Nobel Prize Winner Warns: "It's a Different Reality" — James Webb Found Something Impossible

[00:00:01]We were wrong about everything, not about one thing, not about a peripheral detail that needed adjusting at the margins. About the fundamental central loadbearing story of how the universe began, how structure formed, how the first galaxies assembled themselves in the ancient cosmos. about the picture that took decades of careful observation and theoretical development to build that passed test after test that became the established framework within which every cosmologist worked. That story the story of cosmic dawn of how the universe went from a featureless plasma to the complex structured galaxyfilled cosmos we inhabit appears to be wrong. or at minimum so substantially incomplete that the word wrong captures the spirit of the situation better than any polite qualification. This is not my assessment. This is what the data says.

[00:01:06]This is what James Webb is showing.

[00:01:08]Image after image, spectrum after spectrum as it peers deeper into the universe's past than any instrument before it. And it's what Nobel Prizewinning scientists, people who helped build the theoretical framework that Webb is now challenging, are saying out loud that what web is finding is impossible. That the universe cannot have assembled galaxies this large, this bright, this structurally mature, this early and that therefore something in our understanding of the early universe is fundamentally incorrect. The Nobel laurate in question is John Mather who shared the 2006 Nobel Prize in physics for his work on the Coobe satellite that first mapped the cosmic microwave background radiation with precision providing the observational foundation for our current cosmological model. Mather, who is also the senior project scientist for the James Webb Space Telescope, has said publicly that what web is finding is consistent with a different reality, that the early universe looks different than expected, that the models need revision. When the architect of the observational framework that confirmed the standard model says the observations suggest we were wrong, that's not sensationalism. That's a scientist following the evidence. Let me tell you exactly what Webb found and why it matters and what it means. To understand why Web's discoveries are so disturbing, you need to understand what the standard model of cosmology lambda CDM predicted about the early universe.

[00:02:58]specifically about how galaxies formed and when the big bang occurred approximately 13.8 billion years ago.

[00:03:08]For the first 380,000 years, the universe was an opaque plasma too hot for neutral atoms to form, filled with a sea of free electrons that scattered light continuously, making the cosmos completely opaque.

[00:03:24]This period is the cosmic dark age, illuminated only by the glow of the plasma itself. At 380,000 years, the universe cooled enough for electrons and protons to combine into neutral hydrogen. The epoch of recombination, the universe became transparent. The light from that moment has been traveling toward us ever since the cosmic microwave background. And the universe entered a new phase, the actual dark ages. Not metaphorically dark, but literally dark. No stars, no galaxies, just expanding clouds of mostly hydrogen and helium gas slightly clumped by the quantum density fluctuations from the inflationary era slowly concentrating under gravity. Then at some point between about 100 million and 500 million years after the big bang red shift Z between about 20 and 10 in the notation that cosmologists use where higher red shift means earlier times the first stars switched on the first light in the universe since the CMBB.

[00:04:33]Presumably enormous stars burning hot and fast and dying in spectacular explosions after just a few million years, seeding the surrounding gas with the first heavy elements forged in stellar cores from these first stars.

[00:04:49]The first galaxies assembled, small, irregular, filled with hot, bright, young stars, rapidly forming new stars from the abundant gas, growing over time as smaller structures merged into larger ones under the action of gravity. This is the paradigm of hierarchical structure formation. The bottom up assembly of cosmic structure from small seeds to large. Small galaxies form first, then merge into larger galaxies, which merge into larger. Still, the key predictions of the standard model for the early universe are specific. At red shifts above about 10 in the first 500 million years after the Big Bang, there should be very few galaxies and they should all be small. Star formation has barely begun. Structure has barely assembled. The universe is mostly smooth gas with only the most primitive beginnings of structure. As you go to lower red shifts, as you look at later times, 700 million years, 1 billion years, 1.5 billion years after the Big Bang, the number of galaxies increases, their sizes grow, their structures become more organized. By about three billion years after the big bang, large well-organized spiral and elliptical galaxies are establishing themselves.

[00:06:12]By 5 to 6 billion years, roughly half the current age of the universe, the galaxy population looks broadly similar to what we see today. This is the story.

[00:06:22]slow, gradual, hierarchical assembly of structure with the early universe being a place of small, primitive, poorly organized proto galaxies. Web looked and the early universe looked nothing like this. Let me tell you specifically what web found because the specifics matter.

