How computer models help us understand the universe - with Andrew Pontzen

Thank you very much, thank you for coming tonight. I'm talking about the

universe

in a box tonight. It's really a book about the discovery of our Cosmos and it's a story, but it's probably not the story that most people know about how we found it. Discover this amazing

universe

that we live in because it puts

computer

s at the center and the way that

computer

s and their increasing power in recent decades have taught us totally new things about the Cosmos, about space, about what there's in it, um and about how we should really think about it about what's really happening around us and more than that, it's a story about cooperation, cooperation between many different people working on very different topics at different times, even in fields of study and way that all of these ideas from these different people have really come together on one canvas, a digital canvas that computers provide us with that allows us to

understand

more deeply what's happening in our universe and I want to give you a A taste of that tonight I want to show you some of the discoveries that we've made about the universe, uh, using computers, um, but I guess to start, we first need to take a look at the real Universe, the universe as it really is. out there seen through telescopes and I'll give you a hint that you already know, towards the end of the day I'll address the question that people have raised that maybe the real universe that we see out there itself is part of some higher simulation: a simulation carried out by immensely powerful beings that are beyond our direct reach, but I want to leave that for the moment and just look at the universe calmly, as we see it in reality, and you know if you look up at the night sky if you do it from here If you just go out now and look up you don't see much, do you?

You see some stars, you see some planets, if you're lucky, but what you really need to do if you want to

understand

the universe we live in is pause and allow your eyes to adjust so they can adapt to the dark conditions. In fact, even better if you can get out of London as far away from the city as possible. Light it up as much as possible and if you do that then your eyes will adjust and first you will see a few hundred and then maybe a few thousand if it is a very dark night.

Stars and you will see more than that, not only If you see the individual stars shining before you, you will also see a kind of band of faint light that divides the night sky and this is our Milky Way. In reality, it is not some kind of milky band of light as it seems to us, it is actually composed of hundreds of billions of individual stars, it is just that they are so numerous, so far away that they simply merge into this single band of light, and If you've never experienced that for yourself, I can completely recommend it, I mean, there's really something to see that and you don't need any equipment at all, you just need to get away from the city lights and take a look, but in terms of cosmology in terms of the things that I do in terms of the things. which we are actually trying to put inside the Box inside the computer, this is just the beginning of these hundreds of billions of Worlds and, by the way, many of them may have their own planetary systems, similar to our own system solar, one of the most amazing things.

What has happened in terms of telescopes looking at the real Universe in recent years is the discovery of more or less everywhere you look, virtually any star that anyone studies, they find what is called an exoplanet around it, for there are all these worlds out there and yet, This is really just the beginning, because beyond our Milky Way there are other galaxies, so there is our Milky Way with its hundreds of billions of stars, if you know exactly where to look, then you can see with the naked eye. The Andromeda galaxy, um, which is our kind of closest neighbor galaxy, is similar, it's similar in scale to the Milky Way, it also has hundreds of billions of stars, um, but it's very, very, very far away from Being Beyond our Milky Way, it is very far away in space, so it appears to the eye as a kind of spot of light in the sky, but if you have a good telescope, it starts to look like this and you can see from the outside how You see a galaxy: it is a kind of swirling mass. of gas, dust and stars and, of course, we cannot say it directly, but we have no special reason to doubt that all these stars can also have their own planets.

There is no reason why those star systems would be different from the previous ones. in our own Milky Way, so it's about 22 million million kilometers away and yet we are still on the cosmic threshold, this is still not the scale of the Universe in general because the universe is simply speciesless. of one or two of these galaxies is actually full of galaxies; In fact, if a rough estimate is made, there are probably hundreds of billions of galaxies in the visible Universe, each with its hundreds of billions of stars. This is an image taken by the James web Space Telescope which, if you've been watching the news in recent years, you'll have heard a lot about.

