- 1 day ago
Nobel Prize winner Adam Riess says we may have "misunderstood the universe." The James Webb Space Telescope was supposed to settle a growing contradiction in cosmology. Instead, it made it worse.
Two methods of measuring how fast the universe is expanding give two different answers. One says 67. The other says 73. The gap is too large to ignore, and after Webb confirmed the measurements with unprecedented precision, the possibility of a simple error has been ruled out. Something in our understanding of the cosmos is either missing or wrong.
This video covers the Hubble tension, the impossible early galaxies Webb discovered, the rival teams fighting over the data, the mysterious "little red dots" that fooled astronomers, and why some physicists now believe we may need entirely new physics to explain what we're seeing.
Two methods of measuring how fast the universe is expanding give two different answers. One says 67. The other says 73. The gap is too large to ignore, and after Webb confirmed the measurements with unprecedented precision, the possibility of a simple error has been ruled out. Something in our understanding of the cosmos is either missing or wrong.
This video covers the Hubble tension, the impossible early galaxies Webb discovered, the rival teams fighting over the data, the mysterious "little red dots" that fooled astronomers, and why some physicists now believe we may need entirely new physics to explain what we're seeing.
Category
📚
LearningTranscript
00:00We have misunderstood the universe.
00:03Those are not my words.
00:05Those are the words of Adam Rees,
00:07a Nobel Prize-winning physicist at Johns Hopkins University.
00:11A man who has spent his career measuring how fast the cosmos is flying apart.
00:18And he's not being dramatic.
00:21He's reading his own data and telling us plainly
00:24that something in our model of reality does not add up.
00:28For nearly a century, cosmology has rested on a single elegant framework.
00:35We call it the Standard Model,
00:37a set of equations and assumptions that describes everything
00:41from the first fraction of a second after the Big Bang.
00:45It has worked beautifully for decades,
00:48explaining observation after observation
00:51with an almost suspicious level of precision.
00:54And then the James Webb Space Telescope turned on.
01:00What Webb found did not confirm our understanding.
01:05It complicated it.
01:07In some cases, it contradicted it.
01:10And the deeper we've looked with this telescope,
01:13the wider the cracks have become.
01:16The trouble starts with a number.
01:19It's called the Hubble Constant.
01:22And it measures the rate at which the universe is expanding right now.
01:27Not in the past.
01:28Not in some theoretical future.
01:31Right now.
01:33Today.
01:33At this moment.
01:34You'd think a single number like that would be easy to pin down.
01:40Two methods, two teams, same universe, same answer.
01:46That's how science is supposed to work.
01:49It didn't.
01:51And this isn't the first time.
01:53The Hubble Constant has been causing arguments since the 1920s,
01:57when Edwin Hubble himself first calculated it.
02:02His original estimate was so far off
02:05that it made the universe appear younger than the Earth.
02:09For decades after that,
02:11two camps of astronomers fought bitterly
02:14over whether the true value was closer to 50 or 100.
02:19A disagreement so fierce,
02:21it split the field in half.
02:24By the early 2000s,
02:26the number had finally converged.
02:29Everyone agreed it was somewhere around 70,
02:32give or take.
02:34The war was over.
02:35The measurement problem was solved.
02:39Except it wasn't.
02:41Because as instruments got more precise,
02:44a new split emerged.
02:46Smaller than the old one,
02:48but far more stubborn.
02:50The first method looks backward.
02:53Way backward.
02:55The European Space Agency's Planck satellite
02:58spent about four and a half years
03:00mapping the cosmic microwave background.
03:03The faint afterglow of radiation
03:06left over from when the universe
03:08was about 380,000 years old.
03:12From that map,
03:14physicists extracted the conditions
03:16of the early universe,
03:18plugged them into the standard model,
03:20and ran the equations forward
03:22nearly 14 billion years
03:25to predict how fast the universe
03:27should be expanding today.
03:29The answer they got
03:31was about 67 kilometers per second
03:34per megaparsec.
03:36A megaparsec, by the way,
03:38is 3.26 million light years.
03:42So for every 3.26 million light years
03:46you look outward,
03:47galaxies are moving away from us
03:5067 kilometers per second faster.
