Scott Dodelson spent ten years on the Dark Energy Survey testing the standard model of cosmology. It passed and missed by 2.5 sigma. The Director of Fermilab’s Cosmic Physics Division on why the theory he helped build may be wrong, why nobody can find dark matter, and what it takes to change a scientific consensus. Subscribe if you want science with evidence, not speculation.
Dodelson is Director of the Cosmic Physics Division at Fermilab, Professor of Astronomy and Astrophysics at the University of Chicago, and author of Modern Cosmology, the textbook a generation of cosmologists learned the field from.
Lambda-CDM predicts how the early universe’s tiny fluctuations grew into the structure we see today. The Dark Energy Survey was designed to check that prediction. The answer came back two and a half sigma off. That is both the most precise measurement ever made of how the universe grew, and a crack he cannot stop looking at. His question is not whether the model is close. It is whether close is enough to trust.
We get into the Sigma-8 tension and what it actually takes for a scientific community to change its mind, the Dodelson-Widrow mechanism and his 1994 proposal that sterile neutrinos produced in the early universe could constitute dark matter, and the Neptune vs. Vulcan history of getting dark matter right and wrong. Asked which of those two situations we are in now, Dodelson’s answer is: I have no idea.
What you’ll hear: -Why the April 24, 1992 CMB discovery may have been the only science story ever to lead the New York Times front column -The Dodelson-Widrow mechanism: how ordinary neutrinos in the early universe may have oscillated into the dark matter we see today -Why Brian told Neil deGrasse Tyson to his face that we have already detected dark matter -What it means when a theory can accommodate any result and whether inflation has that problem -Why the particle physics community still does not trust cosmology’s neutrino mass measurements -What cosmology looks like in 2036 if Lambda-CDM breaks
Killing the model is my dream.
CHAPTERS 0:00 None of it has been found in a lab. How many free passes do we get? 0:44 Dodelson helped build Lambda-CDM. His dream is to kill it. 2:10 10 years. One prediction. Two and a half sigma off. 5:42 "You're in charge" — what mentorship in science actually looks like 6:56 April 24, 1992: cosmology stopped being speculation 9:48 Geoff Burbidge went to his grave a steady-state believer 11:04 What a neutrino is and why it barely interacts with anything 12:36 Brian told Neil deGrasse Tyson we already detected dark matter 18:08 How ordinary neutrinos may have become dark matter 23:24 Lambda-CDM predicts Manhattan's density 13.7 billion years later 26:42 Two and a half sigma: technically a 1% chance the theory is right 28:04 Neptune was dark matter. Vulcan wasn't. Which one are we in now? 31:20 Dark matter, inflation, dark energy: none found in a lab 36:46 If both cracks are the same crack, Lambda-CDM is finished
Get the transcript, fascinating bonus content, and my Monday M.A.G.I.C. Message: https://briankeating.com/yt
Support Into the Impossible on Patreon, get my weekly M.A.G.I.C. Message, unfiltered bonus content, and live monthly Office Hours with me: https://www.patreon.com/drbriankeating
Scott Donaldson discusses the tensions within the Lambda CDM model of cosmology, exploring how the Dark Energy Survey (DES) tests theoretical predictions against observational data. The conversation covers neutrino physics, the nature of dark matter, and the philosophical evolution of scientific paradigms, highlighting the gap between theoretical \"postdictions\" and empirical evidence.
What we've done in order to make the story of the universe work is introduce dark matter,
introduce inflation, introduce dark energy, and none of this has been found in the lab.
How many free passes do we get? Exploring the tension, killing the model, is my dream.
That's Scott Donaldson. He runs the Cosmic Physics Division at Fermilab. He teaches at
the University of Chicago, and if you've ever taken a graduate cosmology class,
probably took it from Scott. He also spent 10 years running the sharpest test anyone has ever
made of the standard model of cosmology, the model he helped build. The answer came back,
two and a half sigma off. Close enough to call it a triumph, not close enough to stop staring
at it in disbelief. I'm Brian Keating. This is Into the Impossible. What is new in the field
of dark energy? Yeah. Before we go any further. Well, since you've been in the field, we've had
this fiducial model of cosmology. I guess you, well, you have to establish it, really. So,
it's called Lambda CDM. I've heard you talk about it. And now there, you've had Kyle and other
people on saying that it's under stress. So, the fundamental question that I'm interested in
is not, or one of the questions I'm interested in is, how will we change our mind if we will?
That is, we have this pretty simple model, and then there are these data points which are saying,
oh, this doesn't work. This doesn't work. Oh, maybe it doesn't work. And so, we have to digest
that, and everyone digests it in their own way. And then how are we going to collectively land
that in another model? And so, I know people have written books about this. Tom O'Keefe's
Structure of Scientific Revolutions. But we're living through that time now.