[00:06:41]The first images from web released in July 2022 already hinted at something unexpected.

[00:06:49]background galaxies visible even in casual examination of deep field images at red shifts that implied they were less than a billion years old after the big bang. But those early hints became quantitative results published in rapid succession through 2022, 2023 and 2024.

[00:07:13]The first major shock came from galaxy candidate selection. Web's infrared capabilities allow it to detect galaxies at red shifts above 10 galaxies from the first 500 million years of cosmic history. The Hubble Space Telescope had found a handful of candidate galaxies at these extreme red shifts. Web was expected to confirm some of these candidates and perhaps find a few more.

[00:07:38]Instead, Web found hundreds, many more galaxies at extremely high red shift than the standard model predicted. Not a factor of two more, factors of 10 to 100 more. In some analyses, the number density of massive galaxies, galaxies with large amounts of stellar mass already assembled in the first 500 million to 1 billion years of cosmic history, was dramatically higher than lambda CDM predicted. One paper published in early 2023 caused particular uproar led by astronomer Evo Labbe of Swinburn University. It identified a population of galaxies at red shifts between about 7 and 10 with stellar masses of 10 10 to 10 11 solar masses comparable to the stellar mass of the Milky Way today. These galaxies appear to have assembled the equivalent of a significant fraction of the Milky Way's total stellar mass within the first 700 million years of cosmic history. The problem is not that such massive early galaxies are impossible in some absolute sense. It's that they're extraordinarily difficult to produce within the standard models framework.

[00:08:58]Building a galaxy with 10 10 solar masses of stars by 700 million years after the Big Bang requires an extraordinarily high rate of star formation. Stars being formed at a rate that may exceed the amount of available gas in the models dark matter halos. It requires an efficiency of conversion of gas into stars that approaches or exceeds 100% essentially all the available gas being converted into stars with nothing left over. The standard model predicts that star formation efficiency is low. Typically a few% of available gas is converted into stars at any given time with the rest kept from forming stars by feedback processes.

[00:09:47]Supernova that heat the surrounding gas, radiation from massive stars that pushes gas outward, super massive black holes that inject enormous energy into their host galaxies. These feedback processes are not optional features of the model.

[00:10:03]They're required. Without them, models produce too many stars too quickly. The efficiency must be kept low. And now web is finding galaxies that appear to have operated at near 100% efficiency in the first 700 million years. Either the efficiency was genuinely that high which would require completely reconsidering feedback physics or something else is going on.

[00:10:33]The second major discovery is the morphology, the shapes and structures of these early massive galaxies. And this is where the different reality description becomes most apt. In the standard model, early galaxies should be messy, irregular, and asymmetric.

[00:10:52]They're forming from collisions and mergers of smaller structures. Star formation is chaotic and distributed.

[00:11:01]The stellar populations are young, dominated by hot blue stars. There's no time for the ordered symmetric structures, the grand spiral arms, the smooth elliptical profiles that characterize mature galaxies in the local universe. Early galaxies should look like what they are, works in progress. Some early galaxies observed by web do look like this, but many don't. Web is finding early galaxies with remarkably ordered disclike morphologies, thin discs of stars, flat profiles, rotation dominated kinematics, the kind of organized rotation that's the hallmark of a settled mature disc galaxy, not a recently merged dynamically hot system. One analysis of galaxies at red shift 6 to8 about 800 million to 1 billion years after the big bang found that a substantial fraction of them have disclike morphologies.

[00:12:03]Not protodisks, not incipient spirals.

[00:12:07]Galaxies with the structural properties of mature disc galaxies like the Milky Way assembled in a universe that should have been far too young and too turbulent to produce them. This is the morphological paradox. The structures exist before the model says the structures can exist. Now, let me tell you about the specific discovery that generated the most attention and that most directly challenges the standard model. Something that astronomer Allison Kirkpatre when she first saw the data described as mindblowing.

[00:12:44]something the cosmological community has been calling the impossibly early galaxy problem. It's a set of candidate galaxies identified at red shifts between about 12 and 17 corresponding to the universe being between about 200 million and 350 million years old. If these candidates are confirmed, they represent galaxies in a universe barely a quarter of a billion years old. For context, a quarter of a billion years after the Big Bang, in the standard model, the universe shouldn't really have galaxies at all. It should have protogalactic structures, small clumps of dark matter with gas beginning to cool and condense inside them. Stars should be forming, yes, but in small numbers. in small structures. The first truly galactic scale assemblies of stars shouldn't exist for at least another few hundred million years. The candidate galaxies at Z greater than 12 identified in Web's first deep field images appear to be luminous enough to contain substantial stellar populations.