03:54Clean number.
03:56Tiny margin of error.
03:57Confidence level through the roof.
04:00The second method is more direct.
04:04Instead of modeling the past
04:06and projecting forward,
04:07it measures the present.
04:09Adam Rees and his team,
04:11called SH0ES,
04:14use a technique
04:15known as the Cosmic Distance Ladder.
04:18They start by measuring the distance
04:21to nearby pulsating stars
04:23called Cepheid variables.
04:26Stars whose brightness fluctuates
04:28at a rate directly tied
04:30to how luminous they actually are.
04:34Measure the pulse,
04:35calculate the true brightness,
04:38compare it to how bright
04:39the star looks from Earth,
04:40and you've got a distance.
04:43From there,
04:44they step outward
04:45to galaxies containing both Cepheids
04:48and a specific type of exploding star
04:51called a Type 1a supernova.
04:55These supernovae are useful
04:57because they all explode
04:58with roughly the same peak brightness.
05:01So once you've calibrated them
05:03with Cepheids in the same galaxy,
05:05you can spot them in galaxies
05:07hundreds of millions of light years away
05:10and calculate how far those galaxies are.
05:14Combine that
05:15with how fast they're receding,
05:17and you get the Hubble constant.
05:20Reese's team got 73.
05:24Not 67.
05:2673.
05:286 kilometers per second
05:30per megaparsec
05:32faster than the standard model predicts.
05:35That might sound like a rounding error.
05:38It is not.
05:40In a field where measurements
05:42are routinely precise
05:44to within 1%,
05:46a gap of 8 or 9%
05:48between two independent methods
05:51is enormous.
05:53It's the difference
05:54between a universe
05:55that's about 13.8 billion years old
05:59and one that could be
06:01more than a billion years younger.
06:03It changes the size of the observable universe.
06:07It changes the rate
06:09at which galaxies are separating.
06:12It changes, potentially,
06:14the nature of the dark energy
06:16that's been pushing everything apart
06:18since the late 1990s.
06:23Think about what that means for a second.
06:26If the universe is expanding faster
06:28than our best model predicts,
06:30then everything downstream
06:32of that prediction shifts.
06:34The age of the universe gets younger.
06:37The distances between galaxy clusters change.
06:40The timeline for when the first stars ignited,
06:43when the first heavy elements
06:45were forged inside dying suns,
06:47when the conditions for planets and chemistry
06:50and eventually life became possible.
06:52All of it shifts.
06:54You're not just adjusting a dial on a dashboard.
06:57You're redrawing the entire history
07:00of everything that has ever existed.
07:02For a while,
07:03scientists assumed someone had made a mistake.
07:06The most popular theory was stellar crowding.
07:10Cepheid variables tend to live
07:11in the dense, dusty disks
07:13of younger galaxies.
07:14When Hubble looked at them,
07:16its cameras sometimes blended the light
07:18of nearby stars together,
07:20making the Cepheids appear brighter
07:22than they truly were.
07:24Brighter Cepheids mean shorter calculated distances.
07:28Shorter distances mean a higher Hubble constant.
07:32If you could clean up that blending,
07:34the argument went,
07:35the tension would relax.
07:37So Rees pointed the James Webb Space Telescope
07:41at the same stars.
07:42Webb orbits the sun,
07:44about 1.5 million kilometers from Earth,
07:48far beyond the interference of our planet's atmosphere
07:50and thermal radiation.
07:52Its 6.5 meter gold-plated mirror
07:56collects infrared light
07:58with a sensitivity
07:59that makes Hubble look like a pair of reading glasses.
08:02If any telescope could settle this,
08:05it was Webb.
08:06The expectation among many in the field
08:08was that Webb would gently deflate the tension,
08:11that its sharper vision would reveal the blending,
08:15correct the distances,
08:16and bring the Hubble constant back into line
08:19with the standard model.
08:20Some astronomers were already drafting the obituary for the crisis
08:24before the data came in.
08:26Webb's infrared cameras can cut through dust and crowding
08:30with a precision Hubble never had.
