Lunch Hour Lectures | Universes multiverses and simulations

hello and welcome to this UCL lunch hour lecture on universes multiverses and simulations I'm Andrea SAA I'm a professor of chemistry here at UCL and I'm going to be chairing today's lecture so it's my real pleasure to introduce an old friend Andrew ponson whom I've known for many years and I first met Andrew actually at the cheltam sence Festival a long time uh when he was playing the piano and literally singing the Praises of the universe and I must say I was quite hooked um I've seen his talk

s many times before and um both he and my children when they were younger and as they've grown older have have really loved his his his talks now he's a professor of cosmology and he's written a recent book called The Universe in a box which is just appeared in Italian as L um and his research involves computer simulations to really understand um the deeper parts of the cosmos um he's written for the guardian New Scientist sky at night and so on he's appeared on all kinds of radio and television

programs and um so he's really quite a high-profile figure from a from a physics and cosmology point of view now before we begin we're going to have time let me just tell you that we're going to have time for some questions at the end and you can submit these at any point during the talk by using slido and um you're going to need your your web browser if you scroll down into the to the lecture notes Here on YouTube um you'll actually see the link and there's either a link or you can use the eve

nt code uh with the hashtag multiverses and anyway without further Ado I'm going to hand over to Andrew and um yeah let's hear about the multiverses thank you thank you for that very generous introduction Andrea um I am sharing those slides now so hopefully you can you can see what I'm seeing on the screen shout if you can't um and thank you to everybody who's who's joining um thank you for giving up your lunchtime I hope I can tell you something interesting today I'm going to talk uh to a large

extent about how we go about doing computer simulations uh of our universe um and uh indeed to some extent of of universes beyond our own which is uh the the multiverses in the title it's quite a lot to fit into 40 minutes um so but I will do my best to give you a taste a taste for it and I thought um you know a place to start is is just to to to talk about the power of computers you know I grew up in the computer age and I grew up when um having a a ZX Spectrum was just the height of cool I wa

s so excited we got this it was a little sort of plasticky box that plugged into a television and you could use it to to well of course you could play games but actually for me it was much more exciting to program it to to do uh what you wanted it to do and one of the first program listings that I typed in out of a magazine was all to do with seeing how things orbited how you could get stuff to orbit around a planet and um in some sense you know what what I'm doing my research today is a continu

ation of that we have vastly more powerful computers now so uh the the way we can bring the universe into sharp focus and and understand how all the different bits of the universe work together to produce a kind of vast vastly complicated un and and and all of the phenomena that that go inside it um I just find it incredibly exciting it's one of the best jobs you can to as far as I'm concerned but a few years ago I kind of got got interested in where did where did the idea of simulating the univ

erse inside a computer actually come from and that's what my recent book was partly all about so alongside talking about some of the physics and so on I'm going to try and give you a a sense of that today as well so where am I going to start well I'm going to start with the universe as it actually is before we try to put it inside a computer what do we know about the Universe um in in reality well if you go outside on a dark night here in London you don't get to see very much of it you you see m

aybe a few planets some of the very brightest stars but not more than a handful of stars normally but if you get away from London if you go and uh see the night sky from a much darker area and you allow your eyes to adjust the richness of what's actually out there starts to become clearer you know your your eyes adjust to see hundreds and then thousands of stars and then after a while you also see a sort of faint band of light splitting the the sky into and that's what we call the Milky Way it's

um it's not just it's not sort of a a more strip of light it's actually hundreds of billions of stars far beyond our own um and they're just so numerous that they kind of merge together into this band of light that that we see across the sky and I think to to really appreciate just you know what that means it's it's worth also mentioning that in in recent years we've been able to use telescopes to stare at some of these Stars and despite the fact they're so distant we can actually measure uh th

e fact that they have planets orbiting around them so they're not just sort of stars like giant nuclear furnaces all on their own they're actually sort of ecosystems in their own right they they have planets orbiting around them they're generating light and heat for those planets and as far as we can tell based on the statistics we have today pretty much every Star seems to be surrounded by some kind of planetary system so in our galaxy there are probably hundreds of billions of solar systems um

and just to preempt one of the inevitable questions that always comes up I mean personally I think the chances of finding life out there one day are probably pretty strong just because there are so many of these systems out there it would seem slightly unlikely that Earth would be uh so unique amongst those many many systems but in terms of what we do and trying to understand the universe as a whole and put that into computers we're still here just seeing the beginning you know our own Galaxy i

