If You Only Trust What You Can See, You Miss Most of the Universe

Lisa Randall is a theoretical physicist and cosmologist at Harvard. Her work runs across particle physics, supersymmetry, baryogenesis, cosmological inflation, and dark matter. Most people hear the phrase "dark matter" and picture something occult. What she actually wants to stress is more basic than a missing particle. Physics, she tells Lex Fridman, teaches both our limitations and our abilities: we can deduce the existence of something we do not directly see, and that is a tribute to us. It is also a warning not to over-rely on the senses.

If you relied only on what you see directly, she says, you would miss so much of what is happening in the world. Dark matter is the cleanest case. We cannot see it, yet it is not a single clue. In *Dark Matter and the Dinosaurs* she walks through eight or nine independent ways physicists deduce not just that dark matter exists but how much of it there is, and they all agree. The method is gravity. A single particle's gravitational pull is tiny — gravity is extremely weak compared with the other forces of nature — but there is a lot of dark matter out there, carrying about five times the energy of the matter we know in atoms.

How should we think about it, then? As another form of matter that does not interact with light, or at least not as far as we know. It interacts gravitationally. It clumps. It forms galaxies. It just does not shine, which means that before gravitational-wave detectors, almost everything we detected, we detected because of some interaction with light. The fact that we do not see it, Randall says, makes it no less legitimate. It just means we have a harder job finding out exactly what it is.

Randall's first correction is not to treat dark matter as mystical. Gravitationally it behaves like other matter, but it does not experience electromagnetism, which is why it has a different distribution. In our galaxy it is roughly spherical, unless it has interactions of its own. Ordinary matter can radiate, cool, and lose energy, then, under conservation of angular momentum, collapse into a disk instead of falling all the way to the center. That is why we see the Milky Way disk, and why dark matter, lacking that radiation channel, stays puffed out in a sphere.

Do not assume ordinary matter is the protagonist. Dark matter collapsed earlier and more readily. Ordinary matter has radiative forces that make collapse hard on small scales, so it is dark matter that drives galaxy formation; ordinary matter comes along afterward. There is also simply more of it, so its energy density begins to dominate over radiation sooner than ordinary matter would. In the book she jokes that when we look at a building we think of the architect and forget the workers who did the grunt work. Dark matter did that work in the formation of the universe, and we forget it because we do not see it.

Five times more energy in dark matter than in atoms can make our 5% share sound like leftovers. Randall flips the comparison. We are limited and human, and the fact that the matter we know constitutes 5% of the energy density of the universe — about one-sixth of the energy density in matter — is remarkable. Anything could have been out there. The universe we see is a significant fraction. Physics, she adds, is useful here because so many of our words are based on how we see the world. It trains you to build intuitions that apply at different distances and scales, past immediate visualization.

Fridman brings up the claim that dark matter might be tied to extinction events: gravitational disturbances in the Oort cloud at the edge of the solar system, then a higher rate of objects hitting Earth. Randall is careful before she enjoys the idea. This was a speculative theory. Physicists usually assume dark matter is one thing, basically non-interacting aside from gravity, or very weakly interacting. She wants to get outside that human mindset and ask what else could be there. Our own world is not a simple model. We have nuclei, electrons, forces. Maybe dark matter also has some structure.

The suggestion is that a fraction of dark matter — not all of it — might have interactions of its own. If some of it radiates, and she likes to call that radiation "dark light" because it is light we do not see but dark matter would, it could cool and collapse into a disk the way ordinary matter collapsed into the Milky Way. It could conceivably be a very thin, dense disk. People have since looked for such disks in stellar data. The target is interesting because it is something you can measure. Positions and velocities of stars tell you the structure of the Milky Way.

Then the fun proposal. The solar system orbits the galaxy and, as it does so, bobs a little up and down, like horses on a carousel. Each time it passed through a dense dark disk, the chance of dislodging something from the Oort cloud would go up. At those times you would be more likely to have cataclysmic events — including the one that caused the last extinction we know of for sure, sixty-six million years ago. It was not amazing for the dinosaurs, Fridman notes, or for two-thirds of the species on the planet. It was amazing for mammals, and for us.