And maybe in many ways, with AI, we'll get to that later on.
So, it's not, and of course, it's not just in cosmology. So, cosmology is the lens through
which I can understand stuff. But as you say, in society in general, we're losing faith in
institutions. So, we don't know which institutions we're going to land on, and which to believe,
which to trust. So, I think it's kind of an important question. And so, I've been trying
to explore it in this little corner of our world, cosmology, which is, in some ways,
the simplest thing we do. It's very hard to be a parent, to be a spouse, to be a friend,
but cosmology is really easy, because there's no people, and it's an easier thing. So, that's why
I've been trying to explore it in that context. For about 10 years, I was heavily involved in
this project called DES, the Dark Energy Survey. And that was, started taking data in 2012. And
it takes an enormous amount of time to process and analyze this data. So, we only put out our
final results a few months ago. So, that's what's been occupying me for the last 10 years or so.
Look about the connection between the type of science that I do, which is the first light in
the universe, the cosmic microwave background. I've talked a lot about that. I've talked less
about the kind of science that DES does, though, with the exception of conversations with people
like Kyle and others. But talk about what is DES? Again, it's somewhat strange. Not only are you not
using particle detectors and whatnot, but you're using optical telescopes. So, what does DES do?
What is it comprised of? You mentioned how long it took, but what really went into that? What's
the portfolio diversification between theory, which is what you do, experimental hardware,
observations, big data, machine learning? What are the different ingredients in DES? And first
of all, what does it stand for? DES stands for the Dark Energy Survey. The detectors on the
telescope, which was in Chile, are made of silicone. And so, they leverage a lot of the
silicone technology that Fermilab was already an expert in. So, that's one piece of it. The camera
itself was built by Fermilab. So, that's a tremendous contribution. Then there's the whole
data processing thing. And again, in that realm, high-energy physics has a huge advantage because
they're used to processing tremendous amounts of data. So, there's those two pieces. And then
there's the analysis, which takes us an enormous amount of time. And that piece is ongoing. It's
really thinking back to when it was first started. We thought of it in one way, and
all our analysis tools have evolved partially due to AI. So, that's been kind of interesting
to be a part of. But getting back to the science part of it, there's a strong connection between
what you do and what I do in the sense that one of the things you do is you measure in the cosmic
microwave background the fluctuations. So, I don't know if people can see this, but there are small,
hot, and cold spots in this microwave background. Basically, what that means is the temperature,
say, on this spot over here, was a little bit about one part in 10,000 higher than the temperature
over here. All that means is there were more photons there. And because there were more photons
there, because the photons were also talking to the electrons and protons, there were more electrons
and protons there. So, if you go back to the early universe, what you have provided us with is a
snapshot of the early universe. And we know what it looks like. It was incredibly homogeneous.
So, the universe was basically the same everywhere and with very, very small fluctuations of one part
in 10,000 or 100,000. That's what you've provided us. What this theory that we've landed on Lambda
CDM predicts is how those small fluctuations will accrete more and more matter over the course of
13.7 billion years. And we can test that with DES and see whether those fluctuations have grown
to the size they were supposed to according to the theory. So, that's, to me, the most fundamental
test we're doing. If it hasn't, that means this theory is wrong. So, to me, that's the stress
test that's occupied me for 10 years. Why do we tend to kind of just, as I describe it, you know,
have the series of descriptions to our students, you know, from Nobel Prize to Nobel Prize without
loss of enthusiasm? How do you view the way that we should teach cosmology? I'm not sure this is
direct answer, but this is my experience of it, which is in DES, to me, the most important thing
has not been the papers we've written or that discovery we made that basically the fluctuations
are what they should be in this theory. But the most important thing has been the people.
So, and I think we as mentors get an enormous amount of satisfaction by working with these
young people. So, communicating not just facts to them, but also the very little that we know
has been, I think it's eye-opening to them. I'll give you an example. I was at a collaboration
meeting a couple of years ago, and this first-year grad student from Spain was there, and she was
talking about her work and on this very complex piece of analysis. And I went up to her afterwards
and I asked her a question. I said, well, do you understand this and this? She goes,
you don't understand. I've only been working on this for three months. I said, no, you don't
understand. You're in charge. So, it's like these people, the young people, they're the ones who,
basically, it's not like there's some threshold above which you become an Einstein. We're all
kind of swimming around in the sea of ignorance trying to figure out things. We have a little
more experience than younger people, but they have advantages that we don't have, for example. So,
to me, that's been the best part about DES is working with these young people. It's been great.
