious trouble. Jacob Helton, the graduate researcher at the University of Arizona who worked on the JADES chemistry paper, noted that the fact JWST found JADES GSZ140 [music] inches such a tiny region of sky means there should be more like it everywhere.

[00:26:49]If we looked at the whole sky, which we can't do with JWST. We would eventually find more of these extreme objects, he said.

[00:26:57]More of them, everywhere. The universe absolutely scattered with things that, by our current best understanding, should not exist.

[00:27:06]Now, I want to be honest with you about what this does and does not mean, because there are two easy mistakes to make here.

[00:27:13]The first is to say, this proves the Big Bang never happened, or that the universe is secretly younger, or that dark matter doesn't exist. None of that follows from the data.

[00:27:25]What the data says [music] is that our model of how things proceeded after the Big Bang is incomplete. The early universe is doing something we didn't predict. That is a very different claim from saying everything is wrong.

[00:27:40]The second mistake >> [music] >> is to assume this will all be explained away, that some paper will come out, tweak an equation, [music] and everything will snap back into place. That has not happened.

[00:27:52]Three years of almost daily publication from teams at Harvard, >> [music] >> Johns Hopkins, Max Planck, Texas, Arizona, Leiden, Swinburne, Penn State, [music] Colgate, and the gap between prediction and observation keeps widening, not narrowing.

[00:28:10]The honest position is the exciting one.

[00:28:12]We are at a genuine inflection point in our understanding of the cosmos. The kind of moment that happens once or twice in a century.

[00:28:22]The kind of moment that Hubble himself triggered when he showed that the Milky Way wasn't the whole universe, just one galaxy among countless others.

[00:28:32]Right now, astronomers are not sure whether the solution to what JWST is finding will require modifying dark matter, modifying gravity, introducing some new form of early dark energy, or something even more radical that no one has fully articulated yet. The phrase you keep seeing in research papers is growing chasm between theory and observation. A chasm, not a crack, not a discrepancy, a chasm.

[00:29:02]There is something almost philosophical about where this leaves us. We built the most powerful eye humanity has ever turned toward the sky. We pointed it at the dark. We thought we would see the universe's earliest moments and find them matching the story we had already written.

[00:29:21]Instead, we found a universe that had already been busy, monstrously busy, building and burning and exploding and chemically enriching entire galaxies before our models say had enough time to strike the first match.

[00:29:36]We found the light of stars that may have been powered not by hydrogen burning in their cores, but by dark matter annihilating itself in the first gravitational pockets of the cosmos, objects we have no framework for, no precedent for, and until 3 years ago, no way to see. We found that the universe is expanding in a way that even our best cosmological theory cannot account for. And the gap between what we expect and what we measure is now too large to ignore.

[00:30:09]And sitting underneath all of it is the same unsettling question, the one that every one of these discoveries circles back to from a different angle. If the universe built itself this fast, if those first galaxies were massive and complex and chemically evolved within just 2% of the universe's current age, then what does that tell us about the rules that were in play at the beginning? Were they different rules?

[00:30:36]Were there forces or particles we haven't characterized yet? Is there an era of cosmic history right at the start that we've completely misdescribed?

[00:30:47]The telescope is still running. More data is coming in. More little red dots will be cataloged. More record-breaking distant galaxies will be confirmed. And with each one, the gap between the universe we imagined and the universe we actually live in will come into sharper relief. We built a 10 billion dollar machine, pointed it at a patch of empty sky, and found something that by every model we had should not exist.

[00:31:16]The question now isn't just what those things are. The question is what they're telling us about everything else we thought we knew.

[00:31:23]And I don't think we're ready for that answer yet.