This Will Change How You See the Universe Forever.

Added: 2026-05-06

[00:00:00]When the James Webb Space Telescope delivered its first scientific images in the summer of 2022, astronomers noticed something almost immediately. Tiny red points scattered everywhere.

[00:00:15][music] In almost every single image the telescope took, extremely compact, distinctly red, and far more numerous than anyone expected, they showed up whether the telescope was pointing at galaxy clusters, at empty sky, at nebuli, they were everywhere.

[00:00:34]Astronomers call them little red dots.

[00:00:37]Over 340 have been cataloged. They appear around 600 million years after the Big Bang. They persist for about a billion years and then they disappear completely. Whatever they are, they had a window of existence and that window closed. Nobody can agree on what they are. The first theory was that they were compact galaxies packed with stars at densities so extreme that the night sky from inside one would be blinding. That did not hold up. The second theory was that they were super massive black holes actively feeding on matter surrounded by hot gas discs. But they show none of the X-ray signatures that every known active black hole produces. Their spectra do not match. In January 2026, one team published in Nature that they are young black holes 100 times less massive than previously believed. hidden inside dense cocoons of ionized gas that glow red as the black hole devour its surroundings.

[00:01:35]The same month, a different team proposed something stranger. That they are not black holes at all. That they might be an entirely new type of star.

[00:01:45]Stars so massive, so short-lived, and so unlike anything in the modern universe that they represent a class of object we have never seen before. Stars potentially powered not by nuclear fusion but by dark matter. These objects were completely invisible to every telescope ever built before web. They are rewriting the first billion years and they are just one of at least three categories of impossible objects that JWST keeps finding in the early universe. None of this is speculation.

[00:02:18]All of it was measured. and what it means for how we understand the origin of everything is the part that changes you once you understand it. To understand why these little red dots matter and why some astronomers are now openly using the word crisis, you need to understand what the James Web Space Telescope actually is, what it was built to do, and what it found the moment it started doing it. The telescope's primary mirror measures 6.5 m across. 18 goldcoated burillium hexagons arranged in a honeycomb pattern collecting more than six times as much light as Hubble's mirror. But the critical difference is not size, it is wavelength. Web sees infrared and that matters enormously because the expansion of the universe stretches light from distant galaxies into longer wavelengths. A galaxy that emitted visible light 13 billion years ago does not arrive at Earth as visible light. It arrives as infrared radiation, invisible to Hubble, invisible to your eyes, invisible to every optical telescope ever pointed at the sky, but not to web. The telescope sits at the L2 Lrange point, 1.5 million km from Earth.

[00:03:31]It is cooled to minus233° C, by a sunshield the size of a tennis court. cold enough that the telescope's own thermal radiation does not drown out the faint whispers of light arriving from the edge of the observable universe. It launched on Christmas Day 2021.

[00:03:51]The precise launch trajectory conserved enough propellant that its operational lifetime has been extended well beyond the original 10-year design goal.

[00:04:01]Current estimates suggest 20 years or more of operations. This telescope could still be producing data in the 2040s, but in its first four years alone, it has already done something nobody anticipated. It has shown us that the early universe looks nothing like we predicted, and the little red dots are only one part of that story. Here is what we thought we knew. The universe is 13.8 billion years old. That number is not a rough estimate. It is one of the most precisely measured quantities in all of physics. Pinned down to within about 40 million years by the plank satellites measurements of the cosmic microwave background radiation. It is a hard measurement. After the big bang, the universe expanded and cooled. For the first few hundred million years, it was dark. No stars, no galaxies, just hydrogen and helium gas slowly collapsing under gravity, eventually condensing into the first luminous objects. The process was supposed to be slow. Gravitational collapse takes time.

[00:05:07]Gas has to cool before it can condense into stars. Stars have to form before heavy elements like oxygen and carbon can be produced through nuclear fusion inside stellar cores. And those heavy elements have to accumulate over generations of stars living and dying before complex galaxy scale structures can assemble. The standard model of cosmology called lambda CDM predicted a specific timeline for all of this.