You talk in your sub stack, which we'll link below, and that you kind of had this really
phase change, I think is the only word, revolution, April 1992. Talk about the day,
you know, the music died. Talk about Kobe. I remember that I was in college at the time,
and I was in a summer program or starting a summer program. And I kind of even had a glimpse
that this was something important, but I didn't know I would do it for the rest of my career.
Talk about 1982, 1992, why that was so transformative for you in your career. It
seems to have pivoted you in a completely different direction, which is wonderful to
say about your flexibility intellectually, but why was it so important to you?
I think what happened before 1992 is there were a lot of people who had
really fascinating ideas about cosmology. And one of those ideas, what I just described,
that there were small fluctuations in this cosmic microwave background, and they grew to be the
structure we see today. But there was absolutely no evidence. There was no prediction that had
been made that had been verified. There was only postictions. People observed stuff and they said,
oh yeah, I can explain it this way, that way. So any reasonable person looking at that would say,
I don't trust these guys. And in fact, quite a few astronomers and other people didn't trust
a few cosmologists working in the field for very good reason. So what happened on April 24th, 1992
is the first detection of anisotropies in the cosmic microwave background. So the fact that
spots that are slightly hotter than other spots. So that discovery was a prediction. Stephen Hawking
called it the discovery of the century, if not ever. So I mean, he was one of the driving forces
behind the whole field of cosmology. So he recognized how important it was. And another
thing that I highlighted there is that in the old days, they used to print, you're probably too young
for this, the New York Times on paper. And so the first column of the New York Times on the right
hand side was the lead article. So that was the first, I think the only time in the history of
the New York Times that a science article was on the first column. And I think they got that right.
Because that was the discovery that underpins modern cosmology in a thousand years. That's
the story people will be pointing to, I think. For me, actually, at the time was kind of depressing
because I was working on all these fanciful things that were fun to work on and to understand as a
theorist. The physics that goes into that required me to learn a whole new set of tools,
which I didn't, you know, it wasn't up my alleys. I knew I had colleagues who ended up just staying
and not moving forward, but I kind of had a family of sports. I figured I got to learn this stuff.
So I tried really hard to learn this stuff and ended up writing this book to help me learn it.
And so I was able to stay in the game to some extent. Yeah.
Not pivot from particle theory to theoretical astroparticle cosmology,
which you're one of the major instigators of, I think, to be fair to you. There were also
separate physics and cosmology, physics and astronomy, astronomy and cosmology. They were
all kind of separate things. Yeah. As I told you, the previous occupants of this office was
Jeff and Margaret Burbidge. This plate is one of the Palomar plates that she took with the galaxy.
Wow. That's cool.
I've got some red shifts. So they and they didn't believe in cosmology basically until the day they
died. You know, Big Bang cosmology. Jeff used to go into paroxysms of rage when a speaker would
have the misfortune of mentoring. He was a steady state proponent until he died long after Kobe.
What do you make of that? Are there modern day versions of people like Jeff that are just kind
of eminent, brilliant scientists that just do not accept either inflation or maybe it's string
theory or the standard model? Do you see any parallels between the Big Bang deniers of that
age, which you lived through partially, and today's age? I think so. I think we tend to weight
things according to our experiences. So I just wrote a thing about something that is obvious to
you, but maybe not obvious to most people, which is the sky, the light from the stars or galaxies
comes in different colors. You can view a given galaxy with one filter and see it as one color or
view it with a different filter and see it in a different way. And that's a metaphor to me of the
way we perceive the world, right? We're all perceiving the world via our own filters. So as
one example, there's kind of a raging controversy now about neutrinos, which are these very small,
very light particles that actually there's about a billion of them that just passed through my hand
that were produced in the early universe. We don't know their masses. You and your colleagues
have done experiments which lead to the conclusion that their masses are smaller than they should be
according to experiments that have been done by particle physicists here. And when you raise that
to the particle physicists, they just don't believe the cosmology, basically. So I think
that's an example of a whole class of smart people who don't necessarily buy into this field, really,
or all of the field, you know. I mean, it's partially natural historically. You think,
you know, we've never detected a new particle, weighed the mass of a particle, except in this
accelerator or something like that. So to use the cosmos as your accelerator, which is natural to
people like me, but yeah, it's very sociological in science that it's different. Have you seen
examples of that? Yeah. I mean, I've asked people about that. Like my particle physics,
you know, friends, Don Lincoln I've talked to is in your neighborhood, right? Will you believe it,
when a particle physicist sees a cosmologist say, here's the mass of the nutrient.