What she loves about the story is that it sounds like something a five-year-old would make up: maybe the dinosaurs were killed by a big rock that hit the Earth. Then a serious scientific account grew under it, pulling in geology and the then-novel idea of a cataclysm rather than slow change. She likes the dark disk for the same reason. As far-fetched as it might sound, you can look for the observational consequences and test whether it is true. Critics later said the theory made dark matter weirder than it already is. Her reply is that we have no idea how weird it is. It is not already anything. There is no normalization. People think they know what dark matter is. They do not.

One of the main points of the book, Randall says, and one of the reasons she wrote it, is that she thinks we are abusing the planet, changing it far too quickly. When you alter something, it is good to think about the history of what it took to get here. Structure had to form, then galaxies, then the solar system, then the planet, then humans. We are in some part the result of a large object hitting Earth, dinosaurs going extinct, and mammals developing. It is an incredible story. Something else might come of it, she tells Fridman, but it will not be us if we mess with it too much.

The shape of things still matters for the test. The visible Milky Way is a disk you can see from a dry, dark place. Dark matter is believed to be more or less spherically distributed, and because it is a sphere there is a lot of it sitting there even if little of it is in the disk. The reason it does not collapse, as far as we know, is that it cannot radiate the same way. The suggestion was that some components might. People have seen some evidence of disks of certain densities. These are questions worth asking: if you can figure it out from existing measurements, why not try?

Not all dark matter need be made the same, but that is only a possibility. We do not know what most of it is, or what a fraction is, because we do not see it. Detection of the kind she has been describing would be gravitational. Other candidates have other signatures. Axions are one. The most popular candidate until fairly recently, because no one has found it, is WIMPs: weakly interacting massive particles with masses around that of the Higgs boson, which would give about the right density of dark matter and a better chance of seeing it if it is connected to the Standard Model. That is also a better chance of ruling it out. Underground XENON detectors are still tightening the bound. The Large Hadron Collider was originally one of the hopes. Given what it has already done, she now thinks a discovery there is very unlikely.

Asked for the LHC's biggest insight for humanity, Randall gives a split verdict. It was a major victory. The Higgs boson was proposed fifty years earlier and then discovered. The Higgs mechanism seemed the only way to explain elementary particle masses, and it was right. On the other hand, she has been in physics long enough to hear it as a cautionary tale. When she started, the United States was proposing the Superconducting Super Collider. A lot of physicists, particularly in Europe, said the LHC would have the energy reach to discover what underlies the Standard Model. Many thought they knew what would be there. It happened to be supersymmetry.

That confidence was one of the factors, she thinks, that left the LHC as the only machine in town. The Superconducting Super Collider, if it had achieved what it was supposed to, would have been a much more robust test of the space: higher energies, more events, a more definitive chance to find the next layer or rule it out. For humanity the collider is a tribute to discovery and to believing in things enough to go look for them. It is also a warning. You want to believe in your theories, but you also want to question them at the same time, in ways that make you more likely to discover the truth. Do not assume things before they have been found.

Fridman adds that it is also a grand engineering effort, and a lesson that you could go even bigger. She is glad he said so. The LHC is impressive technologically, impressive at the engineering level, and impressive in that so many countries worked together on a long-term project people actually committed to and made happen. It is a demonstration that when people set their minds to something, they can do something amazing. Bureaucracy, politics, and economics can all speed a machine up or slow it down. Science is a way to make progress; politics is often a way to slow it. She does not want to overstate that, because without politics the LHC would not have happened either.

Sometimes, she says, it is simply good to have a project people work on together. She mentions SESAME, an effort to have Palestinians and Israelis work together in science. Forget the politics and focus on a particular project; sometimes that can work. CERN itself was formed after World War II. A lot of European physicists had left Europe, and they wanted to see Europeans work together and rebuild. It worked. One of the major problems, she often thinks, is that we just do not meet enough people. When someone seems like the other, it is easier to forget their humanity. Connections of that kind matter.

Given all the scales that led to the dinosaurs' extinction, does she worry about future ones? She does think we might be in the middle of an extinction right now, if you define it by the number of species being killed off. It is subtle because the system is complex. Sometimes things evolve; sometimes animals just move. The way we have developed the Earth makes it very hard for species to move somewhere else, and we are seeing a version of that with people too. We do not think about the future very much. We think about what we are doing now, and we do not think enough about the animals we are destroying, the things that are precursors to humans that we rely on.