[00:13:59]If their red shifts are confirmed, if spectroscopic follow-up verifies that these objects are truly at these distances and not foreground interlopers at lower red shift, then the universe assembled massive luminous objects with stellar masses of many billions of solar masses within the first 200 to 300 million years of its existence. Many of these candidates have now received spectroscopic confirmation.

[00:14:28]Web's nerspec instrument has measured precise spectroscopic red shifts for dozens of galaxies at Z greater than 10 and several confirmed galaxies now exist at Z greater than 12. The most distant spectroscopically confirmed galaxy as of this conversation is at a red shift of approximately 13.2 corresponding to a universe about 320 million years old.

[00:14:56]This galaxy J A D SGS130 discovered in Web's Jade's deep field survey has a stellar mass of roughly 108 solar masses. That's not enormous. It's much smaller than the Milky Way, but it's a substantial stellar population in a universe that in the standard model should barely have started forming stars.

[00:15:22]And there are candidates at even higher red shifts, possible galaxies at Z greater than 15, 16, even 17 that would place them in the first 200 million years. These candidates haven't yet been spectroscopically confirmed, but phototric evidence for them is accumulating in multiple deep field surveys. Now I want to address the question that any careful scientist should ask. How confident are we that these objects are what they appear to be? The concern is legitimate.

[00:15:59]Identifying galaxy red shifts from phototric measurements from the colors of objects in multiple infrared filters is subject to systematic errors. A galaxy at very high red shift has a specific phototric signature. It appears bright in filters covering the rest frame ultraviolet wavelengths that are redshifted into the near infrared and drops out of filters at shorter wavelengths. This dropout signature is how web identifies high redshift candidates. The concern is that some of these candidates might not be high redshift galaxies. They might be lower redshift galaxies whose colors happen to mimic the dropout signature for reasons other than extreme red shift. They might be red dusty galaxies at moderate red shift whose dust absorption mimics the high Z dropout. Or they might be phototric artifacts, bright objects whose light has been scattered or processed in ways that create false dropout signatures. This is why spectroscopic confirmation is essential and why the results that have been spectroscopically confirmed are taken most seriously. But here's the important point. Web has produced spectroscopic confirmations of galaxies at unprecedented red shifts. The confirmed galaxies at Z greater than 10 112 are real. They're not phototric artifacts.

[00:17:28]They're genuinely at these red shifts and their luminosities and stellar masses are genuinely unusual relative to standard model predictions. The specific number density of massive early galaxies, the quantitative excess over predictions is still being worked out as larger samples are accumulated and phototric calibrations are refined. Some of the most extreme early claims have been somewhat reduced by more careful analysis, but the excess remains. Fewer of the candidates are at the most extreme red shifts than the initial claim suggested, but there are more massive luminous early galaxies than the standard model predicts. The excess is robust. Let me now tell you about the third major discovery from web that's challenging our understanding of the early universe. Something related to the first two, but distinct something about the chemical composition of early galaxies. The standard model of cosmic chemical evolution predicts that the early universe should be chemically primitive. The big bang produced hydrogen, helium, and trace lithium.

[00:18:42]Nothing heavier. Heavy elements, carbon, oxygen, silicon, iron. All the elements essential for chemistry and biology are produced in stellar cores and distributed through the surrounding medium by stellar winds and supernova explosions.

[00:19:01]It takes time, many generations of stellar evolution to build up significant abundances of heavy elements in the interstellar medium. In the standard model, the earliest galaxies should be essentially metal pore, not metalf-free. Even the first supernova will enrich the surrounding gas somewhat, but with heavy element abundances that are a few% of solar levels at most. Truly primordial chemistry dominated by hydrogen and helium with only the smallest trace of heavier elements. Web is finding early galaxies with heavy element abundances that are far higher than expected for their age. Galaxies at red shifts of 6 to 8 about 800 million to 1 billion years after the big bang show oxygen and carbon abundances that are 10 to 20% of solar levels. Some show nitrogen abundances that are anomalously high nitrogen to oxygen ratios that are difficult to explain with standard nucleiosynthesis models. The elevated metallicity implies multiple generations of star formation before we observe these galaxies. Many cycles of stellar birth, evolution, death, and chemical enrichment all packed into the first billion years. This requires a star formation rate and a stellar evolution rate that is compressed relative to standard expectations. The universe is chemically evolving faster than the models predict and the anomalous nitrogen abundances are particularly puzzling. In the local universe, very nitrogen-rich environments are found in specific circumstances in the immediate vicinity of certain types of evolved massive stars and in the cores of very old globular clusters. Finding high nitrogen to oxygen ratios in young early galaxies is unexpected and doesn't fit neatly into any standard model of nucleiosynthesis. Now, I want to tell you what all of this means for the standard cosmological model. Because it's not a simple story. It's not that the standard model is wrong and we need to throw everything out and start over.