08:33If blending was the problem,
08:35Webb would show it.
08:37The Cepheids would look dimmer through Webb's eyes,
08:40the distances would stretch,
08:42and the Hubble constant would drop back towards 67.
08:46Crisis averted.
08:48That is not what happened.
08:51Webb confirmed Hubble's numbers.
08:53Not approximately,
08:55almost exactly.
08:56The Cepheids looked the same.
08:58The distances held.
09:00The expansion rate stayed at 73.
09:02In early 2024,
09:05the SH0ES team published a paper
09:09in the Astrophysical Journal Letters
09:11announcing that they had now spanned
09:14the full range of Hubble's Cepheid observations with Webb,
09:18extending all the way out to a galaxy called NGC 5468,
09:24about 130 million light years from Earth.
09:28A thousand Cepheids across five host galaxies
09:31of eight supernovae.
09:33The conclusion was unequivocal.
09:37Measurement error could be ruled out
09:38with very high confidence.
09:41Reese put it simply.
09:42With the errors eliminated,
09:44what's left is the possibility
09:46that we have misunderstood the universe.
09:49At a 2019 conference
09:51at the Kavli Institute for Theoretical Physics
09:53in California,
09:55Reese asked David Gross,
09:57a Nobel laureate in particle physics,
10:00whether the field should start
10:01calling this discrepancy a problem.
10:04Gross corrected him.
10:06He said they shouldn't call it a tension,
10:08and they shouldn't call it a problem.
10:10They should call it a crisis.
10:13By December 2024,
10:15the crisis had deepened.
10:17Reese's team published
10:18their largest web study yet,
10:20using data collected across the telescope's
10:23first two years of operation.
10:25They employed three separate methods
10:28to measure distances
10:29to supernova host galaxies,
10:31cross-checking Cepheids
10:33against carbon-rich stars,
10:35and a technique called
10:36the tip of the red giant branch.
10:39All three methods agreed
10:41with Hubble's original measurements.
10:44All three pointed to a universe
10:46expanding faster
10:47than the standard model allows.
10:49So either the model is wrong,
10:52or something is hiding inside the physics
10:54that we haven't accounted for.
10:56Not everyone agrees the tension is even real.
11:00Wendy Friedman,
11:01an astronomer at the University of Chicago,
11:04has spent years building
11:06an independent distance ladder
11:07that relies less heavily on Cepheids.
11:11Her team uses additional types of stars,
11:14red giants,
11:15and a class called JAGB stars
11:18that are found in less crowded regions of galaxies
11:21and are less likely to be affected
11:23by dust and blending.
11:26When she submitted her own web analysis
11:28in August 2024,
11:30her three methods gave different answers.
11:33The red giant and JAGB methods
11:36landed around 68 to 70,
11:39consistent with the standard model.
11:42Her Cepheid measurement came in higher,
11:45near 72,
11:46but with larger uncertainties.
11:48Her overall combined result was about 70,
11:52sitting in a kind of no-man's land
11:55between the two camps.
11:58Saul Perlmutter,
12:00a Nobel Prize-winning cosmologist
12:02at UC Berkeley,
12:03who reviewed Friedman's data
12:05before its release,
12:06noted that the results suggest
12:08there may be a tension
12:10within the star-based measurements themselves
12:12before you even compare them
12:14to the cosmic microwave background.
12:17That's a different kind of problem.
12:19It means the distance ladder,
12:21the tool astronomers have relied on
12:23for a century
12:24to map the nearby universe,
12:26might have unresolved issues
12:28at its foundation
12:29that no single telescope can fix.
12:33Rhys pushed back.
12:35He argued that Friedman's team
12:37had used a small
12:38and potentially unrepresentative subset
12:41of supernovae in their analysis,
12:43which could bias the results downward.
12:45The disagreement between the two camps
12:48has become one of the most closely watched rivalries
12:52in modern physics.
12:53Two teams,
12:54both brilliant,
12:55both careful,
12:56both looking at the same sky,
12:59arriving at different conclusions
13:00about what the sky is telling them.