Fridman wonders whether the real threats are the gradual ones we watch or the sudden ones we barely see coming — asteroids, nuclear war, something that unfolds over days rather than decades. That delay in recognizing catastrophe, she notes, is part of why it took a while to determine what caused the last extinction. Many people thought things were more gradual. Extinction itself was once a novel concept. These events are not always predictable except on a grand scale. People were pretty aware that nuclear weapons were dangerous. She is not sure they are as aware now as they were twenty or thirty years ago, and that worries her.

She was not as worried about AI as other people. Now she understands. As soon as you create things you lose control over, it is scary. The other thing the present is teaching, she says, is that it takes a few bad actors. It takes everyone to make things work well, and not that many things to make them go wrong. Disease is like that: finding the cause is not the same as making things better. For things to be healthy, a lot has to work. For things to go wrong, only one thing has to go wrong. The same is true of democracy. A lot of what we rely on are delicate equilibrium situations. We try to build in robustness. A few extreme events can still alter things. She thinks people are not as scared as they should be about nuclear weapons, a more serious danger than they realize. They are a little more scared about pandemics than they were, and still not enough.

She received an award from a museum of nuclear history and technology in the Southwest, went to visit, and found it was mostly a museum of nuclear weapons. The scary thing is that they look really cool. You have the "this is scary" feeling and the "this is cool" feeling at once, and we have to get around that. She had a tour from people who had worked there in the Cold War, and one or two from the Manhattan Project, and realized how the thing itself gets you excited. You are not thinking about the overall consequences. It was like early Silicon Valley: people thinking "what if we did this?" without keeping track of peripheral consequences. She sees that happening with AI now, full speed ahead. She quotes Rilke: beauty is nothing but the beginning of terror which we are still just able to endure. That applies to nuclear weapons. At a more mundane level it also describes the edge scientists live on, not knowing the answer and not knowing if they will find it — frustrating, not Earth-ending, until some ideas turn out to have bigger consequences than you think.

Is it not still terrifying, Fridman asks, just the size of things: the amount of dark matter, the black hole at the center of the galaxy? Randall's answer is almost a confession of temperament. This might be why she is a physicist, or why she is different from other people. She is not such a big fan of humanity in some ways. The idea that we were everything would be really boring. She loves the idea that there is so much more out there, a bigger universe, lots to discover, and that we are not all there is. Would it not be disappointing if we were? We have no idea how much is out there. We know what we can observe so far. That there is other stuff yet to figure out is exciting, not a source of dread.

Fridman asks an out-there question. Life on Earth is a pocket of complexity that needed a pile of conditions and Darwinian evolution. Could there be pockets of that sort inside dark matter — chemistry and biology evolving in ways we cannot see? Possible, she says. It is one of the things that would be true if the interactions they suggested exist, though it would need more complex ones, and we do not know. The conditions that give rise to life and complexity are themselves complex and unlikely, so it is not as if the odds are great. There is also no reason to know that it does not happen. It is worth investigating whether other forces exist in the dark-matter sector.

As far as we know, that sector does not have the forces of the Standard Model. It might at some low level. It could have its own forces, just as dark matter might not experience our light and might have light we do not experience. There could be other kinds of forces even within our sector, too weak to have been discovered so far, or existing at different scales than we know about. We detect what interacts strongly enough with our detectors. That is one of the reasons we build big colliders: to see whether other forces and particles exist at higher energies and shorter distance scales than we have explored. It is not just a dark-matter question. Even in our sector there could be a whole bunch of stuff we do not yet know.

The Standard Model of particle physics, Randall says, tells us about nature's most basic elements and their interactions: nuclei and electrons, protons and neutrons, quarks held by the strong force, the weak nuclear force, electromagnetism, and now the Higgs boson, associated with how elementary particles get their mass. There is a weird array of masses — not just up and down quarks but charm, strange, top, and bottom; not just the electron but the muon and tau; neutrinos still under intense study. In particle experiments gravity can usually be ignored. Dark matter stands outside that list. It is not interacting through those forces, which is why we do not usually see its effects except that there is so much of it. The Standard Model has worked spectacularly well and been tested to high precision. One of the things physicists do is want it to break down at some level. They look for the precision, or the energy, where it is no longer working, because those deviations can tell us what underlies it.