[00:21:24]It's more nuanced and more interesting than that. The lambda CDM model, the standard model with dark matter, dark energy, and standard inflationary initial conditions has passed an enormous number of tests. the structure of the CMBB, the largecale distribution of galaxies, the abundance of light elements from big bang nucleiosynthesis, the acceleration of cosmic expansion measured from supernova. All of these are well explained by lambda CDM and no serious cosmologist proposes discarding the entire framework. What web is challenging is the astrophysical model built on top of lambda CDM, the specific model of how structure forms and how galaxies assemble within the dark matter framework. The dark matter halos are presumably still there. The inflationary initial conditions are presumably still correct. But the processes that convert dark matter halos into galaxies, the gas physics, the star formation, the feedback apparently operate differently in the early universe than the models assume. The leading candidates for what's going wrong with the models fall into a few categories. The first is feedback efficiency. The standard model assumes that supernova and active galactic nuclei inject energy into the gas surrounding forming galaxies, keeping the star formation rate low and preventing too many stars from forming too quickly. If this feedback is less effective in the early universe, if the gas densities are high enough that the feedback energy is quickly radiated away rather than heating the gas, then star formation could proceed at much higher efficiency. rapidly assembling mass of galaxies. The second is the initial mass function. The initial mass function, the distribution of stellar masses in a newly formed stellar population is observed to be relatively universal in the local universe. But in the early universe with different gas conditions, the initial mass function might be different. Specifically, the earliest stellar generations might be dominated by very massive stars, hundreds to thousands of solar masses that burn extremely hot, live for only a million years, and produce enormous amounts of UV, radiation, and heavy elements during their brief lives. A topheavy initial mass function would produce more luminosity per unit stellar mass than the standard initial mass function, making early galaxies appear more massive than they actually are. The third category is more radical modifications to the cosmological model itself. Some researchers have proposed that the dark matter model needs modification, that dark matter might have different properties than assumed, or that there might be additional physics in the dark matter sector that accelerates structure formation in the early universe. Warm dark matter, fuzzy dark matter, self-interacting dark matter. These alternatives to standard cold dark matter make different predictions about early structure formation. And the most radical proposals involve modifying the initial conditions, the inflationary power spectrum. If the primordial density fluctuations were larger at small scales, then the simplest inflationary models predict structure could form earlier and more abundantly, providing the seeds for the massive early galaxies that web is finding. Let me now connect this to something I find philosophically profound about what web is doing. Not just finding unexpected objects, but forcing a confrontation with the limits of our theoretical models in a way that's deeply uncomfortable and deeply productive. The standard model of galaxy formation was built as all scientific models are from the available data. The Hubble Space Telescope's deep field images provided the deepest view of the early universe before web. Hubble found a small number of candidate galaxies at very high red shift. The models were calibrated to match those observations. Simulations like the ellustrous and eagle simulations, which track the formation and evolution of billions of simulated galaxies from the big bang to the present, were tuned to reproduce Hubble's observations.

[00:26:10]These simulations are extraordinary computational achievements.

[00:26:16]They incorporate detailed physics, gas dynamics, star formation, supernova feedback, black hole accretion and feedback in self-consistent frameworks that make specific predictions about what the universe should look like at different epochs.

[00:26:36]And they've been validated against a wide range of observations in the local and moderately distant universe. But they were not validated against observations at Z greater than 10.

[00:26:50]Nobody could validate them against such observations before web because no instrument could see that far with sufficient depth and resolution. The models were extrapolated to the early universe based on the physics they incorporated and the theoretical assumptions they made. Web is validating these extrapolations and the extrapolations are wrong at least partially in the specific domain of the number, density and properties of the most massive, most luminous early galaxies. This is how science works at the frontier. You build the best model you can. You make predictions. You build the instrument that can test the predictions. The instrument finds discrepancies.