[00:05:35]Galaxies within the first 500 million years would be exceedingly rare. They would be small. They would be dim. They would be, in the words of one researcher, fledgling galaxies, cosmic infants taking their first steps. And for two decades, the data from Hubble agreed with this story. Hubble's observations at moderate distances matched the models beautifully. The universe at 1 billion years old looked roughly the way the simulation said it should. The theory was working. In December 1995, the director of the Space Telescope Science Institute made a decision that several of his colleagues openly called a waste of valuable telescope time. He pointed the Hubble Space Telescope at a tiny empty-looking patch of sky near Ursa Major and committed 10 full days of observation to it. This was the most powerful optical telescope ever launched, and he was aiming it at nothing. The resulting image, the Hubble Deep Field, contained approximately 3,000 galaxies in a patch of sky so small that a grain of sand held at arms length would cover the entire area. They repeated the experiment in 2004 with the Hubble ultra deep field, roughly 10,000 galaxies in an even smaller area. Every time they looked harder at empty sky, they found it was full. Scaling the density across the entire sky gave the famous estimate 100 to 200 billion galaxies in the observable universe. That number held for nearly 20 years. Then in 2016, a team led by Christopher Consulis at the University of Nottingham proved that number was wrong by a factor of 10. They built mathematical models not just of what Hubble could see, but of what it could not see. The corrected number was at least 2 trillion galaxies and 90% of them had never been directly observed by any telescope, any satellite or any instrument ever built by human beings.

[00:07:36]The early universe was packed with small, dim, irregular galaxies. Messy clumps of stars that served as the building blocks of everything we see today. Over billions of years, they collided, merged, consumed each other, and assembled piece by piece into the larger structures we observe. The Milky Way is one of those assembled structures. Our galaxy is not a pristine original. It is a cannibal. Astronomers have identified at least 60 smaller galaxies or remnants that the Milky Way has absorbed over its history. The Sagittarius dwarf elliptical galaxy is being stripped apart and incorporated into our disc and halo right now as you listen to this. But the small galaxies that survived intact in the distant universe, the ones that were not consumed, most of them fell below Hubble's detection floor. They were so faint that their light by the time it reached us was indistinguishable from the random noise of the detector. They existed. They had mass. They contained stars. Some of them hosted planets, but they whispered and the universe was loud and the instruments could not separate signal from noise. Web was engineered to hear those whispers and what it heard was not whispers at all. It was a roar.

[00:08:54]Within weeks of turning on, astronomers saw big, bright galaxies everywhere in the early universe. far more luminous, far more massive, and far more structurally mature than any theoretical model predicted should be possible at those distances. Researchers at Penn State informally dubbed them universe breakers. Because at first glance, the galaxies the telescope was seeing seemed to break the standard model of cosmology. There simply was not enough time since the Big Bang for that many stars to have formed, for that much mass to have assembled, for that much structure to have emerged. In July 2022, just weeks after the first calibration images were released, a team at Swinburn University of Technology in Melbourne identified several candidate galaxies at red shifts above 12, meaning their light had been traveling for over 13.3 billion years. Follow-up spectroscopic confirmation came in 2023 and 2024.

[00:09:52]The distances were real, the galaxies were real, and they should not have been there. Then came the census. In June 2025, the Cosmos Web Collaboration released the largest map of the universe ever constructed. Nearly 800,000 galaxies cataloged across almost all of cosmic time, built from over 200 hours of observation by the James Webb Space Telescope. The image, if printed at the same resolution as the Hubble Ultra Deep Field on a standard piece of paper, would fill a mural 13 ft wide by 13 ft tall at the same depth. And the result was not what the models predicted, not by a small margin, by a factor of 10.