This image is actually of a very small portion of the sky. It's about roughly in terms of how. Large, it appears in the sky, it's about a tenth the diameter of the Moon, so it's similar in scale to a photograph that was taken by its predecessor, the Hubble Space Telescope, something called the Hubble Ultra Deep Field, was on a similar scale and revealed. a similar story that our universe is absolutely full of points of light and each of these points of light that you see on the screen right now are all galaxies beyond our own and all of that is packed into that little piece of sky. it's just that they're so distant that they're kind of, you know, they look like little dots in that area and there's nothing special about that part of the sky either, so if you wanted to know how many galaxies there are in the universe, somehow you have to multiply this by all the areas of the sky that you could have observed and that's where you get these kinds of estimates and they're just estimates of hundreds of billions of galaxies, um, and that's not all. because in a sense, the things that interest us most right now in astronomy and cosmology are not actually the things that are on the screen, because what you're looking at you know what this conference room is made of, what the land. is made of what the sun is made of, what the galaxy is made of and all those other galaxies that are visible to us seems to comprise about 5% of what is really out there and the remaining 95% of our universe is made up of materials that we don't really even begin to understand we don't know what they are they are called dark matter and dark energy um those are two different things they behave in quite different ways and they play very different roles um but uh but we really have very little idea of ​​what they really are, how they tie into the things we're made of and familiar with, and, in a sense, a lot of what we're trying to do in astronomy and cosmology. right now it's understanding it, you know what this 95% additional material is, we don't see it directly, it's called Dark Matter and dark energy, but in a sense it's not even really dark, it doesn't create shadows or anything like that. it's rather transparent it just doesn't seem to interact with light in any particular way and as a result you know you can't see it directly we only have indirect evidence that 95% of the universe is out there, but the indirect evidence is actually extremely strong , is extremely strong partly because some of it is really compelling and partly because there are so many coming from different angles, there are so many different lines of evidence that point in this direction.

So one of those is an effect called gravitational lensing. This is another image from the James Web Space Telescope now looking at a different part of the sky and what you might notice if you look at this is that many of the galaxies you see appear to be somewhat distorted into a sort of ring-like feature in the center of the image, what is happening there is not actually that those individual galaxies have adopted that strange circular shape, actually what is happening there is that the light from those galaxies is traveling long distances through the universe and, as it does , it is deflected by gravity, so our universe has a lot of things in it, things generate gravity and gravity can do all kinds of things, it can drag material, but it can also attract light, so as As light traverses the universe, if it hits a particularly dense area of ​​our universe where gravity is particularly strong, its path through the universe bends and that leaves you with these very particular distortions. in the images of galaxies that we actually see when we capture them with a telescope like the James Web Space Telescope.

Now, if you ask how much gravity is required to allow that level of distortion, then you can do a calculation and you can find. To know what the force of gravity is that allows this type of distortion, you can then go one step further and say how much material you need to generate that gravity because gravity is something that is generated by mass, so the fact that As you know, gravity is pulling us towards the surface of the Earth simply reflecting that the Earth is very massive, so for something like this to happen, you have to have a lot of mass and it also has to be in the right places, so that gravitational attraction allows us to infer what is out there even if what we are looking for is transparent due to its effect of distorting other galaxies and this is one of the lines of evidence that leads us to understand that there is much more out there in The universe that we see directly cannot explain this level of distortion if you simply assume that gravity is being generated by stars and gas and things that you can see directly, so the universe is full of these kinds of additional materials. additional ones we see.

We're actually still just beginning to understand it, but actually, going back to the simulations, one of the strongest evidence we have that this is correct and I told you what makes it compelling is the fact that there are so many lines of evidence. that tell us that the universe has extra things, one of the most compelling is when we look back in time and try to make sense of the history of our universe, this is something that simulations have proven to be incredibly valuable for, as well that we are lucky in a sense that light, although traveling very fast, still travels at a finite speed, so even in the real Universe, if you look at great distances, in a sense you are looking back in time when you receive sunlight.

It has taken several minutes to reach us or if you wait for the light from Andromeda, it could take millions of years to reach us, so if you look far enough into our universe, you are looking so far back in time that you can see the universe as was when it was actually very young, it was about 400,000 years old, which doesn't sound particularly young, but compared to the age of the universe, which is 13.8 billion years, 400,000 years was really the baby universe and again, this is a real image. although in false colors, so this is a false color representation of a kind of real image of the baby Universe taken by something called a flat iron satellite, specially designed to tune us into that really very ancient light that tells us about what was happening right at the beginning of the universe.

Now what you'll see there is that it looks a little unimpressive, just ripples, it doesn't look like much, but that's the point that the Universe back then was a very, very different place, it wasn't a Place full of stars, planets, galaxies and things that we can identify and look at as individual objects. It was actually more like a kind of desert with some little ripples, so they're little ripples that go through the sand more or less. just a flat, very boring, featureless desert with little waves running through it and it's those waves that have been selected here in the false color, so the question then is how is it possible that a universe can start like that? , without distinctive features? objects of some particular interest and they end up as we see today 13.8 billion years later and the answer once again is gravity, gravity is a kind of universal force that acts throughout our universe and means that even if one Part of the universe starts with just a little bit more stuff than its neighbors, so it generates just a little bit more gravity than its neighbors, which means that gravity is now trying to suck material into that small area that just started out slightly denser. than its surroundings, so gravity has the property of being able to pull things together and if you wait 13.8 billion years, that can be quite dramatic, so actually the vague idea is that you start from something like thatas on the left and where your Ripple just accumulated. a little more material together, just a few more grains of sand, over time gravity attracts all the surrounding sand grains and starts to form interesting things like galaxies.