Now we're working on a paper with Shasha Arani that you met over lunch, that really seems to
suggest that we're going to need all three different types of things. Neutrinos, double
beta decay. We're going to need laboratory experiments and we're going to need a long
baseline. We're going to need cosmology. And that whole spectrum actually will make the case much
stronger than only the cosmologists see it. And no, give up all hope. Let's talk about the
mechanism. That's sort of where I first got exposed to you. Very intimidating. It was my
second year of grad school in 1994. Where did this come from? This is at Brown. I was at Brown
and I was trying to understand, well, dark matter. I'm still trying to understand that. We're going
to talk a lot about that. And these are called sterile neutrinos. So first of all, what the
hell is going on here? What are neutrinos? Let's do a recap. Neutrinos, flavor, oscillation. What
does it mean? Where are they oscillating? Are they jiggling around in here? What's a sterile
neutrino? What's a maranon particle? What's a dirac particle? Let's go through it. I want to
get my money's worth. Flew you all the way out of Chicago. Ten questions. And that's my forte.
A neutrino. People are familiar with electrons because that's what they're made of. And so
particle physicists tend to think of neutrinos as being partnered with electrons. So the fundamental
theory of nature says that every thing like an electron has to have a neutrino associated with
it. So for example, you mentioned muons earlier. Muons are kind of cousins of electrons. They also
have their own neutrino with them. There's another thing like that. All these things are called
leptons. There's a tau lepton. It has its own neutrino. So neutrinos are associated with
electrons in this case, or electrons, or muons, or taus. Whereas the electrons are charged,
they have a negative charge. Neutrinos do not have electric charge. That's why they're so hard
to detect. Whereas neutrinos have a mass that we know. Neutrinos have masses with at least a million
orders of magnitude. A factor of a million smaller, probably a billion smaller than the electron. So
that's what they are. They're very, very hard to detect. However, just like the electron,
they do participate in what's called the weak force. And that's important because there are
various decay processes that are important to life and everything that produce neutrinos. So for sure
neutrinos exist. We've seen them. But they only interact very weakly, so they're very hard to
detect. So that answers your first question. Those are called flavor states. The partnering,
pairing between the subatomic elementary particles, mu, tau, and electron, those are the flavor
states, but they are not the mass states, right? Exactly. So what I mentioned is that these
neutrinos, the electron neutrino is paired with the electron. So that's one type of neutrino.
But if you imagine the possibility of a given quantum state that is a superposition, probably
you've had a lot of quantum computing people onto, people are familiar with superposition,
a superposition of an electron neutrino and a tau neutrino. So that's possible. And it turns
out that a superposition of those neutrinos are actually eigenstates of mass. That is,
they are eigenstates of mass. That means the mass eigenstates are the things that propagate
through space. And respond to space-time curvature, for example. Right, exactly. If there was no mass,
then there would only be this one basis, the flavor basis. But because there's mass,
things get mixed up. It's kind of a quantum mechanical effect. Because they get mixed up,
it's possible, and this was first detected from neutrinos from the sun, that neutrino could be
produced in the sun in one type, one flavor, and be detected as another flavor. That's the
oscillations you mentioned. So I hit two of your questions. Yeah, sterile neutrinos. Sterile
neutrinos. Okay, so each of those things are a type of neutral particle. It turns out that it's
quite possible that there is another neutral particle associated with them that does not
experience the weak force. So neutrinos are not charged, so they interact very weakly,
but they do interact. A sterile neutrino is even less weakly interacting than that. It's
completely divorced from the weak force. It's not produced in decays, etc. Why do we think they
exist? There's kind of a complicated technical reason for it. That is, if you think about
electrons, they're made up of left-handed spinning states and right-handed spinning states. The
neutrino that we know are all left-handed spinning, so we kind of expect there to be
right-handed spinning states. Also, those are sterile neutrinos. That's a sterile neutrino.
And the sterile neutrino was hypothesized to be, via the Donaldson-Widrow mechanism,
a potential, but by no means confirmed, dark matter candidate. Now, let me say one thing.
I tell people we've detected dark matter. I told Neil deGrasse Tyson this to his face,
and he was astonished. And I said, they're neutrinos. They satisfy every possible property,
except for the fact that they don't make up all of the mass that is seemingly required to explain
the dark matter gap between the amount of luminous matter and the amount of total matter that we see.
So that's like saying, well, it doesn't solve dark matter. That's like saying, well,
hydrogen doesn't explain all baryonic matter. It's irrelevant. There might be a whole,
as George and as past guests on the podcast, Mike Turner and others have said, there might be a whole
periodic dark matter table, the dark periodic table. So am I wrong? Should I shut up about
this dark matter detection already occurring? Well, let me challenge you with one thing.