[00:10:35]Before Web turned on, Caitlyn Casey at UC Santa Barbara and Jhan Cartalopee at the Rochester Institute of Technology, who co-led the survey, had made their best predictions about how many galaxies would be visible in the first billion years after the Big Bang. They used the best physics they had, the best simulations ever built, and their answer was clear. Galaxies in the first 500 million years should be incredibly rare.

[00:11:05]Casey described it simply. It makes sense. The Big Bang happens and things take time to gravitationally collapse and form and for stars to turn on. There is a time scale associated with that.

[00:11:18]And the big surprise is that with JWST, we see roughly 10 times more galaxies than expected at these incredible distances. We are also seeing super massive black holes that are not even visible with Hubble. 10 times more galaxies, black holes that Hubble could not detect, and the gap between the models and reality kept growing. In January 2026, the telescope confirmed the most distant galaxy ever observed, a galaxy called Mom Z14, sitting at a red shift of 14.44.

[00:11:51]Its light had been traveling for 13.5 billion years. It existed just 280 million years after the Big Bang. It is more luminous, more chemically enriched, and more compact than any model predicted at that epoch. Its effective radius is about 241 lighty years. That is roughly 1% the size of the Milky Way.

[00:12:13]In every Hubble image ever taken of that region of sky, it was invisible, pure, black, nothing there. It was always there. The team that confirmed it was led by Rohan Naidu at MIT. They had named their survey program Mirage or Miracle because they needed to know whether these impossibly bright early galaxies were real objects or artifacts of the instruments. The answer came back, not mirages. When they surveyed their entire observing field and counted how many galaxies of that brightness existed at red shifts of 14 to 15, the number was 182 times larger than the consensus models predicted before web launched. Not 10% more, not double, 182 times. Naidu said it directly. With web, we are able to see farther than humans ever have before. and it looks nothing like what we predicted, which is both challenging and exciting. A colleague on the team, Jacob Shen, used a phrase that had been circulating through the field for months. There is a growing chasm between theory and observation related to the early universe. And that chasm was about to get wider because the impossible galaxies were only part of the story. The little red dots, the ones scattered through almost every JWST image, were the other part. Calling something very red in astronomy means the object emits most of its light at long wavelengths in the infrared. Hubble cannot observe those wavelengths. Web was designed specifically to reach them, which is why the little red dots appeared the moment web turned on and why Hubble had missed them entirely. The first attempts to explain them focused on what astronomers already knew. If they were galaxies, they would need to be packed with stars at impossible densities. If they were active galactic nuclei powered by super massive black holes feeding on matter, they would need to produce X-rays. They do not. They would need to match the spectral signatures of known dust reddened black holes. They do not. and the black holes required would need to be enormously massive relative to the tiny galaxies hosting them, some of which are a 100 times smaller than the Milky Way.

[00:14:32]Despite the disagreements, everyone agreed on one thing. More data was needed. The initial JWST findings offered images, but understanding the physics required spectra, detailed breakdowns of how much light these objects emit at different wavelengths.

[00:14:48]The data came in January 2026 and it came from multiple directions simultaneously.

[00:14:56]A team at the University of Copenhagen led by Professor Darak Watson published in Nature. Their conclusion was that the little red dots are young black holes roughly 100 times less massive than previous estimates suggested hidden inside dense cocoons of ionized gas. The gas absorbs the radiation from the black hole and remits it, giving the objects their distinctive red color. Watson described it plainly. The little red dots are young black holes enshrouded in a cocoon of gas which they are consuming in order to grow larger. This process generates enormous heat which shines through the cocoon. This radiation through the cocoon is what gives little red dots their unique red color. But that same month, a completely different team proposed something more radical.

