Sir David Attenborough :The Disturbing Truth About Trillions of Planets — No BS

Right now, as you watch this, there are more planets in our galaxy than there are stars. Not a guess. Not a vibe. A measurement. A 2012 gravitational microlensing survey led by The Planet Collaboration, published in the journal Nature, concluded that on average every single star in the Milky Way hosts at least one planet. And that estimate has only climbed since. The current working number, based on Kepler data and refined by the NASA Exoplanet Archive through 2024, sits at somewhere between 100 and 400 billion planets in our galaxy alone. That is, the Milky Way. One galaxy. There are an estimated two trillion galaxies in the observable universe. Multiply them out, and you arrive at a number with no emotional weight attached to it. Because human brains do not process exponents. Roughly 10 to the power of 24 planets. A 1 followed by 24 zeros. That was January 2024. By late 2025, the James Webb Space Telescope had identified planetary candidates orbiting stars so old and so distant that the light reaching us left before our solar system existed. And the part everybody misses. The overwhelming majority of those worlds are invisible to us. Not hidden behind clouds, not too far to reach. Invisible by the physics of how we look. What that means for how we understand reality, and what recent data is quietly suggesting about the planets between the stars, is the part nobody talks about. Mount Wilson Observatory, California, 1920. For most of human history, astronomers assumed our solar system was unusual. A star with planets in tidy orbits around it felt like a special arrangement. Astronomers called it the solar nebula hypothesis, and for decades it was treated as plausible but unconfirmed. We had a sample size of one. We knew planets could exist because we were standing on one. Beyond that, everything was theory. The first confirmed detection of a planet orbiting another main-sequence star came in October 1995. Michel Mayor and Didier Queloz, working at the Haute-Provence Observatory in France, announced the discovery of 51 Pegasi b, a gas giant, larger than Jupiter, orbiting its parent star every four days at a distance so close that its surface temperature runs above a thousand degrees Celsius. Nothing in the solar nebula model predicted anything like it. The planet existed in a region where classical physics said it could not have formed. One planet. That was where the count started. Mayor and Queloz won the Nobel Prize in 2019 for that discovery, but by then the number had already collapsed under its own momentum. In 2009, NASA launched the Kepler Space Telescope with one job. Stare at a single patch of sky, constantly, for years. Wait for stars to dim. When a planet passes between its parent star and the telescope, the star's apparent brightness drops by a tiny fraction of a percent. Measure the drop. Time the drop. Work out the planet's size and orbit from the numbers. Kepler stared at 150,000 stars. By the time its mission officially ended in 2018, it had confirmed the existence of over 2,600 exoplanets in that one small patch of sky. That is the critical detail. That is not the galaxy. That is one rectangular window, smaller than your hand, held at arm's length, pointed at a random neighborhood of the constellation Cygnus. Scale that density up to the full sky. Adjust for the fact that Kepler could only detect planets whose orbits line up precisely between us and the star, which is a small fraction of all planets that actually exist. Factor in the faint ones and the distant ones that slip below the detection floor. The math was done, redone, and then done again by independent teams at Caltech, Penn State, and UC Berkeley between 2013 and 2021. Every time, the number came out higher than the last. The current best estimate is that the Milky Way contains roughly 100 billion stars and at least that many planets orbiting them. Some stars have one, some have none. Some, like TRAPPIST-1, have seven packed tightly around a dim red dwarf 40 light-years from Earth, three of them sitting in the zone where liquid water could exist on the surface. The TRAPPIST-1 system was confirmed in a 2017 Nature paper by a team led by Michael Gillon at the University of Liège. That single star, a tiny ultra-cool dwarf most people have never heard of, hosts more potentially habitable worlds than our entire solar system. And this is where the uncomfortable part starts. Everything described so far assumes one thing. Planets orbit stars. The whole detection framework, transit, photometry, radial velocity, microlensing, all of it is built on the assumption that if you want to find a planet, you look at a star and watch what it does. But in 2011, a team of Japanese and New Zealand astronomers doing microlensing surveys at the Mount John Observatory noticed something that should not have been there. Gravitational signatures from objects with planetary mass. The problem was, the signatures had no host stars attached to them. Nothing was anchoring these worlds to any solar system. They were drifting through interstellar space entirely alone. Rogue planets. That was the name that stuck. The working theory, supported by computer simulations at MIT and the Max Planck Institute, is that during the chaotic early years of any solar system, gravitational interactions between forming planets frequently fling some of them out. A young gas giant gets too close to another gas giant, the orbital mechanics go unstable, and one of them gets ejected at speeds high enough to escape the star's gravity permanently. Once out, it coasts into the interstellar medium and keeps going forever. With nothing to orbit, its surface goes dark, its atmosphere freezes, whatever oceans it had turn to solid ice. The planet becomes a frozen wanderer travelling through galactic space at tens of kilometres per