Have we detected cosmic neutrinos? We've detected, I believe we've detected.
Oh, we have. Yeah. Yeah, we have. Yeah. You have detected them.
Not direct detection. Not direct. You have indirectly
detected them. So yes. So you have indirectly detected neutrino dark matter. There's a
difference between direct detection, where you actually build a detector that tries to see these
cosmic neutrinos, a billion of which have just passed through my hand. That, do you know Joe
Formaggio from MIT? So he has told me that every experimentalist in their career goes through a
period of two years where they try to detect cosmic neutrinos and then they realize it's
impossible. And unfortunately their startup runs out. Yeah, exactly. Anyway, but you have detected
that you and your CMB colleagues have detected them indirectly. If they weren't there, then the
pattern of anisotropies that we see would look much different. So kudos for a hundred percent
agree. Yeah. So what is the Donaldson-Woodrow mechanism? In the late eighties, early nineties,
there was a guy named John Simpson and he provided evidence for detecting in the lab,
a neutrino with a mass of 17 kiloelectron volts. Did you ever hear about this? No. People were
astonished. It turned out to be an experimental artifact due to magnetic fields. But for about
three to five years, everyone was talking about it. As you probably know, such a heavy neutrino
makes a lot of problems for cosmology. So I started thinking about what would happen,
would they be produced and stuff? And so then Larry and I came up with this idea that
putting aside the 17 kV neutrino, which turned out to go away, maybe it's possible that the
ordinary neutrinos in the very early universe oscillate into these sterile neutrinos. And
maybe they have a mass, so produce enough of them so that they constitute the dark matter today.
I still think it's a good idea because we know neutrinos exist, right? As opposed to every other
dark matter candidate around where we don't know exists. So to some extent, my prior on that is
higher than other things, but of course I'm pretty biased. So yeah. Talk about the mechanism. There's
something called a mixing angle, which is quite beautiful when you think about it, that these
states, these quantum mechanical eigenstates, which are superpositions, that they have this
ability to effectively rotate, just like an ordinary rotation of a ball or any object in
two-dimensional order. It's the simplest thing you could imagine. I guess one-dimensional would
be simpler, but they can basically be thought about as rotating in this abstract space.
How do you think about that? How do you visualize this? Is it purely a mathematical thing in a
theorist's brain? Yeah, that's a great question. And you're absolutely right that you don't need
flavors or anything like that. It's actually just a two-dimensional space, the regular neutrino and
stereoneutrino. Actually, this comes back to the way different people perceive and think about
different things and how everyone's opinion, everyone's brain, sees things differently.
So my mind works best very linearly and mathematically. I'm not good spatially,
so I just think about a two-by-two matrix, which is a pretty simple mathematical thing.
So I just think of these things as, if you take a two-by-two matrix with off-diagonal elements
and diagonalize it, that's what I think of as an oscillation. But people like, I'm sure you think
of it more in a physically intuitive way. I don't think I have that physical intuition.
So what's physically happening in the way I think about it? I do think about it mathematically also,
but in quantum mechanics, phase is an important quantity. Even though we can't directly measure
phase, you can measure phase differences, right? And so, you know, stop me if I get this wrong,
but energy differences lead to phase differences, which then can be imprinted,
and you can get physical oscillation. I mean, we get oscillations of solution to the solar
neutrino problem, right? It was effectively this oscillation, both abstractly in the phase space
of quantum mechanical rotation space, but physically these things are oscillating as
they travel, you know, if the distance between the earth and the sun were different, we would
have gotten a different answer, right? We could have been in this weird position where it exactly
always came out to be an electron or something like that, we just didn't happen to live there,
right? So how do we explain this, that the neutrinos interact with matter,
but sterile neutrinos wouldn't interact with matter, say, in the same way?
It turns out matter plays a quantitative role in it, but it's not, I don't think it's
a qualitative thing. So I think the qualitative thing is exactly what you said, that in the solar
neutrino issue, the electron neutrinos that the sun produced convert as they travel into muon
neutrinos that Ray Davis, et cetera, did he win a Nobel Prize? So that is an oscillation in a
two-dimensional space. It's the exact same thing without the flavor thing. The two dimensions are
that the regular neutrino and the sterile neutrino, the same exact process that we think might've
happened in the early universe to produce, we had a lot of regular neutrinos around in the early
universe, they oscillated and produced these sterile neutrinos that could be the dark matter
today. So it's the exact same process. Now, a year ago, Kyle Dawson sat in that very chair with Dan
Green and we had the spirited conversation about DESI, right? And a lot of the conversation was
a little bit, he's a very statesman-like and wonderful scientist and just exceptional person,
but I detected a little bit of hedging. He's saying, yes, there's a 4.2 sigma tension.