13:03One possibility that keeps surfacing
13:06is called early dark energy,
13:08a hypothetical burst of extra expansion
13:11that occurred in the first few hundred thousand years
13:15after the Big Bang,
13:16then vanished.
13:17If something like that happened,
13:20it would change the conditions
13:21in the early universe
13:22just enough to shift
13:24the Planck prediction upward,
13:26closer to 73,
13:28without breaking the other things
13:30the standard model gets right.
13:32Mark Kamionkowski,
13:34a cosmologist at Johns Hopkins
13:36who has helped develop this idea,
13:38has compared it to the universe
13:40receiving an unexpected kick
13:42right after it was born,
13:44a nudge that would have rippled forward
13:47through 14 billion years of expansion,
13:50slightly altering the rate we measure today.
13:54Nobody knows if early dark energy is real.
13:57No one has detected it,
13:59but it's one of the few ideas
14:01that could solve the Hubble tension
14:03without demolishing everything else that works.
14:06The tension would be enough on its own,
14:09but Webb delivered a second blow.
14:12Almost as soon as the telescope started observing,
14:15it found galaxies in the early universe
14:18that had no business being there.
14:21Not small, dim, infant galaxies,
14:25which is what the standard model predicted
14:27at those distances.
14:29Massive ones.
14:30Galaxies seen as they existed
14:32just 500 to 700 million years
14:35after the Big Bang,
14:37some of them rivaling the Milky Way in mass.
14:40A galaxy that took our own neighborhood
14:4213 billion years to build
14:44had apparently assembled itself in under a billion.
14:49Mike Boylan Colchin,
14:50an astronomer at the University of Texas at Austin,
14:54ran the numbers.
14:55In a study published in Nature Astronomy,
14:58he showed that six of the earliest
15:00and most massive galaxy candidates observed by Webb
15:04pushed right up against the absolute limit
15:07of what the standard model allows.
15:10To form galaxies that heavy that fast,
15:13nearly every available atom
15:15in their surrounding dark matter halos
15:18would have had to convert into stars.
15:21In a normal universe,
15:23that conversion rate sits around 10%.
15:2690% of the gas just never gets around
15:29to forming anything.
15:30It's too hot
15:32or it gets blown out by radiation
15:34or it simply drifts.
15:37These galaxies seem to be running
15:39at 100% efficiency,
15:41like a factory that somehow turns
15:43every scrap of raw material
15:45into finished product with zero waste.
15:49Physics doesn't work that way,
15:51not in any model anyone had built.
15:54The initial reaction was, frankly, alarm.
15:57If galaxies this massive really formed this quickly,
16:02then either our models of galaxy formation
16:04are deeply incomplete
16:06or the standard model of cosmology itself
16:09needs revision.
16:11Some physicists proposed that new particles
16:14or forces might have existed in the early universe,
16:18accelerating star formation
16:20beyond anything current models predict.
16:23Then came a correction.
16:25Catherine Chwarowsky,
16:27a graduate student also at the University of Texas,
16:30led a study published in mid-2024
16:34that re-examined the data.
16:37She found that many of the apparently overmassive galaxies
16:40were being fooled by their own black holes.
16:44These galaxies, nicknamed little red dots
16:47because of their color and compact size,
16:50hosted black holes that were rapidly devouring gas.
16:54The friction from all that in-falling material
16:57heated it to extreme temperatures,
17:00and that heat radiated outward as light,
17:03lots of it,
17:04blending with the starlight of the galaxy itself.
17:08From billions of light-years away,
17:10through Webb's cameras,
17:12a galaxy with a hungry black hole at its center
17:15looks almost identical to a galaxy
17:18with ten times more stars.
17:20The telescope sees brightness,
17:23and brightness is usually a reliable proxy for mass.
17:28Usually.
17:29In these cases,
17:31the black hole was doing the work of billions of suns,
17:34inflating the galaxy's apparent size
17:37like a flashlight taped to a candle.
17:41Once those little red dots were removed from the sample,
17:44the remaining galaxies fit within the standard model's predictions.
17:49Steven Finkelstein,
17:50Chorofsky's advisor,
17:52was blunt about it.