second. And unless something else pulls it back into a system, it never gets captured again. The initial 2011 estimate, based on the microlensing data alone, suggested that for every star in the Milky Way, there might be roughly two rogue planets out there in the dark. Then Webb went to work. In late 2023, a team using Webb's infrared instruments surveyed the trapezium cluster inside the Orion Nebula, a dense stellar nursery about 1,300 light years from Earth. What they found, and published in a 2024 paper in the Astronomical Journal, upended the rogue planet count. They identified over 500 free-floating Jupiter mass objects in that one region alone, and several dozen of those objects appeared to be orbiting each other in pairs, binary planets. Two giant worlds bound only to each other, with no parent star, wandering the void in a slow, eternal dance. The discovery was so unexpected that the theorists had to invent a new name. Jumboes. Jupiter mass binary objects. No existing model of planetary formation predicts them. Conventional theory says you cannot form two Jupiter-sized bodies together in isolation without a host star's gravity well to shape the collapse. Yet, there they are. Roughly 40 pairs of them, confirmed in infrared imaging, orbiting each other in a region where standard cosmology said they should not exist. And the implication of that survey, if you extrapolate across the galaxy, is that the number of rogue planets in the Milky Way is not 2 per star. It might be closer to 20 per star. Or more. Which means the total planetary population of our galaxy is not 100 billion. It is potentially several trillion. And the vast majority of those worlds are not orbiting anything. They are not near any light source. They emit almost no detectable radiation of their own. They are essentially undetectable except by the faint gravitational lensing they create as they drift in front of distant stars for brief moments every few thousand years. Pay attention to this next detail. Because it explains everything. These rogue worlds are not rare. They are not exotic. They are, statistically, the dominant planetary population of the galaxy. Every time you look up at the night sky you are looking through a fog of dark, wandering worlds. None of them are visible. All of them are there. So, if there are trillions of planets out there, why does the galaxy look so empty? The answer is deceptively brutal. Every method we have for detecting planets requires the planet to interact with something we can already see. And most planets refuse to cooperate. Method 1 Transit Photometry This is how Kepler found its thousands of planets. The planet crosses in front of its star and dims the light. It works. But only for planets whose orbital plane happens to line up almost perfectly with our line of sight. For random orientations, that alignment happens less than 1% of the time for close-in planets and far less for distant ones. For rogue planets with no star to cross, transit photometry does nothing. The method is blind to them by definition. Method 2 Radial Velocity A planet's gravity pulls on its star, causing a tiny wobble. Sensitive spectroscopy can detect that wobble as a shift in the star's spectral lines. This works, but it is strongly biased toward massive planets in close orbits, because those produce the largest wobbles. Earth-mass planets in Earth-like orbits create wobbles so small they sit just at the edge of what current technology can resolve. Again, useless for rogue planets. No star, no wobble. Method 3 Direct Imaging We photograph the planet itself. Sounds simple, works almost never. A star outshines its planets by factors of a billion or more in visible light. Imagine trying to photograph a firefly next to a stadium floodlight from 10 miles away. Coronagraphs and star shades can help by blocking the starlight, but even at their best, direct imaging only works for young, hot, giant planets in wide orbits. Earth analogues remain out of reach, and a rogue planet emits almost no light of its own, so there is nothing to image. Method 4 Gravitational Microlensing This is the one that actually sees the invisible ones. When a massive object passes in front of a distant background star, its gravity bends the background starlight, briefly magnifying it in a predictable pattern. The mass does not have to be glowing, it just has to be massive. Microlensing has detected rogue planets, and in principle it can detect any planet-mass object anywhere in the galaxy. The catch is that microlensing events are one-shot and unrepeatable. The alignment has to be essentially perfect. The event lasts for hours to days, and then it is over. You cannot go back and look at it again. The planet continues its journey into the dark, and unless another background star aligns behind it a few thousand years from now, it is never seen again. That is not a typo, by the way. Every rogue planet we have detected has been seen exactly once in human history, during a single brief alignment, and then lost forever. In practical terms, the sky is filled with planets, and we catch occasional glimpses of them through a cosmic accident. Our entire catalogue of confirmed exoplanets, roughly 5,700 worlds, as of the NASA exoplanet archives last update, is a statistical artifact of what our instruments can see. It is not a map of the galaxy, it is a map of our instruments' blind spots, inverted. If you are wondering how that is even legal, so is everyone else. The number of planets we have confirmed represents a fraction of the true population, so small it is hard to hold in your head. If the galaxy contains trillions of planets, and we have catalogued fewer than 6,000, we have seen roughly one out of every several hundred million. Up to this point, we have been counting. Now we need to look at what is actually out there, because the catalogue of confirmed exoplanets includes objects that none of our formation models predicted would be possible. HD 189733b A gas giant 64 light-years from Earth. From a distance, it appears cobalt blue, which sounds almost Earth-like, until you understand why. The blue colour comes from silicate particles in its atmosphere. Glass. The planet has clouds of molten glass, and because its atmospheric temperature runs around 1,000 degrees Celsius, and because wind speeds on that world reach 7,000 kilometres per hour, the glass does not fall as droplets. It gets blown sideways at twice the speed of sound. It rains sideways glass. That is a planet that actually exists. Confirmed. Spectroscopically characterised. 