There's only one problem and that is that this process couldn't take you from the left side of this. To the right side of this in 13.8 billion years wouldn't be fast enough unless you have something that provides some sort of extra pull and again this is where Dark Matter comes into the picture, it's the idea that we have some extra things that can create additional weight and speed up this process of building interesting structures. So how did we learn all this? Part of it is because of very expensive telescopes that

help

us observe the real universe, but in terms of understanding what it is. feasible, can you go from the left image of the universe to the right image of the universe in 13.8 billion years?

The actual equipment you need to figure it out is more like this. This is something very close to my heart. I actually spent the first few years of my life playing these little rubber keys messing around with learning to program and stuff like that and um, yeah, I probably just revealed myself to be the world's biggest nerd, but that's literally what I was doing for most of my adolescence. years um and uh, it just seemed amazing, you know you can create your own worlds inside these machines because they accept instructions, you can, you can tell them what you want them to do and they do it, which is kind of mind-blowing when, when "When you were young is still mind-blowing now and I think you know I was under the impression that this was a totally new idea at the time and it was kind of a new frontier, of course, it's not you." You can go back in time and the idea of ​​calculating things about the universe is extremely old.

This is the antia mechanism that was discovered in a shipwreck off the coast of Greece, in the early 20th century, I think it was discovered. and uh, it's some sort of very intricate set of interlocking bronze gears that is believed to have been made around 50 BC. C. and shows that people were trying to make very sophisticated astronomical calculations using mechanical computers 2000 years ago. So the idea of ​​trying to calculate the universe is by no means new, but there is a big difference between what I was doing in the ZX spectrum and what is possible with something like this.

You know, this is a fixed set of meshing. gears and it could do very sophisticated calculations related to planetary orbits and eclipse predictions and things like that, but what it couldn't do was go outside of what it was pre-programmed to do, so it couldn't take anything. I like this mechanism and calculating something completely new in it, like calculating something that has to do with I don't know the way two stars can orbit each other in the distant Universe because you would just have to completely redesign the gears from scratch. They are specially designed to do one thing, so what has made computers exceptionally powerful in our quest to understand the universe is the fact that they are actually very flexible today and the idea of ​​a kind of machine Flexible computing dates back to at least the 19th century, the work of Charles Babbage and A.D.

Love Fit when imagining a machine, in his case it would have been powered by steam, would have made its calculations by turning giant gears and large metal bars, etc. . but the whole idea of ​​it being customizable is that you can tell it what to do without you having to change any of the gears or anything to give it a different instruction, all you would have to do is give it a different instruction. um instructions encoded on pieces of paper you would actually use punching, punching holes in pieces of paper and um and providing instructions that way without having to change the machine so that you could change your mind about what you wanted to calculate without having to um without having to than redesigning the computer itself so that was a big step forward unfortunately the thing was never fully built this is just a part and you know accounts differ slightly on why it was never fully built but probably a factor What contributed most is that Charles Babage actually wasn't a particularly nice person, by most accounts he was a bit grumpy and fired all the engineers who were working on it because he decided they weren't good enough and then quit. to his professorship at Cambridge because he decided Cambridge wasn't good enough and then he fell out with the Prime Minister because he decided the government wasn't good enough and basically he was very, very difficult to work with for some reason um famous adah love uh still managed to work with him despite all of this uh and in fact she wrote a lot of what we now consider to be the most fundamental kind of ideas about what computers are good for and why they are very exciting and for us, that We're trying to understand the universe, one of the ideas he wrote about was that you could actually use this machine as a kind of prediction engine to predict what you might expect to happen. any given circumstance from the laws of physics and she wrote, you know that if you built this machine, then you could consider physics not simply as a vast body of abstract and immutable truths.

I mean, personally, I think it's a vast body of abstract, immutable truths. Immutable truths are pretty good, but it's not just that if you can already put them to some kind of practical use, you can start making calculations that are actually useful and, in fact, one of the first tangible ways they were put into practice. I use computers. Combined with physics, it was about predicting the weather, so predicting the weather is incredibly practical. You know, we take it for granted today, but we shouldn't because it's not just a convenience. It literally saves lives if you know a storm is coming and you can get out of its way, so uh.

Computing the weather actually saves lives. It's as practical as you can imagine and perhaps a little more surprising than predicting the weather. a lot in common with understanding the universe, so I want to draw that out a little bit, and as I do in the book, I want to draw that parallel, so to start, you get the idea that physics can be codified into what you call differential equations. Differential equations are a special type of equation that tells you how things change over time and that's what physics is all about. It's how things change over time.