s in fact just the start of of what's out there and uh at the start of the 20th century it began to be realized that there are many many other galaxies beyond our own and that those galaxies too contain hundreds of billions of stars so um this is our nearest neighbor large Galaxy it's it's called Andromeda and if you want to see this for yourself you certainly can um you need to know where to look you need to be fully dark adjusted and ideally have a have a good pair of binoculars but if you kno

w what you're doing you can find this on the night sky and um it's uh it just looks like a sort of smear of light but it too is this collection of hundreds of billions of stars very much like our own Galaxy and as I say that's just the beginning you know that the universe is absolutely packed with these things in fact what I'm showing you now is an image from the James Webb Space Telescope that launched a few years ago and uh this image was released just last year and shows a tiny patch of the s

ky it's round about the full width of this is around about a tenth of the diameter of the Moon so very tiny bit of of the night sky and the James web Space Telescope was was was pointed at that one patch of sky and left to take a picture and this is what came back now everything that you see pretty much in this image even the tiny tiny dots they are all galaxies far beyond our own and all of that is just crammed into that tiny little patch of sky so if you if you take this at face value and extr

apolate up to the total number of galaxies in the universe then uh you can put a lower limit on that at many hundreds of billions so many hundreds of billions of galaxies each containing their own hundreds of billions of stars um Each of which may host planets and that is the kind of scale of what we're looking at and the scale of what we're trying to recreate inside today's computers um but that still is the tip of the iceberg in some sense because everything that you see directly as far as we

can tell accounts for only five% of the actual material that the universe is made out of we have extremely good evidence along multiple lines of evidence that what you see is about 5% of of what there really is and there is another sort of 95% to the universe and that 95% is made up of materials that are extremely unfamiliar so unfamiliar that we've never really or never at all been able to experiment on them in a laboratory there're they two materials known as dark matter and dark energy and th

ose are going to have to be in our computer simulations if we want to have any chance of explaining what's going on in the universe because although we don't see them directly they have a profound influence on what's going on so the role of dark matter is that it essentially generates a great deal of gravity it's just it just carries a lot of mass around with it the mass generates large gravitational pulls and those gravitational pools uh manipulate and and brag around the material that we do se

e in in the real Universe the other material is called Dark Energy and that uh again is invisible just like dark matter we can't see it directly but it seems to have the effect of of pushing material apart uh and making the fact that the the universe is expanding as it's the first thing to say you know the universe is is getting larger but it seems to be getting larger at an ever increasing rate and dark energy is the material uh that is responsible for this and you know these two things they're

just made up names in some sense they they are standing in for the fact that there must be additional materials to explain what we see and I'm out to tell you what some of the evidence for that is but you know that don't take the names too seriously they don't really mean very much dark matter and dark energy they're just standing in for what we hope one day will be a proper understanding where we can put these materials alongside the materials that we know and love the atoms and molecules and

uh photons and protons and neutrons and all of the things that that we know and can experiment on in a laboratory we hope to be able to fit d matter and dark energy into that picture and really specify and pin down what they are in fact this is one of the main goals that we have when we're performing computer simulations we can essentially do experiments where we imagine that the dark matter and dark energy is really made out of something specific we put forward a theory for what uh these myster

ious materials are actually made out of then we try programming a computer to follow through what the consequences of that would be for what we do see in the real universe and to a large extent this is driving forwards this whole program of uh capturing the universe inside computers is is our quest to understand what these missing materials really are anyway I've I've gone off on on a bit of a tangent there I want I want to kind of dial it back we've seen in short you know what the universe is m

ade out of and how vast it is and the kind of things that that we're trying to capture let me just dial it back for a moment and talk about what the evidence for uh things like dark matter is and and you're looking actually at one of the strong pieces of evidence we have for the universe containing this additional stuff Dark Matter this is another image taken with the James web Space Telescope uh here you're seeing a bunch of galaxies and again in a dist part of the universe but if you look at t

his you'll immediately see that they're kind of bent into a particular shape there's a sort of sense of a sort of circular pattern in the sky in this particular place that is not because uh we believe the the galaxies themselves have been Bent by anything in fact we have very good evidence to to show that the galaxies themselves are perfectly normal galaxies they have perfectly normal shapes what's going on here is that the light coming from distant parts of the universe is being bent so it's it