[00:15:48]Researchers at Harvard Center for Astrophysics suggested that the little red dots might not be black holes at all. They might be super massive stars, gigantic, short-lived stellar objects from the first generation of stars in the universe. Stars so enormous that they could only have existed during the cosmic dawn. and their collapse could have directly seeded the super massive black holes we find at the centers of galaxies today. The team published in the astrophysical journal showing that a simplified model of super massive ancient stars matches the spectral signatures of little red dots. And then came the dark stars. In January 2026, a study published in the journal Universe, proposed that three of JWST's biggest mysteries, the little red dots, a class of ultra bright early galaxies called blue monsters, and the impossibly massive black holes found in the young universe, could all be explained by a single theoretical object, dark stars.

[00:16:51]hypothetical stars powered not by nuclear fusion but by the annihilation of dark matter particles in their cores.

[00:16:58]Dark stars are not a fringe idea. They were first proposed as a serious theoretical framework over a decade ago.

[00:17:05]The concept is straightforward. In the early universe, before normal stars could easily ignite because gas clouds were too hot and lacked efficient cooling mechanisms for collapse. Dark matter was far more concentrated. If dark matter particles collided and annihilated inside a gas cloud, the energy released could have sustained enormous stellar objects, objects that could grow to be millions of times the mass of our sun. Objects that would have been extraordinarily luminous and when they finally collapsed, they could have directly produced super massive black holes without needing billions of years of conventional growth. The team led by Cosmin Ilier at Colgate University described in detail how dark stars could account for the properties of blue monster galaxies whose extreme brightness implies impossibly dense star packing that no current model can explain. They could account for the little red dots whose compactness and spectral features match what you would expect from objects transitioning from dark star to collapsed remnant. and they could account for the early super massive black holes whose existence is difficult to explain through nuclearpowered stars alone. Two separate earlier studies published in the proceedings of the National Academy of Sciences in 2023 and 2025 had identified phototric and spectroscopic dark star candidates in JWST data. Distinctive helium absorption features in the spectra of two high red shift objects J A D SG140 and J A D SGS130 matched what dark star models predict.

[00:18:47]Dark stars have not been confirmed through direct observation, but the circumstantial evidence is accumulating.

[00:18:55]And if they existed, they would stitch together three of the biggest puzzles in modern cosmology into a single coherent story. While the little red dot debate was consuming the theoretical community, a completely different team dropped a completely different result. Astronomers at Texas A and M University using JWST observations found something in the early universe that was not supposed to be there. Not a single impossible galaxy. A system. Five galaxies locked in the act of colliding, merging, smashing into each other, surrounded by a halo of gas rich in oxygen. The system was dubbed JWST's Quintet, published in Nature Astronomy in January 2026. And the problem was the timing. This was happening 800 million years after the Big Bang. At that point in cosmic history, galaxies were supposed to be small, isolated, barely formed. They were not supposed to be interacting.

[00:19:54]They were certainly not supposed to be scattering heavy elements like oxygen into the surrounding void. Oxygen is only produced inside stars through nuclear fusion. To find it outside the galaxies, floating in the surrounding gas, means that something had to remove it, rip it out. The team's analysis showed that gravitational interactions during the merger were driving the enrichment. The collision itself was stripping elements forged in stellar cores and flinging them into intergalactic space at a time when the models said none of this should be happening. The lead researcher Casey Papovich, professor of physics and astronomy at Texas&M, said it directly.

[00:20:36]This discovery tells us our theories of how galaxies assemble and how quickly they do so need to be updated to match reality. This finding helps explain something else that had been nagging astronomers. Why does web keep finding massive galaxies that appear largely inactive just a few billion years later?