[00:27:32]You revise the model. The discrepancy between Web's observations and the standard model predictions is not a failure of science. It's science working exactly as it should. The models made specific quantitative predictions.

[00:27:49]Web is testing those predictions. The predictions are wrong in specific ways.

[00:27:56]Now we figure out why and the model gets better. I want to now tell you about something that adds another layer to the web puzzle. Something that connects the early galaxy problem to the Hubble tension that we discussed in an earlier conversation. You remember the Hubble tension, the discrepancy between the two measurements of the universe's expansion rate. The CMBB base measurement gives H0 of about 67.4 kms MPC. The distance ladder measurement gives about 73 km MPC. The two measurements disagree at about five sigma and no satisfactory explanation has been found. Now, here's the connection. Some proposed solutions to the Hubble tension involve modifications to the early universe's expansion rate, adding a component of early dark energy that slightly increase the expansion rate before recombination, shortening the sound horizon, and shifting the CMBbased H0 prediction upward. These early dark energy models don't resolve the Hubble tension perfectly, but they reduce it somewhat.

[00:29:11]The relevance to the early galaxy problem is this.

[00:29:14]A faster expansion rate in the early universe as early dark energy would produce means that dark matter halos would form faster, providing more opportunities for early galaxy assembly.

[00:29:26]The same early dark energy that might partly explain the Hubble tension might also help explain the overabundance of massive early galaxies. It's a speculative connection. Neither the early dark energy explanation for the Hubble tension nor its implications for early galaxy formation are established, but it's an example of how the multiple anomalies in current cosmology, the Hubble tension, the early galaxy overabundance, the anomalous CMB features we discussed in the CCC conversation might not be independent problems but symptoms of a single deeper modification to the standard model, something might be wrong with our picture of the early universe in a way that affects multiple observations simultaneously.

[00:30:15]Finding out what that something is, following the data wherever it leads is the central project of observational cosmology right now. Let me close this first part with something that captures what I think is the most important aspect of what web is doing. Astronomy has always advanced through the opening of new windows. Galileo's telescope opened the window of visible light at high magnification.

[00:30:41]Radio astronomy opened the window of radio frequencies. X-ray astronomy opened the window of high energy photons. Each new window revealed phenomena that were invisible through previous windows. Phenomena that required theoretical revision to accommodate. web has opened a new window. Not just infrared light at unprecedented sensitivity, but the specific window of the universe's first billion years at the resolution needed to study individual early galaxies in detail.

[00:31:16]This window was theoretically anticipated. We knew web would see early galaxies, but its contents were unknown.

[00:31:22]We had models, but we hadn't seen. And what we see is not what the models predicted. We were wrong about everything about the number of massive early galaxies, about their structures, about their chemical compositions, about the apparent efficiency with which the early universe converted gas into stars.

[00:31:43]Not wrong in the sense that the laws of physics are different. The physics is the same. Not wrong in the sense that dark matter doesn't exist. It probably does. wrong in the sense that our specific models of how the physics played out in the early universe are incomplete. Something was happening in the first billion years that we didn't adequately model. Something that made structure form faster, made galaxies bigger earlier, made star formation more efficient, made chemical enrichment more rapid. what that something is. Whether it's modified feedback physics, a different initial mass function, modifications to dark matter, modifications to inflation, or something we haven't thought of yet is the question that's now driving observational and theoretical cosmology.

[00:32:34]In part two, I want to take you deeper into the specific candidates into the physics of each proposed explanation and what each implies about the early universe into what Web's continuing observations are revealing as the sample of early galaxies grows and into what it would take to resolve the tension. What observations or theoretical breakthroughs would settle the question of why the early universe looks so different from what we expected? We were wrong about everything. The question now is wrong in exactly which way. And that question asked with the full force of the best telescope ever built is one of the most exciting questions in the history of cosmology. So, we'd arrived at this place where web has revealed an early universe that doesn't match our models. More massive galaxies, more structured galaxies, more chemically enriched galaxies assembled earlier than Lambda CDM's astrophysical framework predicts. The discrepancy is real, confirmed by spectroscopy, and growing more precise with each new web observation. Now I want to go deeper into the specific physics of the proposed solutions into what each candidate explanation actually predicts and what it would mean for our understanding of the universe if each one turns out to be correct and into something that I think is the most profound aspect of this whole situation.