17:53There is no crisis in terms of the standard cosmological model,
17:57he said.
17:58Not from the galaxies, at least.
18:00But the story didn't end there cleanly.
18:03Even after removing the little red dots,
18:05roughly twice as many massive galaxies remained
18:08as the standard model expects.
18:11Not enough to break the model,
18:13enough to bend it.
18:15Chorofsky suggested that stars may have formed faster
18:17in the early universe than they do today,
18:20that the denser conditions shortly after the Big Bang
18:23made it harder for gas to escape during star formation,
18:26allowing the process to accelerate.
18:28A reasonable explanation,
18:31but one that still requires adjustments
18:33to how we think galaxies grew up.
18:35And then, in early 2025,
18:38Webb found something else.
18:40A massive spiral galaxy,
18:42quickly nicknamed the Big Wheel,
18:44that existed within the first 2 billion years of the universe.
18:49Spiral structure takes time to develop.
18:52It requires relatively calm, orderly growth,
18:55not the chaotic merging you'd expect in a young universe.
18:58Themia Nanayakara,
19:00one of the astronomers who discovered it
19:02at Swinburne University of Technology,
19:04said the galaxy either assembled its mass
19:07in an unusually neat way,
19:09or it formed most of its stars in place,
19:12without the violent collisions
19:13that typically build large galaxies.
19:16Either explanation challenges conventional thinking.
19:19Meanwhile, a separate team at Cambridge
19:22used Webb to study over 250 galaxies from that same era
19:27and found the opposite pattern.
19:29Most of them were chaotic, turbulent,
19:32clumpy systems that hadn't settled into smooth, rotating disks.
19:36So the early universe contained both messy, adolescent galaxies
19:41and at least one remarkably mature spiral,
19:44coexisting in the same epoch.
19:47That's not a contradiction the standard model handles gracefully.
19:50And none of this addresses the elongated galaxies.
19:54A study published in Nature Astronomy in late 2025,
19:58led by Alvaro Pozzo and including researchers from MIT,
20:03Harvard and Taipei,
20:05found that many young galaxies observed by Webb
20:07appear strikingly stretched out,
20:10like cosmic cigars.
20:12These prolate shapes don't match predictions
20:14from the cold dark matter model,
20:16which is the backbone of standard cosmology.
20:19They do, however, match predictions from alternative models,
20:23warm dark matter and wave dark matter,
20:26in which the filaments of the large-scale structure
20:29of the universe are smoother,
20:31allowing gas and stars to flow along them
20:34in elongated streams.
20:36That distinction matters more than it sounds.
20:39Cold dark matter isn't just one ingredient
20:42in the standard model.
20:43It's the scaffolding.
20:44It determines where galaxies form,
20:47how they cluster,
20:48how the large-scale structure of the universe
20:51looks on the biggest scales.
20:53If dark matter particles are lighter
20:56or behave differently than the cold model assumes,
20:59the ripples from that change would propagate
21:01through almost every prediction cosmology makes.
21:04It would alter the mass of galaxy clusters.
21:07It would shift the expected distribution
21:10of dwarf galaxies around larger ones,
21:12a distribution that has already been
21:14a source of tension for years.
21:16It would change how gravitational lensing
21:19bends light around massive objects.
21:22You can't swap out the foundation of a building
21:24and expect the walls to stay where they are.
21:27So where does this leave us?
21:29The standard model isn't dead,
21:32not even close.
21:33It still explains the cosmic microwave background
21:36with extraordinary precision.
21:38It predicted the abundances of hydrogen and helium
21:41in the universe.
21:43It accounts for the clustering of galaxies
21:45across billions of light years.
21:47No competing model can match its track record.
21:51But it's developing fractures
21:53that weren't there a decade ago.
21:54The Hubble tension refuses to go away.
21:57The early universe is producing galaxies
22:00that are bigger, more structured,
22:02and stranger than expected.
22:04The nature of dark matter,
22:06which makes up roughly 27% of everything that exists,
22:11may not be what we've been modelling
22:12for the past 25 years.
22:15And dark energy,
22:16the 68% of the universe
22:18responsible for its accelerating expansion,
22:21remains a placeholder name
22:23for something we cannot identify,
22:25measure directly, or explain.