's coming from a normal Galaxy but on its way to Earth it's being bent and it's being bent just by gravity so one of the things that Einstein's theory of relativity established is that as well as pulling material around gravity can actually pull light around and so as light makes its way through the universe towards our telescopes it can get deflected and we call this effect gravitational lensing what you can do with gravitational lensing is work backwards from the distortions in the images to a

n estimate of the strength of gravity that just generating those distortions and then in turn to um uh to to the amount of stuff the amount of material you would need to generate that much gravity and this is one of the key pieces of evidence that we have that there must be extra stuff in the universe that we're not seeing directly but that is contributing to generating this gravity this very strong gravity that's required to distort the images of galaxies in this way so that's one piece of evid

ence but actually that the evidence starts to get really convincing when you throw simulations in there because you know generally speaking you know when you when you present a piece of evidence like that people can counter it by saying well okay but that's all based on assuming that we understand how gravit works it's assuming that although it's a very well tested Theory we're we can't be 100% sure that Einstein's theory of relativity is correct so maybe we're just miscalculating the effects of

gravity it's a it's a slightly um it's possible to sort of undermine this evidence for dark matter in that way but there's another piece of evidence for dark matter which is based on simulations of physics that's much harder to undermine and that's based on contrasting the universe as it was when it was young with how the universe is today and trying to fill in the gap between those two pictures using computer simulations so in a sense we're very lucky that light although it travels very fast i

t still travels at a finite speed so that means if you look to large distances in the universe the light simply has got to take a long time to reach you it it like like like anything although it's going very fast you know the further it has to travel the longer it has to travel for to reach its destination and we can use this to look back through time we actually have a special telescopes this is an image from the the plank satellite uh that was an EA mission that flew a few years ago that reall

y tuned into light from the very earliest uh age of our universe a time when the universe was only about 400,000 years old which sounds like a fairly substantial age but in in terms of a fraction of its current age which is about 13.8 billion years uh this really was a very very young universe and we can see directly what was going on because the light uh you know if you tune into the right kind of light you find the light that's been traveling for all of that time so you can see a picture there

of uh what was seen by the uh the plank telescope and you can immediately tell it doesn't look much like the universe today I mean I should say it's false color uh the the the universe wasn't painted bright red and blue in fact all the light we're receiving is in the form of microwaves which is a very long wavelength type of light and then false color is used to highlight what are actually very very tiny differences between one part of the early universe and another part of the early universe s

o the red and blue patches here different differ only by about one part in 100,000 it's that kind of level they are very very similar actually but um you know one one that the red bits have slightly more material in them and the and the blue bits have slightly less material in them just at the level of about one part in a 100,000 other than those small rippley type variations the early Universe was a very very uniform place it wasn't at all like the the universe today where we have um you know g

alaxies and stars and planets and then vast spaces in between the universe today is a place of huge contrasts some parts of it you know if you go to the center of a star it's millions of degrees but if you go out into the furthest reaches of space it's only 3 degrees above absolute zero whereas in the early Universe one part of it was very much like any other and you have to kind of color it in this F color at the level of one part in 100,000 to see anything at all so there's a there's a fundame

ntal question there which is how did the universe get from its early stages uh a very very smooth uniform boring place to the remarkably Rich varied place that we find around us today and the answer to that question has in large part been given to us by computer simulations where we PR program computers with what we think are the most important physical laws notably gravity consistent with Einstein's theory of relativity um and we try to understand what happens if you start with the universe bei

ng very smooth as we believe it did what happens next and I will show you simulations of this in in a little while but just to give you a sense of it the most important thing that happens is if you have some region of the early universe which has just slightly more stuff in it than a neighboring region of the universe there's a net gravitational force towards the the region with more stuff it's just got more stuff therefore it generates a stronger gravitational pull and it starts to suck up and

sort of Hoover up the material around it um and uh and and that process is a kind of runaway process that uh what happens is that over time those very slightly more dense regions become more and more and more and more dense and gravity gets stronger and stronger and the process runs away and and it builds the galaxies that we see now the the reason this is such good evidence for dark matter is that the the process cannot happen without something like dark matter governing it because we find if w

e if we try to make this happen using just regular matter then it it it it doesn't work essentially the the pressure from the regular matter is too high and the pressure tries to force the material back out again there's insufficient gravity to actually pull everything together and so what computer simulations of this process have have been able to to do is predict the process predict how it works on the assumption that something like dark matter is there and in fact they they really done beauti