And I often say, we got the Hubble tension, we have the sigma eight tension, now we've got the
W tension. So what do you make of these different tensions? First of all, I've had eminent scientists,
including a partner of one of your former partners, Mike Turner, Lawrence Krauss,
alleges, he sometimes says he came up with dark energy and on his weaker moments, or maybe he's
being more accurate than I'm giving him credit for, but with Mike Turner, that they sort of
independently came up with some ideas that suggested dark energy was real. And he doesn't
believe it. Lawrence Krauss said that he thinks they're wrong. It's a cosmological constant.
Now, Einstein was wrong once before. It's too bad. He could have had a good career, right? But tell
me, Scott, where do you come down on this? Is it a legitimate tension? Where does it rank in the
that I mentioned? Sigma eight, which is clumping a matter, Hubble tension, which is disagreement at
early times and late times, and now this new tension between dark energy and a cosmological
constant. Just to focus on one thing, this S8 tension you mentioned, that's what I've been
spending the last 10 years on, this idea that, think about Manhattan in 1790. There were 50
people every square mile in Manhattan, very overdense compared to the rest of the country.
Today there are 50,000. So why is it that the inhomogeneities grew like that? That's a fascinating
question, right? Could someone in 1790 have predicted that there will be exactly 50,000
people? No way. But we've done that. Lambda CDM predicts that what you measured in the
microwave background evolves to be precisely the inhomogeneities we see today, that precisely
modulo the fact that it's off by about two sigma or something. So that's what you call the S8
tension. So I guess my question back to you is, is it tension or is it, wow, that is unbelievable
we're able to get that close. I'm kind of depressed that it's not a five sigma tension,
because then it would be whatever, just like kind of Kyle's thing. But on the other hand,
we're so close on this incredibly amazing story that we've created. We've created a story of how
we got here. I guess a pushback, I would say. Sometimes I feel like we got to grow an extra
arm to pat ourselves on the back, and you're doing a great job of that with the experimental
kind of kudos to us. But by the same token, I have kids, you have kids, right? And you remember when
your kids were young, you'd take them to the doctor at age two, and they'd measure their height,
and they'd say, basically, they're going to double in size, right? And how do they know that? Well,
they've had about 100 billion humans have lived since time equals zero when Og and Magog came
out of the cave somewhere 250,000 years ago. They estimate about 100 billion people. That's
pretty good statistics on this, right? Now, we only have one universe, and we always complain
about how hard it is to be a cosmologist, world's smallest violin. There's only one universe.
I always say, at least astronomers have 100 trillion stars perhaps in the observable universe,
so there's a lot of different statistics that they can do. You're making my point that because
there's so many regions of that size, you and your colleagues are able to measure it at one
time in the universe, and you've measured it incredibly precisely to a few percent,
that's what your point is, that there's so many different regions that you can measure it so
precisely. And we haven't measured quite the area you have, but we've measured a lot, and we can
measure it at a time 13.7 billion years later, and we get an answer that's 10,000 times bigger,
but precisely on what the prediction gives. So it's the precision that, to me, is astonishing
and is a great success. So why do you still work on it? What's left to unravel? I mean,
I'm not an expert in some way. Your time is very valuable, right? So
why are you still so deeply invested in it? Yeah. So you're looking for something that,
if you found them, these B-modes and the cosmic right backward would prove,
beyond a shadow of a doubt, inflation, something that happened in unimaginably small times,
it would be incredible. The rest of us are focused on unkinging Lambda CDM,
like basically just dethroning it. So there are people like Kyle who do it in one way,
by trying to figure out whether the distance that they measure to distant places agrees with what
you guys in Lambda CDM predicts. And I've been doing it in this other way, by seeing whether
that precise prediction, which is so precise, is correct. And so the answer is, it's still,
it's called an SA tension. That is, we're not exactly right, so I overstated a little bit,
but we're, so we're two and a half Sigma off, which means there's technically only a 1% chance
that theory's right, but we kind of know that that's probably, yeah, yeah, yeah.
Millions of things happen at the end of the day here.
So I mean, exploring the tension, killing the model is my dream.
Is that because there are certain dependencies on Sigma-8 that go as like, what, Sigma-8 to the
seventh power? There's some interesting, either it's in the power spectrum or, I seem to recall,
aren't there some very crisp tests that are available because of the deep sensitivity on
Sigma-8? Like in other words, if you nail Sigma-8, you can get other things to extremely,
eight times higher precision.