If I throw something, after a certain time it will fall again. back to the ground, um, that's the kind of predictions that physics is interested in, what physics can't really do is tell you a starting point, so this is kind of obvious in the case of climate if you want to know if If it's going to rain tomorrow, for example, then you won't do very well just by looking at the laws of physics in the abstract - even if you have access to the world's largest supercomputers, the laws of physics won't

help

you. What you also need to know is the state of the weather today because the state of the weather today is what will determine the state of the weather tomorrow.

You know if there is a storm. If you're worried about a storm coming, you need to know. Where is he today? What are you doing? What is your trajectory? And more generally, you know, this is known as the initial conditions problem. Initial conditions are not something that can normally be calculated from physics. There is no law of physics that tells you what. what the weather is like today, unless we know what it was like yesterday, and actually in terms of moving towards a world where physics can be calculated and useful predictions can be made about both the weather and our universe. and about other things, one of the most important step was to address the problem of initial conditions and therefore, in terms of weather, it was not really possible to make good weather forecasts until one could gather an image, a snapshot how the weather was. today and that was made possible by the invention of the telegraph system and there was an exhibit, one of those stellar exhibits at the Smithsonian museum in Washington DC, it was a kind of map of the US where they could get remote information, um uh. observation stations meteorological stations would telegraph the current weather and then terator Joseph Henry would gather all the information and make a map much like a weather map we would be familiar with today just to show what is happening. is happening right now in the climate in the US and then from there it's kind of a small step to go well, now we can start making predictions about what's going to happen next, but you have to take the first step . know what the initial conditions are like and to reveal what is really important to us in cosmology, we have a problem with initial conditions, you know what we mean when we talk about initial conditions for our universe, it is one thing to tell you know for a forecast of the time we know what the weather was like yesterday, that's fine, but if we really want to talk about the initial condition of the universe, then we have to somehow know what was happening in its first moments, you know, in the in in which sometimes We call it the big bang and that is extremely difficult, almost by definition, physics doesn't really give us an exact answer to what was happening then, so we have to rely on a combination of different ideas based in part on observation of the real universe. , but I'll come back to that in a moment.

I just want to put aside the idea that initial conditions are really important if you want to use a computer to make any kind of predictions for the future, so the fact that you can take those initial conditions and combine them with physics or mathematical terms , differential equations and coming up with a prediction, it's not obvious at all, you know, you could imagine writing down a bunch of equations that are supposed to tell you how things change over time. what you think the initial state of the universe was like or the weather yesterday, plug all this together and then you get meaningless garbage, you could imagine that would happen if the equations didn't behave well enough, but fortunately we have something called uh, the theorem of Koshi Kovalev who tells you that if you have good enough equations and good enough initial conditions, then it is possible to make some kind of predictions.

This is a theorem in mathematics that tells you that the kind of thing that ad LEL was talking about making practical predictions about how things change over time is within reach, at least in principle, so all of these ideas were in the background, but at that time, of course, there were no computers available, so it wasn't really possible. To make any kind of weather forecast that was based on really good physics, you could look at weather maps and guess, but actually doing the math seemed out of reach, except that there was a very, very unique figure named Leis Fry Richardson.

He's kind of an absolutely extraordinary guy, the more you read about him the weirder he seems and he just wondered why you don't know if it's possible, in principle, to be able to predict tomorrow's weather from today's observations using the laws. of physics then why not, why don't we go ahead and try to do it now, as you can imagine, doing the calculations that allow this is a nightmare, I mean, it really is, I mean, this is an excerpt from his book, a sort of a table of numbers, but there were dozens of these, you would start with a tabulated set of information about today's weather and then you would sort of follow through a bunch of calculations, it almost seems like some kind of doing a really complicated task. . spreadsheet but of course doing it all by hand and to top it all off I was actually doing this on the front lines of World War I um and uh it was a pretty extraordinary thing to get into now there's no way I can even the most talented mathematician could calculate fast enough to predict the weather in advance, in fact it took him weeks and weeks of calculation to predict the weather, onwards only a few hours, so in a sense he was not trying to do anything practical.

He was more trying to prove the point that it works in principle and when it came time to figure it out well, then okay, if it works in principle, how do you actually make it work in practice? He didn't seem familiar. with the ideal ideas about computers, so he came up with the idea of ​​a giant amphitheater like this, where you have lots and lots and lots of people doing lots and lots of calculations and when someone finished a calculation, they had precise instructions on who to pass it to and then the next person would take the result of that calculation and feed it into the next calculation, and so on, and so on, he also made a kind of elevated extraction pit for himself.