Where do you look? Higher energy, including beyond the LHC, because higher energy means shorter distances and particles you did not have enough energy to make. Precision is the other place. If the model has been tested to 1%, look at a tenth of a percent. Some processes should not happen at all in the Standard Model, or only at a very suppressed level; those are worth hunting too. The big open questions include why the particles have the masses they do, and why the Higgs is so light compared with the mass it might have been expected to have from special relativity and quantum mechanics. How much of a structure is environment versus initial information is, she notes, a big question in pretty much every field — including whether our universe is one of many self-contained gravitational systems we do not even interact with.

Fridman then reads her disagreement with Carlo Rovelli. Rovelli wrote that electrons do not always exist; they exist when they interact, materializing in a place when they collide with something else. Randall's reply is that stocks may not achieve a precise value until they are traded, but that does not mean we cannot approximate their worth. Electrons might not have definite properties, but they do exist. It is true that an electron does not exist as a classical object with a definite position until the position is measured. Something was there, which physicists use a wave function to describe. She likes to think of the universe as being out there whether or not she looks. It would be really weird if things came into existence only when she saw them. She could believe the Middle East does not exist because she is not there now. That would be ridiculous. There is only so much we can attribute to our own powers of seeing.

The disagreement is not even about the Standard Model. It is about how you interpret quantum mechanics. She would say those wave functions are real. Quantum field theory says electrons can be created and destroyed — that is what happens at colliders — but that does not mean the electron in an atom is not there. Its charge is there. Is there a bottom turtle? She might not seem so, but she thinks she is more humble than a lot of physicists. She is not sure we are ever going to get to that bottom level. She does think we will keep penetrating layers. The notion of distance itself might break down. When she says we have measured everything, she means it with a grain of salt: in everyday life, at the level of known ingredients. Consciousness, and many other phenomena, sit at higher layers. When she throws a ball she does not tell you where every atom is; she says there is a ball. For a far-out theory such as panpsychism — consciousness as a fundamental force — the thing you do is test the possibilities within the constructs that exist. You do not just jump to the most far-out option. You can, but then you need evidence, or a showing that the phenomenon is impossible without it. We are very far from that.

What are the limits of science? She is smart enough, she says, to know she has no idea, and it is not even clear what science means. Particle physics looks for fundamental things. Biology asks different questions at a higher level. In a more numerical age, or with AI, answering a question might mean predicting or reproducing, and that shifts the definition. It would be premature to declare limitations. At the beginning of the twentieth century we did not know the universe was expanding, or about quantum mechanics, or about special relativity. That is a lot in a relatively short time. Various people have declared everything solved before. Various people were wrong.

Top-down theoretical physics starts from a theory you think is so perfect that you can predict everything from it, or at least some salient features. Bottom-up is closer to the mass questions they had just been asking: we measure things and want to put them together. A good approach is usually to combine the two on a specific question. The community tends to fragment into people who do one or the other. Some of her best collaborations have been with people more top-down than she is, so they came up with ideas neither would have thought of alone. Fridman asks whether the truly big leaps happened top-down, like Einstein. Einstein, she says, was not a top-down person in the beginning. Special relativity was him thinking about how Maxwell's laws could make sense given the symmetries they seemed to have. He was very much bottom-up, and he resisted top-down for a long time.

When he tried to incorporate gravity into general relativity and needed feedback, he was helped by a mathematician who had developed differential geometry and helped him write the equations down. After that he thought top-down was the way to go, and he did not actually make that much progress. It is naive, she says, to think it was just one or the other. A lot of people who made real progress were rooted in actual measurements. Physics and mathematics overlap more than outsiders expect, but the questions differ. Mathematicians like the structure itself. Physicists concentrate on the consequences for the world. Insights can be beautiful. If you saw Einstein's equations written out in components you would not think them so beautiful; written in a compact way, they look nice.

Successes and failures of string theory? Talking about any science in those terms often misses the point, because there is not some absolute endpoint. She does think string theorists were a bit overly arrogant in the beginning, thinking they could solve many problems they were not going to solve. That is not to say the methods and advances do not exist. They were not able to immediately solve all the problems they thought they could. The theory has given tools and insights, but it becomes almost a sociological question of how much effort should go one way or the other. You can get caught up in the problems themselves, or in the methods, and just do other examples. The real physics insights often come from people who are thinking about physics as well as math.