55 Cancri A rocky super-Earth, about 40 light-years away. It orbits its star so closely that a year on the planet lasts 18 hours. Surface temperatures on the day side exceed 2,400 degrees Celsius, hot enough that the crust is partially molten. Early analysis by the team at Yale, later complicated by follow-up observations, suggested that the interior could contain enough carbon under enough pressure that layers of it might be crystallised into diamond. That is not science-fiction packaging. That is a peer-reviewed hypothesis built on the planet's measured density and its star's unusually high carbon-to-oxygen ratio. The interpretation has been disputed. The planet has not. KELT-9b An ultra-hot Jupiter whose day-side temperature runs hotter than the surface of most stars. Roughly 4,300 degrees Celsius. At those temperatures, the planet's atmosphere does not just get stripped away by stellar radiation. It gets torn apart at the molecular level. Water molecules cannot exist there. They dissociate into atoms on contact with the heat. KELT-9b is a planet where the concept of chemistry partially breaks down on its sunlit face. And if that sounds bad, keep watching. The catalogue includes planets orbiting pulsars, meaning rocky worlds that survive the supernova explosion that created their host. It includes worlds in binary star systems with two sunsets. Worlds with retrograde orbits, moving opposite to their star's rotation. Which no standard formation model predicts. Worlds that have been partially stripped down to their metallic cores by the radiation of their parent stars. Worlds that escaped one stellar system and were captured by another. An event so rare, it was considered essentially impossible until the gravitational signature of one was measured by the Gaia mission in 2022. Every model of planetary formation we have has exceptions in the confirmed catalogue. Not edge cases, not statistical flukes, real planets, real spectroscopic confirmation, real gravitational signatures, and they all imply that the list of ways a world can form is much longer than we thought. The key word in that sentence is confirmed. These are the ones we have detected and analysed. If we are seeing roughly one in every several hundred million planets, the catalogue of confirmed impossibilities is a tiny window into a population of planets that is almost certainly stranger by orders of magnitude than anything we have yet named. That is the kind of number that makes accountants cry. If this is the kind of content you actually want more of, hit the subscribe button. The next video in this series breaks down what the Webb telescope found in the atmosphere of a planet that should not have an atmosphere, and why the data is making astrophysicists quietly revise how they think terrestrial worlds form. While all of this was happening, something else was brewing in the data. The number of planets sitting in the habitable zones of their stars. The habitable zone is a narrow orbital region where a planet's surface temperature could, in principle, support liquid water. Not guarantee life. Not even strongly imply it. Just allow the basic chemistry to remain in a liquid state long enough for something interesting to potentially happen. Earth sits inside our sun's habitable zone. Venus is just inside the inner edge, and is the reason we know what runaway greenhouse collapse looks like. Mars sits near the outer edge, and is the reason we know about thin atmospheres and frozen oceans. By 2024, analysis of the Kepler dataset by Natalie Battaglia and her team, combined with refined radial velocity surveys, had converged on a working estimate. Roughly 20% of sun-like stars in the Milky Way host a rocky planet, roughly Earth-size, orbiting within the habitable zone. 20%. Out of 100 billion stars in our galaxy, a meaningful fraction of which are sun-like that is at least several billion potentially habitable worlds in our galaxy, which is one of two trillion galaxies. The arithmetic on the full universe produces numbers that exceed the capacity of human intuition to even feel. Most people never heard about this part. The habitable zone math is not a fringe claim. It is the mainstream, cautious, peer-reviewed, heavily conservative estimate, and it only accounts for habitability under Earth-like rules. It does not include the moons of gas giants that might hold subsurface oceans. It does not include worlds where the chemistry of life could run on something other than liquid water. It does not include any of the trillions of rogue planets floating in the dark, some of which under internal radiogenic heating, could plausibly maintain subsurface liquid water for billions of years with no star at all. This is not speculation. It was modeled in 2019 by a team at the University of Chicago. They ran the thermodynamics on a Mars-sized rogue planet with a hydrogen-dominated atmosphere and concluded that such a world could, in principle, retain liquid water at the surface for over a billion years, warmed entirely by the planet's internal heat and a dense blanket of hydrogen trapping it in. No star required. No daylight. Just a dark wandering ocean under a hydrogen sky, traveling through interstellar space in total silence, potentially for