I mean, that's probably true for galaxy formation, the halo mass function and stuff, but that's,
that's not what I'm most, I'm most interested in measuring this very simple statistic. It
actually comes back to something you mentioned earlier. We're doing it with the simplest of
statistics, what's called the two-point function or the power spectrum, which is some of the same
thing that you use. And I wonder if we're going to be supplanted with AI and opticians that use
more of the information. So essentially what we're doing is we're compressing the data
into a hundred numbers, right? And from those hundred numbers, we're extracting out this one
number and comparing to the one number that you guys measure. So that may not be the best way to
test the theory. There might be better ways to do it. So people are thinking about that.
Higher order functions.
Last time you gave a colloquium here, which is, you know, decades ago or now, but it was
extremely well received. You made this beautiful kind of point about the discovery of dark matter,
you know, in our solar system, which was, uh, goes by the name of Neptune and Neptune's
discovered by Le Verrier famously use. I went back and I looked at his paper from 1843 or whatever.
And they basically, you know, it's, it's just this lionizing hero worship of the, he discovered a
planet with a pencil, you know, it's basically just this lone genius that set this thing.
And then he made the case that like, okay, so that was discovery that there was this
weird sort of effect that was happening to the planet, George Uranus, which would,
I think still should have been called George, but, but anyway, he took us back to that time,
you know, when, when it could have been some Primal Newton, Isaac Newton's gravity,
or it could have been dark matter that was just unseen and turned out to be dark matter.
Then he made the analogy, let's fast forward to 1911, 1913, 14, Einstein's coming up in GR
and Le Verrier and the acolytes of Le Verrier were still actually Le Verrier was proposing
that Mercury was being perturbed by an unseen companion, just like Neptune called Vulcan.
And then you said, well, that's how he was wrong, right? So the hammer theory is the,
everything looks like a, like a nail when all you have is a hammer. Where do you think we are with
dark matter? What is your take? I've had Mordecai Milgram on the podcast. What do you make of the,
you know, kind of alternative, you know, history rhyming again and again,
looking for dark matter, uh, looking for changes to gravity. Where do you think we're going to end
up? What do you, what do you, I mean, you worked on this for a long time. Where do you, what is
dark matter? Yeah, I have no idea. I actually had this debate with Stacy that this is an organization
that promotes civil debates. So we had this debate and I don't, so just to give context,
Stacy McGaugh is one of the leading astronomers who doubts the existence of dark matter and he
and Mordecai Milgram and other people, most notably from a theoretical perspective in Israel,
a Mexican-Israeli physicist, Jacob Bekenstein, put forth alternative models of gravity that would not
require dark matter. The problem with those is they don't explain what you guys have seen.
And to some extent, it's another example of what we've been talking about. It all depends on your
filter. What they look at is, oh, I can see that galaxy over there. Let me look at that and I can
fit this better with my theory of modified gravity. What we tend to think, what I think is
simpler is the stuff that you measure because there's no people there. It's just electrons,
protons, and photons. It's really simple to understand. But from their perspective, that's
like a gazillion miles away. How can we possibly understand that? So it's two different lenses on
the same universe. And from their perspective, we're never going to convince them that the,
you know, this distant stuff has any information about what is guiding stars in a galaxy near us.
So I don't think there's much trouble bridging that gap. However, Stacy, who's a very reasonable
person, does not believe Mond is correct. And if you look at the generalizations of Mond,
such as the ones that Bekenstein set up, they're actually so close to theories of dark matter,
they introduce new fields. It's the same thing, basically. They're just introducing new stuff.
And so I actually think another possibility is that we're just looking at everything wrong,
that someone's going to come around and say, we have to just rethink everything because
what we've done in order to make the story of the universe work is introduce dark matter,
introduce inflation, introduce dark energy, and none of this has been found in the lab. So,
you know, how many free passes do we get? I don't know.
No, it's extraordinary that you're, again, candid and you're honest and you're willing to,
you know, kind of admit where these lacunae, these gaps, these flaws. But, you know,
I often say the most exciting thing, and you hinted at this earlier, is a flaw, right? Because
I tell my students a flaw leads to a law, and we just don't know it yet. Like the Big Bang was,
you know, kind of a solution to a lot of the problems in the steady state model.
The inflationary universe is sort of a patchwork of fixes and, you know, kind of home improvements
to the Big Bang model, fiducial model, and lambda CDM, and there's alternatives to it.