Is there hope that AI might find interesting insights, and how special are humans at discovering novel things about the world? It depends on the kind of insight, she says, and we are going to find that out. It is hard to think about something that does not quite exist yet. She reaches for the internet as a lower-dimensional analogy. The internet took on a life of its own, and it is also something we have been able to tame in places. She would not have been able to write her books if it did not exist, because she would not have had the time to go to the library and look everything up. It helped her enormously. In a nice world, AI could be that: a tool that helps us go a step further, more efficiently. It has already done that to some extent. Or it could be like the parts of the internet we cannot control, ruining politics. There are plenty of indications it can do that.

In terms of actually figuring things out, we are in the early stages. Fridman splits the landscape into precise theorem-prover tools and large language models that come up with convincing-looking things we do not know are true — a big problem for physics. She agrees. Large language models are more or less generalizations of stuff we have. There are still breakthroughs in AI waiting to happen, and maybe they are happening. Maybe in some cases it is just pattern recognition that leads to important things. Sometimes it could be something more insightful than that, which she cannot even put her finger on. We do not really understand how smart we are, or how we think. One thing is true: we are a lot more efficient right now than computers at coming up with things, and we require a lot less energy to do that. If computers figure out how to do that, it is going to be a totally different ball game.

There are also less abstract questions: how AI will be used in a world based on capitalism and a certain political system, how global politics will deal with it, what we will focus on doing with it, and how much control researchers will have to ask different sorts of questions. While it was starting out, people were doing toy problems. What will it actually be applied to, and what will it be optimized to do? Those questions, she says, it is really important we start addressing. Right now she does not think we have discovered new laws of physics this way. Could we in the future? Maybe.

The most beautiful unsolved problem depends on whether you mean the most beautiful, or the most beautiful she thinks we can make progress on. Most of the big questions have to do with what underlies things, how things started, how far science will take us, what fraction we are actually seeing, whether there are other forces or other ways of seeing the world, whether there are universes beyond our own — and if they are totally different, how we would even comprehend or detect them. She would like us to understand more about dark matter, dark energy, extra dimensions, the things she actually works on. One result from *Warped Passages* still excites her: there can be a higher dimension, but only locally do you think there is the gravity of a lower dimension. We might only locally think we live in three dimensions. For the observable universe, physics looks the same. Beyond it, who knows.

What advice would she give young people, in high school or college, on a career and a life they can be proud of? The weird thing about being a scientist, or an academic in general, is that you have to believe really strongly in what you do while questioning it all the time. That is a hard balance. Sometimes it helps to collaborate. To really believe that you could have good ideas, at the same time knowing they could all be wrong, is a tough tightrope. You still have to walk it, because otherwise you cannot really test them. The other thing: sometimes you look out and think there is so much out there. Sometimes it is good to bring it back home and just think, can I have as good an idea as the person next to me, rather than the greatest physicist who ever lived.

There are lots of big issues out there, and it is hard to balance them against the role of physics. Robert Wilson said it well when they were building Fermilab: this will not defend the country, but it will make it worth defending. In all this chaos it is still important that we make progress in these things. When major world events are happening it is easy to forget that. Those events are important too. You do not want to forget them, but you try to keep the balance, because we do not want to lose what it is that makes humans special. She does lose herself in the simple joy of puzzle solving. Everyone likes solving puzzles.

What drives her in research is the inconsistencies. When things do not make sense, it really bugs her, and she will go into different directions to see how they could fit together. The irritation is the motivation, because she has an underlying belief that it should make sense, even though the world comes at you in many ways and tells you nothing should. If you believe that it makes sense, then you look for underlying logic. That, she thinks, is just good advice for everything: try to find why it is the way it is. In *Knocking on Heaven's Door* she talks a lot about effective theory. Rather than only the big questions, sometimes you ask about the immediate things you can measure. You can sometimes tell in advance that a direction will fail. When we approach the big questions, it is good to do it from an effective-theory point of view, while still having the big questions in mind, one small step at a time.

We think things in the world we live in are beautiful. She is not sure we would, if we had different senses. She has to say, though, that no matter how many times she sees a sunset, she always finds it beautiful. There are things that clearly resonate with us as humans, and we were maybe evolved that way, but that is about us. In terms of figuring out the universe, it is amazing how far we have gotten. We have discovered many wonderful things, and there is a lot more out there. She hopes we have the opportunity to keep going.