fully well at making those predictions overall with some notable exceptions but overall they've been extremely predictive and extremely successful in accounting for this kind of huge contrast we see in the the young universe and the old Universe like I say I'm I'm going to show you that in just a moment but let's let's now step back and you know where did this idea of trying to simulate things come from well like many things in computation you can trace it back to the work of Charles babage and

a love lace in the 19th century um Charles Babbage was inventing uh a machine that would be a sort of general purpose computer you'd be able to instruct it to carry out any computation that you liked without physically changing the machine you would just give it different coded instructions on on Punch Cards um and unfortunately the machine never got built or at least not in full and and there are all sorts of reasons for that um but partly that Charles babby was a absolutely horrible person to

work with by most accounts um he he was extremely rude to the people around him he sacked uh half his engineering team because he decided they weren't good enough then he resigned from his professorship in Cambridge because he decided Cambridge wasn't good enough uh and uh and and then actually I think he um he he got cut off from his Government funding because he became so critical of the government he decided that the government wasn't good enough either uh and and his grants were cut and the

thing never got built but its Legacy lives on largely thanks to the work of a love lace who's a rather more sympathetic character who had really thought about what this could be used for and one of the things that she highlighted was that it was a way to turn physics from being study of abstract and immutable truths uh you know sort of with with limited applications you know applications just really Limited by the handful of things that you could actually calculate by hand which is relatively sm

all um and and it sort of promotes it from that abstract study into being something much more practical because if you're not limited by the human ability to calculate if you have a machine that's able to do calculations of any type for you um at um you know more or less any speed if you combine enough of them then all of a sudden uh all of these laws of physics that had been acre become much more amable to to practical application the um one of the first envisaged applications for this was was

actually weather forecasting um and weather forecasting has played a huge role throughout the history of simulations and has has been deeply influential on the way that we think about um the universe but one of the things that it really highlights one of the reason I talk about it a lot in my book is that it's not enough actually just to have the physics and just to have a computer that can do all the calculations that the physics implies you also know you you also need to know what we call the

initial conditions and um in the case of weather that means having really good accurate measurements of what the Earth's atmosphere is up to today if you want to be able to predict what it's going to be doing tomorrow you know if you want to predict what a storm is going you know whether there's going to be a storm here in London tomorrow for example then it's very very useful to be able to track them coming across the Atlantic Ocean and in fact one of the first people who really followed this d

ream was Joseph Henry who is uh I think the inaugural director of the Smithsonian uh museum in Washington DC and and he had a a weather map one of his sort of prize exhibits was a weather map where every morning they would use the telegraph system to collate weather reports from across the US and uh put together a weather map of what the weather was like across the entire uh west across the entire us that day this was a huge novelty at the time people would come just to see this weather map to k

ind of Marvel at the idea you could know what the weather was like um such long distances away but Joseph Henry went a step further and started predicting weather you know he he could look at it and you know they were in Washington he could see if there was a storm in Cincinnati then he he would be able to have a good guess that it was going to track and and uh come over and they would soon have a storm in Washington as well um so this idea of weather forecasting and the way it relies on the ini

tial conditions was also something that sort of arose in the the 19th century and the first person who tried to put these ideas together tried to put the idea of just using Brute Force coupled to the you know the the laws of physics and the relevant laws of um things like the Navia Stokes equations these are equations that describe how things how fluids move around and how how masses of air move around the earth and so on um so you take those bits of those sort of abstract equations of physics y

ou combine them with some initial conditions which represent the state of the weather today and then you just do loads and loads and loads of calculations and by Brute Force you try to extrapolate what will the weather be like tomorrow and the first person who attempted to to do this was leis fry Richardson and um he uh was quite a singular figure I mean what you're seeing here is is just a tiny fraction of the calculations that he had to do he um had pages and pages and pages of these calculati

ons it looks like a giant spreadsheet of course long before there were any such things as spreadsheets and he just did loads of manual calculations to uh to try and um and figure out a weather forecast now he there was no way that he could do this fast enough for it to be a practical proposition he was working with weather uh a weather forecast that would only kind of be predicting weeks and weeks weeks and weeks and weeks of calculation would be taken to make a prediction for just a few hours o

f how the weather would unfold and Incredibly he actually did this while he was serving um on the front line of World War I he uh he he was um actually wounding people backwards back from the front so so he was um fing wounded soldiers back from the front um working for the friend's ambulance service but in in his sort of downtime he was working away on these calculations and just wanted to prove a point that this idea of calculating the the weather in this very sort of Brute Force approach coul