[00:20:57]Galaxies that have already finished forming stars that have gone quiet as though they burned through their fuel early and shut down. If systems like JWST's Quintet merged rapidly and exhausted their gas supplies in the process, they could evolve into exactly those massive silent galaxies observed at later epochs. The violence of the early universe was not random. It was productive. Galaxies were assembling faster, colliding earlier, and burning brighter than anyone predicted. And this was not the only structural surprise. In February 2026, a team at the University of Pittsburgh identified what may be one of the earliest barred spiral galaxies ever observed. A galaxy called Cosmos 74706 dating back 11.5 billion years containing a stellar bar stretching across its center. A structure similar to the one in our own Milky Way. 2 billion years after the Big Bang, barred spiral galaxies were not supposed to form that early. The gravitational dynamics required to create a stellar bar were thought to require a more evolved, more settled galactic environment. Meanwhile, a separate team at the University of Waterloo identified the most distant jellyfish galaxy ever observed. A cosmic oddity streaming long tentacle-like trails of gas and newborn stars as it races through a dense galaxy cluster. Its trailing tentacles of gas, stripped away by the pressure of the surrounding cluster environment, offer a rare glimpse into how galaxies were reshaped by their surroundings in the ancient universe. And it raises questions about how many galaxies in clusters were killed by this process, having their star forming gas ripped away entirely. And in January 2026, NASA announced another unexpected finding from web. The dwarf galaxy Sexton's A, one of the most chemically primitive galaxies near the Milky Way, was producing types of dust that should not have been possible in an environment so starved of heavy elements, iron dust, silicon carbide, complex carbon-based molecules called polycyclic aromatic hydrocarbons, all forming in a galaxy with only 3 to 7% of the sun's metal content. The implication for the early universe was direct. If a primitive galaxy near us can forge these building blocks with almost nothing to work with, then the first galaxies after the Big Bang could have been producing the ingredients for planets far earlier than anyone assumed. The early universe was not settled, but it was building things that look like they belong in a settled universe. And that contradiction is at the heart of everything JWST is revealing. So here is where we stand.

[00:23:47]The Cosmos Web Survey cataloged nearly 800,000 galaxies spanning almost all of cosmic time. 10 times more early galaxies than expected. Super massive black holes that Hubble could not see.

[00:23:59]Galaxies interacting and merging billions of years before the models predicted. And entirely new classes of objects, little red dots, blue monsters, possible dark stars that no one predicted at all. The standard model of cosmology, Lambda CDM, is not broken.

[00:24:16]Most cosmologists are careful to say that the basic framework dark energy plus cold dark matter driving the expansion and structure of the universe still holds. Hubble data from moderate distances still matches beautifully. The cosmic microwave background still fits.

[00:24:33]But the early universe is where the cracks are showing. The number of bright galaxies at red shifts above 10 is significantly higher than preJWST models predicted. There are more very bright, very massive early galaxies than theory allows for the time available after the Big Bang. Some researchers have proposed that the early universe simply had more efficient star formation. That gas cooled faster, collapsed more readily, and converted into stars at higher rates than the modern universe does. Others suggest that the initial conditions after the Big Bang produced more concentrated matter over densities than previously assumed, giving galaxies a head start. A smaller group has questioned whether the standard model needs more fundamental revision, whether the parameters themselves need adjusting. Those are three different conversations and they are all still happening. None of them are settled.

[00:25:28]What is not in dispute is that JWST has forced a productive rethinking of how and when the first massive galaxies assembled. The debate is driving some of the most vigorous theoretical activity in cosmology in decades. Papers are being published weekly. Models are being revised monthly. The field is alive in a way it has not been since the discovery of the accelerating expansion of the universe in 1998. And here is the part that matters beyond the physics. For most of human history, we looked at the night sky and saw darkness. And we built an entire narrative around it. The narrative was isolation, emptiness, a vast and indifferent universe in which we are a rare and fragile exception.

[00:26:13]That story shaped religions. It shaped philosophy. It shaped how we think about our own significance. In 1990, Carl Sean asked NASA to turn the Voyager 1 spacecraft around at 6 billion kilometers from Earth and take a photograph. Earth appeared as a single pale pixel suspended in a scattered beam of sunlight. Sean pointed to that image and told the world to consider how small and alone we are. The emptiness of the background made the point land, but the emptiness was wrong. The black background behind that pale blue dot is not empty. It is filled with trillions of galaxies, each containing hundreds of billions of stars. The void that made us feel small and alone was never a void.