But one thing I thought about, you know, 61 years ago this month or last month, the CMB was
discovered and the announcement was discovered. And many people, including Jeff Burbage, who used
to occupy this office, and his longtime colleague, Fred Hoyle, and the recently deceased, and also,
I'm very proud to say, past guest, Johanna Narlikar, you know, they went to their graves
believing in the steady state, in a quasi-steady state. They added some stuff to it. They were
eminent scientists. I mean, these are people that were Nobel-worthy, adjacent to Willie Fowler,
did win the Nobel Prize for the, you know, BBFH, and incredibly eminent scientists. Hoyle,
of course, coined the term Big Bang as a pejorative. And my question to you is,
when the CMB was announced, Hoyle said, well, they found it to be, you know, 2.7 Kelvin, but
if it had turned out to be 27 Kelvin or, you know, 270 Kelvin, they would have explained that too.
And it reminded me of this quote that I have in my book from you, from eminent cosmologist Scott
Dottleson. During BICEP2, you said, inflation can produce a B-mode signal as small as you like. And
I kind of used that in the rhyming sense with our friend, Fred Hoyle, who said, you know,
they would have found any value. So it was almost like anything that we found would be consistent.
And you're not the only one who said this, by the way. And many people said things much, much
more hyperbolic, like, hello, multiverse, Max Tegmark said. And Lawrence Krauss said this now
means there's proof that God does not exist. It was an incredible time for me personally,
obviously. I talk about a lot of these in Losing the Nobel Prize, my first book. But what do you
make of that? If a theory inculminates any result, is it a theory? I mean, what is it?
What are we to make of it? Yeah, I don't want to go too far towards ignorance and the fact that
smart people don't believe in the microwave background doesn't mean they're correct.
Yes. Sure. I think here, there's a sidebar. I think the human need for meaning is the most
paramount need that humans have. Viktor Frankl's Manifest for Meaning, right? I want to just kind
of take this counterfactual, hypothetical, you know, 10 years from now, 2036, DESI's done,
Simon's Observatory's done, CNBS4 never happened, but unfortunately, rest in peace,
it should have happened. Lightbird, you know, there's many different projects, hopefully,
that will augment and complement and compete with us. And so, I think the most healthy thing is to
have a competitive scientific environment. And so, I welcome our, you know, kind of competition
from China. We have a Chinese team that's trying to do us, so. And there's many other great and
brilliant scientists. But we're there. What is, like, cosmology being done look like? What is
the solved model of cosmology look like? And is it on the horizon in our grad student careers?
Okay. I have no idea. But. What would satisfy you?
Let's agree that cosmology is a millennial-old pursuit, maybe more. Maybe those people from
75,000 years ago, maybe they also. And we've been just incredibly fortunate to live in a time where
the, you know, the amount of information we've gotten has been, you know, extraordinary about
the universe. We have very good reason for believing that the universe is not on the back
of a turtle, right? We're not any smarter than the people who made up that theory, right? Because we
have data. So, we've just been incredibly fortunate. But it's still a thousand-year-old science. And
it's probably going to be evolving. So, the chance that we're going to identify dark matter in my
career is, I would say, it's pretty small. The chance, well, you tell me. I mean, the chance
that we find B-modes, so you're going to be extending the reach by a factor of 5 to 10,
is that right? So, that, I mean, you have to have a prior on what inflationary models produce that.
I mean, it would be amazing. As you probably know, I went all Goo Goo Ga Ga the first time
you guys reported something. So, it would be, you know, it would be amazing. We have to be clear
about what we know and what we don't know. If you guys discovered B-modes, that would not be the end
of our studies of inflation. Then we'd put up a gravitational wave thing in space to detect
gravitational waves at higher frequencies, right? For sure, what they used to call the Big Bang
Observatory, right? So, we would definitely hone in on those B-modes because they would teach us
about physics 12 orders of magnitude larger than can be probed at the Large Hadron Collider. So,
that would open things up. In dark matter, if we detected something in the 100 GeV range that
pointed to supersymmetry, that would open up the pathway, I think, to more colliders being built
to understand things. So, I think the possibility of, especially this marriage of quarks in the
cosmos, the possibility of discovery would be unlikely to shut things down. But as I said,
I really don't know. I tell my students a flaw leads to a law, and we just don't know which law
yet. Scott spent a decade in the trenches working on one flaw, whether the lumps in the universe
grow the way the theory says they should. There's a second way the stress tests the
same model. A year ago, Kyle Dawson sat in the same chair and told me what DESI found,
4.2 sigma. If Scott's crack and Kyle's crack are the same crack, lambda CDM is finished.
That conversation is on screen now. Click it, watch it, subscribe and share it, and then tell
me which tension you'd bet the most on. I'm Brian Keating. This is Into the Impossible. Tune in next time.