d in principle work um he wasn't really aware of the fact that computational technology was around the corner In fact when he got round to writing up this idea his vision was for a giant purpose-built Amphitheater where a bunch of mathematicians would sit each in their own room um if you there's a sort of zoom into this picture you can you can see an individual calcul calculator sitting at a desk um so he had this sort of vision of a giant Hy theater lots of people working um and passing each ot

her their results all sort of coordinated from a central pull pit where he would tell people what what they ought to be doing and somehow by working together the idea was you could achieve the speed of calculation needed to make predictions for the weather ahead of time rather than what he had been trying to do which which had just been you know to spend weeks calculating the equivalent of of a few hours um and he proposed this in in a book quite seriously as far as as far as I can tell he was p

retty serious about this Vision but um in fact it it it wouldn't be necessary because within his lifetime rather than anything like this coming along digital computers actually came along and um the the first digital computer to perform a weather forecast was the eniac the famous computer that you might have heard about in connection with the end of World War II and the race to develop the hbomb and so on uh that computer was actually used to produce a numerical weather forecast using a lot of t

hese same ideas and that wasn't at all a coincidence the military were deeply deeply interested in weather forecasting first of all because just being able to forecast the weather is of course in itself deeply practical but they actually had other ideas up their sleeve that what they were hoping to do was to instigate a program of weather control which was believed to be a sort of even even greater threat in some sense to um world stability was the idea that the the USSR might be able to start i

nfluencing the weather and if they could influence the weather then essentially uh they would have something even more powerful than the hbomb at their disposal so the US started a kind of race to try and figure out um how can we uh influence the weather um and as it turned out that line of research got shut down fairly quickly because it's extremely hard to influence the weather uh which is probably just as well um but that that was what was at the back of the minds of the people who were devel

oping these first numerical weather forecasts on on the eniac so we've taken a bit of a detour again about into into very brief history of some of the new maracle weather forecasting so to take it back to cosmology um of course there were people working parallel with this having ideas about how can we get the idea of sort of reproducing Cosmic phenomena inside computers or even before computers were available there were were various attempts like this one by Eric Holberg he he used a bunch of li

ght bulbs actually he had a laboratory at the the Lund Observatory and he darkened it he blacked out all the light used a bunch of light bulbs that he would was using to stand in for stars so they they were supposed to be like a kind of uh an analog for Galaxies little stars that he dotted through through his laboratory and then he used the intensity of the light coming from the different stars as a way to kind of estimate the strength of the gravitational pull that they would produce it turns o

ut there's a very exact correspondence in fact between the intensity of light you receive from something which drops off uh technically as 1/ r s and the strength of gravity that it generates which also drops off as 1 r s so he was able to use a bunch of light bulbs to kind of stand in for uh the stars in galaxies and try and figure out using that um rather than have to calculate by hand what gravity would do to galaxies he was able to do it sort of experimentally and figure out what happens if

two galaxies collide together and he came up with the answer that well essentially they they they fly towards each other and they throw out spiral arms into space and this is a phenomenon that's now well recognized it's something that happens in the real universe and that we see in our modern simulations a great deal here's an example from the real Universe sometimes known as the mice of two galaxies that have kind of smashed together and thrown out a stream of material out into deep space so Er

ic homberg was the first person in a sense to simulate that but he did it with this s of very analog approach rather than a digital computer so if you fast forward a couple of decades then you find the the first first digital simulation of galaxies the first thing that starts to touch on what we do today was done by beatric Tinsley um uh who ended up at Yale University and she actually changed basically everything that we that that was thought to be known about cosmology at the time by asking pr

etty straightforward question which is do galaxies shine with the same brightness over their entire life and she came to the conclusion by putting together some quite simple um uh bits of physics with some best guesses because it wasn't really possible it still isn't possible to capture all the bits of physics that matter she had to to write down some best guesses for example for the way that new stars can be formed out of clouds of gas things like that um and she showed that it was just untenab

le the idea that galaxies could somehow shine with a steady brightness over their entire lifetime was untenable and that really mattered because at the time people had the idea that galaxies essentially are unchanging things that might you know they might vary slightly over time but but but to a to a good approximation they were thought to be essentially shining bright shining with the constant brightness over time and a bunch of observations and the interpretation of observations had been based