[00:34:16]what it tells us about the relationship between dark matter and the visible universe we can observe. But I want to start with something concrete with numbers because I think the quantitative magnitude of the discrepancy is not fully appreciated and getting it right changes how you think about what's needed to explain it. Let me give you the specific numbers for the excess. The quantity cosmologists use to characterize the abundance of massive galaxies is the stellar mass function, a statistical description of how many galaxies of each stellar mass exist per unit volume of space at a given cosmic epoch. It's essentially a population census. How many Milky Way mass galaxies per cubic gigapseek exist at redshift 10? How many at red shift 8? How many at red shift six? The standard model makes specific predictions for these numbers.

[00:35:18]The predictions come from combining the dark matter halo mass function which is well understood from endbody simulations with models of how dark matter halos convert their gas into stars. The predictions at Z greater than 6 are uncertain because they depend on extrapolating the star formation and feedback models to regimes that haven't been directly observed, but they're the best predictions we have. and they're taken seriously by the community. Web's early results suggest that the observed stellar mass function at Z greater than six exceeds the model predictions by factors of roughly 10 to 100 for the most massive galaxies. At Z greater than 10, where the predictions are most uncertain and the web data is most surprising, the excess might be even larger, but the sample sizes are smaller and the uncertainties are correspondingly larger. A factor of 10 to 100 excess, not a 10% discrepancy, a factor of 10 to 100. This is not something that tweaking a single parameter in the model can fix.

[00:36:28]Models have free parameters. The efficiency of star formation, the strength of feedback, the threshold for stellar wind launching. Reasonable variations in these parameters can change predictions by factors of a few.

[00:36:43]A factor of 10 to 100 requires something more fundamental. Let me now walk you through the leading explanations one by one with the physics. The first and most conservative explanation is modified star formation efficiency. In standard models, the efficiency of converting gas into stars, the fraction of available gas that becomes stars per unit time is low, perhaps 1 to 10% per freef fall time. This low efficiency is enforced by feedback. Supernova explosions that heat and expel gas radiation from massive young stars that ionizes the surrounding medium and prevents it from cooling and collapsing.

[00:37:30]In the early universe, the physical conditions are different from the local universe in several ways that might allow higher star formation efficiency.

[00:37:41]The gas density is higher. The universe is smaller, more compact, and gas densities in the dark matter halos were correspondingly higher. Higher density means faster cooling. Faster cooling means gas can radiate away the energy deposited by feedback more quickly before the feedback can halt star formation. The feedback is less effective not because it's absent, but because the dense environment neutralizes it rapidly. If early star formation efficiency was 30 to 50% rather than 1 to 10%. The mass of galaxies observed by web could be produced without violating any fundamental physical laws. The gas was there. The dark matter halos were there.

[00:38:23]The efficiency was just higher than we assumed. This explanation has the appeal of requiring no new physics. Just different conditions leading to different efficiencies of known processes. Its weakness is that it's somewhat ad hoc. We don't have a detailed physical model that reliably predicts efficiency of 30 to 50% in early universe conditions. And some analyses suggest that even 100% efficiency, all available gas converted to stars, is insufficient to produce the most massive of the observed early galaxies within the available dark matter halos. The second explanation is the topheavy initial mass function. And this is where the physics gets genuinely interesting. The initial mass function, the IMF, is the distribution of stellar masses produced when a cloud of gas collapses into a cluster of stars.

[00:39:18]Observations in the local universe suggest the IMF is roughly universal approximately a krupa or chabriier function with most of the mass in intermediate mass stars but most of the light produced by massive stars. The theoretical expectation is that the IMF depends on the temperature of the gas from which stars form. Hotter gas can't fragment as easily thermal pressure resists the collapse of small clumps. So only larger clumps can collapse and form stars. The minimum stellar mass scales roughly with the genes mass. The minimum mass for gravitational collapse which depends on gas temperature. In the early universe, the gas is enriched but only slightly. The cooling channels are different. The cosmic microwave background temperature is higher at red shift 10. The CMBB temperature is about 30 Kelvin compared to 2.7 Kelvin today.

[00:40:15]This higher CNB temperature sets a floor on how cold the gas can cool. Molecular hydrogen cooling. The primary cooling mechanism for the first stars can only cool gas to temperatures limited by the CMB floor. In this warm, dense early environment, the genes mass is higher.