It's in his imagination, by the way, no, of course, none of this was built, but he made a raised pulpit where you know someone would be almost like a conductordirecting, telling people where to pass their messages, etc., and so on, this is a picture, uh. that was drawn in the late '80s, trying to summarize what this would have looked like and, actually, you can zoom in, you can see the individual people doing their individual sums and, uh, it's a beautiful vision and it was actually something like that. It's hard to tell how serious he was, he wrote extensively about this in his book and he also described that outside the amphitheater there would be nice fields and things because people would get tired and need to go out and take a little time. running and breathing fresh air and um, he was a strange guy, but clearly you know he was onto something, he had the right idea, he just had the wrong way of solving it on a technical level, um.

And who could blame him? I was doing this at the beginning of the 20th century, don't forget that of course Charles bab gen love lace's work had never produced any machine that could do calculations so it just wasn't possible, but actually computers in this moment they were not that far away. So the so-called eniac machine, which was largely built during World War II, largely to help the nuclear weapons development program, one of its first apparently non-military uses was to start making weather forecasts. taking all these ideas but now putting them into a digital computer, so this now follows a lot of the ideas we've been talking about, including the ideas of Charles Babage's analytical engine and so on, but now that's it. done electronically and suddenly it is possible to do these calculations that we are talking about and Jul Chy, who was a meteorologist accompanied by a great team, did this work and they actually started producing forecasts that could practically keep up with the weather. so it wasn't quite practical yet, but it was getting there and of course since then the rate at which computer technology has accelerated and developed means that by now you know the kind of forecasts they were making, that you could do this. on your smartphone in your pocket in milliseconds those things that would take them a day or more, so the weather forecasters were essentially ahead of the game now, it's actually not a complete coincidence that this military machine was used. for weather forecasting purposes, um, that there was an idea at the time that the next kind of um, the next step in the war was going to be over the weather and that, beyond nuclear bombs, was the next kind of control at the state level that could be exercised. would be manipulating the climate of a country and um Von Neyman, who was instrumental in all of this, had the idea of ​​putting people to work predicting the climate on the basis that they would then see if the military carried out a certain intervention .

If we explode a bomb here, what effect will that have on tomorrow's weather? Asking these kinds of "what if" scenarios and if we tried this, "what effect will it have on tomorrow's weather" is actually a pretty scary thought, but luckily it turned out that the results showed that it is actually extremely difficult to manipulate the weather. , it is extremely difficult to do it intentionally, in any case, any type of intervention you try to do on the climate tends to have very unpredictable effects, this is an effect known as chaos that, although some things are predictable, but the interventions are generally not very predictable as to what they're going to do about the climate, so that's good, but unfortunately, of course, what we've discovered is that these same calculations that gave rise to the weather forecasts also started to predict climate change and the mess in the which we find ourselves in now and I'll come back to that in a second, so right now understanding the universe was moving in parallel with these developments in weather forecasting and there are many other figures that could I mention that I just picked one, Eric Holberg, who maybe a little bit like Richardson wanted to do this kind of idea of ​​simulations to predict what will happen in different scenarios, but he didn't have access to a computer and what he did was he had a lab where he set up a lot of individual light bulbs and they were supposed to light bulbs replaced stars in a galaxy and he actually set up two separate galaxies, so two, if you could imagine, two constellations of light and to figure out how those two galaxies would affect each other, he used a light intensity meter. to measure how much light is coming from different directions in different parts of his apparatus and he realized that light could be used in a It makes sense to replace Gravity, where you have a very intense kind of light coming from here that corresponds in a very mathematically clear.

In fact, the intense light coming from here corresponds to a strong gravitational pull if they were real galaxies, so he was able to do some very complicated calculations, but using the light to do the hard work for him and figure out how much gravity should come from different parts of these galaxies and then moved his light bulbs by hand, he discovered how gravity should attract and move them and in that way, over time, his apparatus evolved, it went from being two separate galaxies, the galaxies came together and merged into the half and they launched. they put out what we now call spiral arms, um, bulbs that shoot outward as they pass through a galaxy and fly out the other side, and that's what you can see down here, so it goes from their configuration superior. from two separate galaxies to the one below where they are merging and this is absolutely a phenomenon that occurs in the real Universe.

This is a pair of galaxies known colloquially as the mouse galaxies really merge and it really throws up these tidal tales and although again you know he didn't have access to a digital computer, but he was doing the same thing that we're essentially doing with the computers of Today, he was using the kind of power of a machine to calculate what should happen next, taking steps over time seeing how things change, but of course he had to set some initial conditions, he had to say by hand: I'm going to imagine that I have these two galaxies, um, I don't.