[00:26:57]It was a limitation of the detector. We are the detector. Your eyes evolved to respond to wavelengths between roughly 300 and 80 and 700 nanome. That is your entire window on reality. And the most ancient, most distant, most numerous light in the universe falls completely outside of it. Your retinas register an ocean of 1 3 billiony year old light as empty black space. Not because the light is not there. It is there. It fills every cubic cm of the space you are sitting in right now. Because you are the wrong instrument. If you could replace your retinas with infrared sensors, the night sky would not be dark. It would be a continuous wall of light from horizon to horizon. Every direction saturated, every patch occupied. Two trillion galaxies, their light stretched across billions of years of cosmic expansion, all arriving at your position simultaneously. The sky would look nothing like what you see tonight. It would look like the inside of a furnace. Three mechanisms, three independent processes, all producing the same output. The universe is 13.8 billion years old, creating a hard horizon beyond which light has not had time to arrive. The expansion of space is stretching photons from distant galaxies into wavelengths our eyes cannot detect. And the overwhelming majority of galaxies are too faint, too small, and too ancient to register on any instrument we had ever built before web. All conspiring to produce a sky that appears empty to a species whose eyes evolve to find fruit and avoid predators. Not to detect one 3 billion year old infrared radiation from the edge of the observable universe. The darkness between the stars has never been a sign of cosmic loneliness. It is a map. A map of the age of the universe, the expansion of space, and the sheer number of objects too faint for biology to detect. And the map says something that 300,000 years of anatomically modern human existence could not have revealed on its own. You are not alone in a void. You are surrounded. You have always been surrounded by trillions of galaxies whose light has been traveling toward you since before Earth existed, arriving every second, filling every cubic cm of space you will ever occupy.

[00:29:30]You just could not see it until now. Web is now targeting red shifts above 15 objects from the first 270 million years of the universe, the absolute frontier of its detection capability. Any galaxy found at those distances would represent the earliest structure ever observed.

[00:29:52]Constraints on the very first stages of cosmic assembly. Programs are underway to map how the epoch of reionization progressed. The period when ultraviolet radiation from the first stars and galaxies ionized the neutral hydrogen gas filling the universe, making it transparent to light for the first time.

[00:30:12]Web is providing the first direct mapping of that process. Early results suggest that bright galaxies played a more important role than expected.

[00:30:21]Another surprise in a field that has had nothing but surprises for 4 years. The Nancy Grace Roman Space Telescope is targeted for launch in late 2026 or 2027.

[00:30:34]It will survey much wider fields of sky than Web can cover, identifying thousands of candidate galaxies that web can then follow up with detailed spectroscopic confirmation. The partnership between wide survey and deep pointing is about to enter its most productive era in the history of astronomy and web itself may break its own record again before Roman launches.

[00:30:59]The team that confirmed MOME Z14 wrote in their paper that previously unimaginable red shifts approaching the era of the very first stars no longer seem far away. 4 years in, 20 more to go, and the early universe has already proven to be far stranger, far more violent, and far more creative than the models predicted. The little red dots are still being debated. The blue monsters are still being cataloged. The dark star hypothesis is still being tested. The chasm between theory and observation is still growing. And with every new image web sends back from its position 1.5 million km from Earth, the universe reveals another piece of itself that we were never supposed to see. Not because it was hidden, because we did not have the right eyes. Now we do.

[00:31:49]Everything you just heard was measured.

[00:31:51]Every galaxy was confirmed. Every number was published in peer-reviewed journals by teams at MIT, Texas&M, the University of Copenhagen, Penn State, Harvard, and dozens of other institutions across the world. The James Web Space Telescope is not speculating about the early universe. It is photographing it, and the photographs do not match the predictions. If you want to see what Webb found hiding in the darkness of our own galaxy and why some of it is equally hard to explain, that video is on screen now. Subscribe so you do not miss what comes next because the next record is coming and it will not take