[00:40:37]The minimum stellar mass is higher. The IMF might be systematically shifted toward more massive stars. is a topheavy IMF where the typical star is 10 or 100 times more massive than in the local universe. A topheavy IMF has several effects that are relevant to Web's observations. First, it dramatically increases the luminosity per unit stellar mass. Massive stars are vastly more luminous than low mass stars. So a population dominated by massive stars appears far brighter than a population with a standard IMF of the same total mass. Second, it increases the supernovi rate per unit stellar mass which accelerates chemical enrichment explaining the high metalicities web is finding in early galaxies. Third, it changes the ionizing photon budget.

[00:41:34]Massive stars produce more UV photons per unit mass, which may help explain the reionization of the universe. If early galaxies have a topheavy IMF, then their true stellar masses might be much lower than we're inferring. We're inferring stellar mass from luminosity.

[00:41:53]And if we're using the wrong IMF to convert luminosity to mass, we'll overestimate the mass. An apparent galaxy of 10 10 solar masses might actually contain only 10 nine solar masses of stars making it much less challenging to produce in the standard models dark matter framework. This is one of the most important systematic uncertainties in the current web results. We don't observe stellar mass directly. We observe luminosity and convert it to mass using population synthesis models that assume a standard IMF. If the early universe IMF is different, those mass estimates are wrong. Testing this hypothesis requires detailed spectroscopy measuring the strengths of specific spectral features that are sensitive to the IMF like specific stellar absorption features and specific ionization state ratios.

[00:42:50]Web spectroscopic capabilities are being used for exactly this purpose and early results are mixed. Some early galaxies show spectral features consistent with the standard IMF. Others show hints of a more topheavy distribution. The third explanation is early dark energy and this is where the web observations intersect with the Hubble tension in the way I described at the end of part one.

[00:43:18]Standard cold dark matter assumes that dark matter is cold, that it was non-relativistic at the time of matter radiation equality when the dark matter halo mass function was set. Cold dark matter produces a specific power spectrum of density fluctuations. This specific shape of the clumpiness of matter across different scales. This power spectrum has been measured exquisitly in the CMBB and in the large scale structure of galaxies and the standard cold dark matter model fits these measurements beautifully. But the mass function of dark matter halos and therefore the abundance of early galaxies is sensitive to the amplitude and shape of the density fluctuation spectrum at small scales. If the primordial fluctuation spectrum has more power at small scales than the standard model predicts, more smallcale halos form earlier and more massive galaxies can be assembled earlier. A concrete way this could happen is through features in the inflationary potential bumps or wiggles in the slow roll dynamics of the inflaton field that amplify fluctuations at specific scales. Such features are not predicted by the simplest inflationary models, but they're not forbidden either. They're constrained by CNB measurements at large scales, but at small scales, scales too small to be directly probed by CNB observations. The constraints are weaker. If the primordial power spectrum has an enhancement at the scales corresponding to early massive galaxies, the dark matter halos that host those galaxies would be more abundant than standard models predict. and the discrepancy with web would be reduced.

[00:45:06]Early dark energy, a component of dark energy that was present in the early universe before recombination, also affects structure formation. Early dark energy temporarily increased the expansion rate of the universe before recombination, shortening the sound horizon and affecting the growth of structure. The specific effects on the mass function of early dark matter halos are model dependent. But in general, a higher early expansion rate tends to suppress rather than enhance early structure formation, which is the wrong direction to explain the web excess. So early dark energy in its simplest forms doesn't help with the early galaxy problem, even if it partially helps with the Hubble tension. more complex modifications involving both a higher early expansion rate and enhanced primordial power at small scales might potentially address both problems simultaneously.

[00:46:03]But these are speculative require fine-tuned coincidences between multiple modifications and haven't been worked out in detail. The fourth explanation is perhaps the most conceptually radical.

[00:46:16]It involves modifying the nature of dark matter itself. Standard cold dark matter has one key property that's relevant here. It's cold. Cold means non relativistic at matter radiation equality. Non-relativistic dark matter produces a power spectrum of density fluctuations that has more power at small scales, more small scale clumping than warm or hot dark matter would produce. But what if dark matter is slightly warm? Warm dark matter, dark matter particles with a non-negligible thermal velocity of matter radiation equality would produce a power spectrum that's suppressed at small scales. This suppression would reduce the number of small dark matter halos and delay the formation of the first galaxies.