I don't know if this actually happens in the real universe, but I'll imagine these two galaxies approaching each other and see what happens next. It's very, very similar to what we do today and I wanted to mention one more type of foundation. work on how we think about computer simulations of our universe, which goes back to the late 1960s and was actually the first example where I think a computer simulation completely changed the way we thought about our universe, so this is Beatric Tinsley, she. She was working at the time on understanding the brightness of galaxies, the brightness and colors of galaxies seem quite esoteric, why do some galaxies have slightly different brightnesses and colors?

And she decided, because she had access to a computer, that she would try to simulate this and the particular simulations she had had to make a lot of very, very difficult assumptions to simulate all the details of what happens inside a galaxy; in fact, we still can't do that today, so she had to assume some initial conditions again. Well, let's say the Galaxy starts out with a lot of gas and then it also had to assume some things about physics that at the time weren't really known very well, things like how fast if you have a box of gas, a big box of gas, how fast do stars form because we know that over time gas can collapse under gravity it becomes very dense and when it becomes very dense it can ignite nuclear fusion and it generates new stars how quickly does that happen?

Well, Beatric Tinsley didn't know it and we still don't know it today, but you can guess and the really powerful thing about computer simulations, which is still true today, is that you can ask "what if" questions and that It's what she was doing, she was saying, let's try different things. Imagine we have so much gas and it turns into stars at this rate and generates so many heavy elements. So what happens next and what this kind of experiment inside the computer convinced her was that the assumptions people were making about life in all galaxies were wrong and it really mattered because people were trying to map galaxies. throughout the Universe, they were trying to use that to figure out how the universe is expanding because we know that our universe is expanding over time, that's definitely true, but at the end of the day there was a consensus that that expansion was going to end. in a few billion years and then the universe would collapse back on itself in what was sometimes called the Big Crunch and the Universe would end after about seven or eight billion years. had an expiration date, what the Beist Tinsley simulations showed is that very strong assumptions are made about what you're seeing in the real Universe; in particular, a galaxy was assumed to be very similar to any other galaxy, regardless of where a galaxy was viewed. whether it was close or far from you, at least on average you could regard them as interchangeable and what Beatric Tinsley's work showed is that that was simply unsustainable because you couldn't understand the brightnesses and the colors. of galaxies and at the same time assuming that intrinsically they were essentially very similar objects, it just didn't work, it just didn't work, whatever she tried in her simulation, so even though there wasn't any kind of um, this wasn't kind of of recreation of the galaxies. universe in a literal sense, was just a bunch of numbers on a computer, but what the computer allowed him to do was experiment and show that the assumptions that people previously made were simply wrong and that led to people realizing which actually the universe doesn't look like it's going to end certainly not soon um in fact now decades later we don't believe the universe has an expiration date, it looks like it's not only expanding but it's expanding at an accelerated rate, so which really doesn't look like it's going to end.

I certainly won't end up in that kind of big crisis scenario, so I have one last thing to say about weather and climate, which is the link between what we know and what we don't know, It is something incredibly powerful and important within simulations. This is Sikur Manab, he won the Nobel Prize for his early work in simulating climate change and essentially relying on weather forecast type

models

. He has a lot of modules. Now these are becoming much more sophisticated simulations. They have modules. inside them they do different things, some of those modules do physics that we know and love, you know physics that you can learn at school or university, things like thermodynamics and the way radiation moves through the atmosphere etc. ., but like Beatric Tinsley, there were other things that his machines were not powerful enough to capture, but that he knew were important, so Beatric Tinsley could not simulate the formation of a star and, similarly, Sakura Manab could not simulate cloud formation in our atmosphere.

The clouds are actually very, very small scale. characteristics compared to the entire Earth, so if you have limited computer power, you really have to simplify things and it's not possible to start with the laws of physics and learn where clouds will form in the atmosphere. What you have to do is make some reasonable guesses about how clouds form when they form, what they do once they form, and they're really important because they reflect a lot of light back into space, so if you neglect clouds in a climate. model or in a climate model you are going to understand it very badly, so there are many other examples of this in weather forecasting, there are things that cannot be captured with physics alone because the level of detail that would be obtained is too detailed.

What is required even today is too fine to the point of knowing that individual ice crystals form in the atmosphere. If you really wanted to use physics, you'd have to capture all this stuff happening and no computer is remotely powerful enough. To do that, you have to make some guesses and see if your results are sensitive to what you're inputting into those assumptions, so let's go back to cosmology, so here are the ingredients of a weather forecast. The initial conditions that I have talked about, which are the current observations, there are the physical things like the radiation of light, how the gas behaves and then there is what is called the sublattice, which I was talking about the things that you will never capture. because you just don't have the computer power to do it, as it turns out that universe simulations work the same way we have some extra bits of physics that we have to include.