22:29There's even a second, quieter tension
22:31that most people haven't heard about.
22:34It's called the S8 tension,
22:37and it concerns how clumpy the universe is.
22:40The standard model predicts
22:42a certain amount of clustering
22:43in the distribution of matter across space.
22:46When astronomers measure that clustering directly,
22:49they consistently find slightly less of it
22:52than the model expects.
22:53Adam Rees has called it
22:56the little sibling of the Hubble tension,
22:58less dramatic, but worth watching.
23:01If both tensions turn out to be real,
23:03they may be symptoms of the same underlying problem,
23:07two cracks in the same wall.
23:10Adam Rees put it carefully in December 2024,
23:14as he often does.
23:15The discrepancy between the observed expansion rate
23:19and the predictions of the standard model
23:21suggests that our understanding of the universe
23:24may be incomplete.
23:26Two NASA flagship telescopes
23:28now confirm each other's findings.
23:31We have to take this problem seriously.
23:34What's remarkable about this moment in cosmology
23:37is that the anxiety is productive.
23:39Nobody is panicking.
23:41The prevailing mood is closer to the feeling you get
23:44when you've been assembling a jigsaw puzzle for years,
23:47confident you had the picture right,
23:50and then you find a piece that doesn't fit anywhere.
23:53It doesn't mean the puzzle is wrong.
23:55It means the picture on the box
23:57might not be the whole picture.
23:59New instruments are coming.
24:02NASA's Nancy Grace Roman Space Telescope
24:05will conduct wide-field surveys
24:07designed specifically to study dark energy.
24:10The European Space Agency's Euclid mission
24:13is already mapping the geometry of the universe
24:16on a scale no telescope has attempted before.
24:20Future data releases from the Gaia Space Telescope
24:23will let astronomers calibrate Cepheid distances
24:26with geometric precision,
24:28removing another potential source of error.
24:32And the Atacama Cosmology Telescope
24:34has already produced the most detailed ground-based map
24:38of the cosmic microwave background ever made,
24:41slightly nudging the Planck estimate upward,
24:44but not far enough to close the gap.
24:47We're living inside a question.
24:49The most precisely measured universe in history
24:52is telling us two different things at the same time,
24:55and both of them appear to be correct.
24:59Something in the space between the beginning and now,
25:01in those 14 billion years we've never directly observed,
25:05is doing something we don't understand.
25:08Some ingredient we haven't identified,
25:11some process we haven't modelled,
25:14some property of space-time we haven't imagined.
25:18Cosmology has been here before.
25:20In the 1990s,
25:21the field nearly collapsed under the weight
25:24of a different contradiction,
25:26measurements suggesting the universe was younger
25:28than its oldest stars.
25:30That made no sense.
25:32The stars couldn't be older than the thing they lived inside.
25:36Then came the discovery
25:37that the expansion of the universe was accelerating,
25:41driven by dark energy,
25:43and the whole picture snapped into focus.
25:46What looked like a fatal flaw
25:48turned out to be the shadow of something enormous
25:51that nobody had seen yet.
25:53That's the thing about crises in physics.
25:56They feel like endings.
25:57They almost always turn out to be doors.
26:01If you could somehow stand outside the universe
26:04and watch this moment,
26:05you'd see a species on a small, rocky planet,
26:08orbiting an unremarkable star
26:11in the suburbs of a mid-sized galaxy,
26:14using mirrors and math
26:15to argue about whether reality
26:17is expanding 6% faster than it should be,
26:21and getting genuinely upset about it.
26:23That's either absurd or magnificent,
26:27and I think it might be both.
26:29It's not a failure.
26:31Honestly, it's the opposite.
26:33This is what it looks like
26:34when a science reaches the boundary
26:36of what it knows and keeps pushing.
26:39The universe isn't broken.
26:41Our map of it is just incomplete.
26:44And the next version, whenever it arrives,
26:47is going to be stranger and more beautiful
26:50than anything we've drawn so far.
26:53Thanks for watching,
26:54and I'll see you in the next one.
Comments