the age of the universe. If even a fraction of a percent of rogue planets support that configuration, the number of potentially habitable environments in our galaxy is not in the billions, it is in the trillions. The old picture, which held habitability to be a rare property of specific planets in narrow orbital bands around specific kinds of stars, has been steadily eroded by the data for two decades. The current picture is that habitability might be closer to the default state than the exception, and that most of it is happening in places we cannot see. You would think someone would have said something. Nobody did. The reason nobody said it loudly is that, until recently, the tools to look anywhere else did not exist. Every major habitability search, SETI, the Breakthrough Listen Initiative, the TESS mission, the characterization work on TRAPPIST-1, has focused on nearby bright stars with detectable planets. This is reasonable. It is where the signal-to-noise is best, but it is also a streetlight effect. We are looking where the light is. The dark is where most of the galaxy lives. Here is the part that has nothing to do with physics, and it is the part that is hard to put down. For roughly 300,000 years of anatomically modern human existence, we looked up at the stars and counted a few thousand. That was the catalogue. Ptolemy had about a thousand stars in his Almagest. Hipparchus, working in the 2nd century BCE, had around 850. Galileo, pointing the first telescope at the sky in 1609, added thousands more, but the total never ran into the millions, because the eye and the lens are finite instruments. During that entire period, we assumed the planets we could see were roughly all the planets there were. 6, then 7 with Uranus, then 8 with Neptune, then briefly 9 with Pluto, now 8 again. 8 was the number for most of my lifetime. 8 was the number in the textbooks. 8 was the number that felt like a complete accounting of what a solar system contained. 8. Out of roughly 10 to the 24th. The ratio is a rounding error. Our entire species-level sense of how many planets the universe contains was built on a sample size so small it did not even qualify as a sample. And when the instruments improved, the count did not go up by a factor of 10 or 100, it went up by a factor of roughly 10 to the 23rd. That is not a revision. That is an entire category of reality we did not know we were missing. And this is the part that stays with you if you think about it too long. The worlds are not hypothetical. They are there. They have always been there. Every night you have ever looked up at the sky, there have been rogue planets passing silently between you and the stars you were seeing. There have been ocean worlds in the dark. There have been rocky worlds in habitable zones. There are surfaces warmed by starlight, potentially with weather, potentially with chemistry, that has been running for longer than our planet has existed. All of them, all of them have been invisible to you. Not because they were hiding, because you are small and your eyes are small, and the universe does not calibrate itself to your sensory hardware. Every patch of night you have ever called empty is dense with worlds. Every rogue planet is a separate geological history, a separate chemical record, a separate atmosphere, a separate temperature profile, a separate set of possibilities for what could happen on a world over four billion years of isolation. Multiply that out across trillions of cases and you are looking at a number of planetary histories so large that human civilization's entire recorded history, every war, every language, every moment of grief and joy, would fit comfortably inside the geological story of a single random rogue world for the duration of a single rainy afternoon on its invisible surface. The story that the universe is empty, that life is rare, that Earth is an outlier, was always a hypothesis based on an almost empty catalogue. The catalogue is no longer empty. And what it is showing us quietly, paper by paper, is that the emptiness was in us. We have not answered the question of whether any of those trillions of worlds host life. The honest answer, as of late 2025, is that we do not know. Webb has characterized the atmospheres of dozens of exoplanets. None of them have produced an unambiguous biosignature. The data is suggestive in a few cases, including the ongoing debate around K2-18b's atmospheric composition. But suggestive is not confirmation, and the scientists closest to the data are openly careful about that distinction. What we do know is that the arena is larger than we thought. The number of possible stages for whatever complex chemistry can do anywhere in the galaxy is measured in the trillions. The number of those stages we have ever managed to look at carefully is measured in the thousands. The ratio between those two numbers is the gap between what we have seen and what is actually there. And the gap is closing, slowly, one paper at a time, through instruments that keep detecting worlds that our models said should not exist. The next time you look up at the night sky, remember what is actually happening above you. Trillions of planets, the vast majority not orbiting anything, some of them warm under hydrogen atmospheres, some of them drifting in eternal darkness, some of them statistically certain to be in configurations that could support the chemistry of life, based only on the cold math of the numbers involved. You will never see them, no human eye will ever see them directly. But the data says they are there, and the data is getting better every year. If you want to see what Webb found inside the atmosphere of a world orbiting a dying red dwarf, and why its chemical signature is making physicists quietly rewrite the textbook on how terrestrial worlds form, that video is on screen now. Subscribe, so you do not miss what comes next.