Gravitation is perhaps one of the most important. that gravity attracts different objects together, our subnetworks, which are quite different things, like stars, as Beatric Tinsley was also studying today inday, black holes, we know that black holes have a very important role to play in sculpting the dust of our universe, believe it. or not, it really matters, it's just little grains of uh of um molecules that have stuck together in space, cosmic rays, the list goes on and on, there are a lot of things that are really difficult to capture, but the big question is, does What do they do?

What do we do with our initial conditions? We don't have that ability that weather forecasters have. They now have a very sophisticated network of satellites and Earth weather stations. They can find out exactly what is happening today and use it to predict. tomorrow, what can we do in cosmology? That goes in that direction. Well, it all goes back to what I showed you earlier about the cosmic microwave background, so we don't know exactly what was going on in the early parts of the opening O, part of our universe. But we have the ability to look back in time and see some of it and we have been able to relate those observations to ideas from quantum theory that talk about what the universe might have been like in its initial fraction. of a second and it essentially would have been a disaster.

This boils down to something known as the uncertainty principle in quantum mechanics which says that the Universe when it was born could not have been completely the same from place to place. The uncertainty principle forbids this, you can't make things exactly the same. Everywhere, there must be some level of uncertainty as to how much material there is in different parts of the universe. Now I'm oversimplifying because it's not something I have time to do. until today, but just to say that you can use this idea to make calculations, you can check those calculations with what we have seen of the early Universe through the Plank satellite and then when you put all this together, we think we know approximately what the initial value is .

The conditions for our computerized Universe should be like now. There are a lot of extremes there and you can go into the details of that, but we have an idea of ​​what the early Universe was like, so now we can bring all of this in. together we can put together the ideas about physics encoded as differential equations and the fact that with the initial conditions we just created we should be able to make predictions, we can add those subgrid rules that just say we don't have the computer power to do it all OK, so we have to make some reasonable guesses about things like how stars form and how black holes behave.

We can put it all together and we can produce a computer simulation now that we don't do this anymore. At zedx spectrums, as you can imagine, we're doing this on some of the largest supercomputers in the world. This is actually a movie I made 10 years ago and it held up quite well despite the advancement of technology and shows a simulation of that era. starting in the early Universe, you see that the universe is expanding, we have encoded the expansion of the universe in the computer code and what you also saw is that in the early T, in the first part of this video, everything looked pretty uniform.

It just looked like a green blob expanding towards you, but as time goes on, in addition to general expansion, you start that process by which gravity pulls things together, so that the slightly denser part of the early Universe begins to attract all the material. and makes this pattern very distinctive of the UN. I am rotating around the universe so you can see the 3D structure. It's not actually spinning. This is just for visualization, but you can see that it has this very distinctive 3D network-like structure. something that is very, very specific to dark matter, what you are seeing here right now is not something that you can see directly through a telescope, it is dark matter and of course we have the freedom to tell our computer what we want It is assuming that dark matter really exists and behaves in this or that way and there is so much that we can do those types of scenarios that I was talking about before, what you can also do, of course, is change. views and see what the universe would look like through a real telescope and what you now see is the same structure of the web bike because this whole structure has been sculpted by Dark Matter dragging things around, but now you can see individual points of light . and that's where individual galaxies are starting to form.

We're only about two billion years into the history of this universe, so it's still pretty early. Obviously one of the nice things about doing a simulation is that you can compress time into as short a period as you like and see how things play out if I take a trip through this universe and fly through it, you can see Since there are already these collections of many millions and billions of stars and they are all held in place by the gravitational power of dark matter, we cannot structure a universe in this way unless we have dark matter.

This was one of the breakthroughs in the 1980s, in fact what really confirmed the case that dark matter had to exist was the fact that there is simply no other way to get a Universe structured like this and make no mistake. when we look at the real Universe, it really is structured that way. The galaxies I showed you before are not simply scattered randomly, but are scattered along very distinctive filaments of a webik structure, so simulations in that way contributed to the growing evidence for dark matter in the 1980s. and then in the 1990s, so I've reset time and now what you'll find is that the simulation shows that the dark matter is still there.

I don't. It's no longer visible to you, but it's still there making its presence felt and it's bringing these galaxies together and helping them merge, so this is pretty much the same thing. in the same way as in that laboratory experiment, you know, throwing two galaxies at each other, watching them merge and throwing out these kinds of spiral-shaped features, um, but now it's being done not starting from some kind of assumption that two galaxies are going to collide, but from something well motivated. ideas about what was happening in the early Universe combined with those guesses about what the ingredients of the universe are and if we fast forward to today, almost 14 billion years later, then all these little elements come together and form many galaxies different and here we are simply flying into an individual galaxy.

There is no way, by the way, that a human spacecraft could do this. We just flew. Think thousands of light years. You just flew. So, no, sorry, that's, yeah, that's right, thousands of light years, uh, in the last few seconds, um, so there's no way any human telescope could fly through our galaxy like that, but we have telescopes like the Gaia telescope that is trying. make a 3D map of our galaxy only from a relatively close to Earth perspective, I'm making a 3D view and the fact is that these simulations actually do a pretty good job of recreating that view, it's not perfect, we never expected it to be perfect . too many compromises along the way, we've had to make guesses, we've had to make compromises, particularly at that subnet level that I was telling you about, where we have to be a little creative about how we use the finite power of the computer to make It's a kind of almost infinite universe, but in a kind of first approximation we are doing quite well.

The name of the game now is refining that picture, figuring out what we're doing right and what we're doing wrong, and there are a lot of things. You see these kinds of messages pop up in the media from time to time, but it's a long and difficult process to figure out and I think you know what a lot of people are really trying to do is understand what the dark matter is that we can put into it. in the simulations by simply telling the computer that there are additional things that do not interact except through gravity, they are there, please tell us what happens next, but what we would really like to know is what it is and of course we can do that kind of things.

What if in type experiments we can try putting in slightly different flavors of dark matter and see if that affects the type of galaxies that form? Does it lead to a better concordance or perhaps a worse concordance with the real universe and that way? We hope to understand much more about what this type of invisible mass in our universe is. Another way to do this is to build a giant collider. But you know you can choose whether you want a giant collider or a Zedex spectrum. I mean, if you're on a budget, I can recommend the latter, so that's where we are, we're trying to figure out what's really going on, but before we finish I wanted to say something about the real Universe again.

Because of all this progress, all this ability to create digital worlds that look so convincing both in cosmology and simulations and in the field of computer games, etc., it seems that we are creating realities that look more and more like the reality. something real and of course that leads you to wonder, well, when we look at our universe, is it that real or is it some kind of digital imitation? Is it possible to imagine that the universe itself was created within someone else's universe? Computers and quite serious people have raised this as a possibility. Well, okay, some people not so serious, but Elon Musk's take on this is that games are becoming indistinguishable from reality and then you know it if you think about them all. all the possible computers that could be built and all the possible computer games that could be made, so chances are you're in one of those and you're not actually in the kind of base reality, real reality, at all, but um, if you're not impressed with Elon Musk, there are other more serious people who take this point of view.

Richard Dawkins has written that he concludes that we could very likely be part of a simulation run by future humans inside some future cosmologists um uh computer. code, so I mean it sounds crazy, but serious people take it seriously, so I started to worry a little about this and it's true that, as I was saying before, the remarkable thing about physics is that it is computable and you can make predictions. therefore, you can imitate reality inside a box inside a computer, so maybe it's not crazy, and in fact, philosopher Nick Bostrom, who is certainly a serious person, wrote an article arguing from a point of philosophical view that this is actually extremely likely. because of all the future simulations that are going to be done and again it's that kind of counting argument, there are many of those future Sim simulations and yet only one reality so maybe we are already inside the simulation so I want to tell you.

Why don't I think that's really true? So if you look at our best computer simulations today, they are made up of about 10 to 14 bits, so one with 14 zeros after the bits and one bit is a unit of digital storage. sort of an irreducible unit of calculation, if you want, and you can look at the universe and take it at face value and calculate, well, how many bits you would have to have to recreate the entire universe without any of the approximations that I make. I've been talking or the assumptions I've been talking about and the answer is 10^ the 24 bits and not only that, but they can't be regular bits, they have to be slices, which just means they have bits in them. a quantum computer instead of bits within the computers that we have today, which are classical class computers, so it's a giant leap and some would say don't be afraid to take giant leaps, but I would say it's so giant that when we start claiming to reason about the simulations that people could be running if they had that kind of power available.

I mean, they would need literally every last bit of material in the universe to build this computer, apart from anything else, so when we start to reason about what they're doing, I think we're actually on very, very shaky ground. and you know it could on some level, of course, it's just a curiosity and not something that really matters, people can make these arguments if they wanted to, but on the other hand I think there is a serious aspect to this and that is that we live in a time in which science and indeed our very notions of truth and reality are constantly questioned;

It's a really horrible trend in that sense. that basically we're all being enlightened all the time about what's really going on around us and I think when we as scientists start to intervene in that and start to tell them, oh look, you know maybe the reality that we live in If a fact is real, we are going down a very dangerous path, so I think extraordinary claims require extraordinary evidence, and other than that, I don't think we live in a simulation, but I do think we all know more about simulations. and, fortunately, if you want to know more there is a very good book, thank you very much