on that idea including something as important as the fate of the entire universe was was kind of bound up in this because to map out the way that the universe is expanding today you need to understand what you're looking at you need to understand how bright intrinsically are all the galaxies that uh that you're looking at um and people have be just been making wrong assumptions about that and by running some of the first ever simulation Tinsley was able to show that that there's no way that the

se things can shine with exactly the same brightness through their whole life it's just it just wouldn't wouldn't work um and um and that basically uh changed the viewpoint on what the ultimate fate of the universe is because it changed how we thought it was expanding and therefore changed how we extrapolated that expansion into the future it changed the picture from being a future where everything would ultimately stop EXP standing and collapse back down and and the universe would end in a few

billion years that was the sort of the old picture it completely upended that and said no actually you if you interpret this more carefully it looks more like the universe is going to carry on expanding forever and that's still what we think is is happening today um we still believe the universe will probably carry on expanding forever so it was an early example of simulations really changing the way we thought about the universe but I think one of the things that that story teaches me when I Un

earthed it is just how important it is to realize that simulations don't have to be literal Recreations of the universe no Tinsley had to make loads of very rough guesses at the way that galaxies behave and it wasn't just it wasn't that she was trying to perfectly reproduce the way that galaxies behave it was more that she was trying to understand the range of possibilities know what flaws ible for how galaxies might change over time and it was by doing that that she was able to completely chang

e cosmology this is still the case today we still have a bunch of things inside our simulations where we really um we really don't quite know how to handle them there's an example in weather and in fact in this specific case climate change simulations so this is sikura manab who won uh the Nobel Prize a few years ago for his early work on simulating climate change and this is a a diagram from one of his papers explaining how the simulations work part of them yes is about physics that's well unde

rstood and that uh is sort of uncontroversial but another part of them is what we call sub Grid it's it's stuff that is happening that's sort of happening on scales that are just too small to capture inside a computer so that might be in the case of we for example you know small scale convection the generation of small clouds for example basically impossible to capture on a computer um even today it's very challenging so you have to make assumptions reasonable assumptions about how those clouds

might form and you know there's other things like the way that surface water drains away from the surface of of the earth once it's rained for example that's really important but is uh extremely hard to capture in terms of just laws of physics you have to make some reasonable assumptions and then test how sensitive your simulation is to varying those assumptions so this is this this is the kind of very important thing to realize about simulations they are not literal Recreations of our universe

where where we're just sort of revealing the truth about our universe they're much more like imperfect experiments where we're capturing some useful insights but we are not literally recreating anything so so just to recap here's what you need to perform a good simulation you need some initial conditions which in the case of weather are given to us by present day observations of what Earth's atmosphere is up to you need some Physics which in the case of weather forecasting it's about the way the

gas moves I mentioned the uh Navia Stokes equations things like that and the way that light and radiation behaves but you also need what we call a subgrid uh model which is all the stuff that we our um computer isn't powerful enough to get from physics um that that we have to make some best guess assumptions about and I don't think this will ever change particularly even though computer power is increasing very rapidly there will always be processes that are just too small relative to the scale

of the thing that we're simulating for us to be able to capture them uh using completely tried and tested uh well-known laws of physics for the case of um uh the universe this is much the same pattern we have some extra physics that we care about like gravitation and magnetic fields um but we have these kind of subgrid uncertainties I mentioned the way that stars form is one of them the way that black holes form and evolve is another one uh things like formation of dust in the universe cosmic r

ays the list could actually go go on and on and one of the really difficult things about simulation is making sure you're doing the best job that you possibly can of of that subgrid level but I wanted to mention also that we also have a an issue with initial conditions for cosmology because we have to be careful about what we assume about the early Universe we have some idea about what was going on in the early Universe thanks to uh things like the plank satellite which I showed you ear on but u

m uh we also need to supplement that with some account of what was going on in the early universe and I'm not going to have a huge amount of time to talk about it but just to say that's given to us by um uh quantum mechanical idea known as inflation and Loosely the idea is that in the very early Universe because the universe was so small there were there was sort of quantum uncertainty you can think of it is like ripples going through the universe because of Heisenberg's uncertainty principle th

e universe couldn't have been completely uniform it must have had some variations in density in there Rippling Rippling along and that it's those variations in density which we later see in the cosmic microwave background and which then get Amplified to become the galaxies and so on that we know today so if you put all of that together then here is a a simulation as they look today you just saw the big bang you saw the univers is expanding towards you you'll see it starts out fairly smooth but u

h pretty quickly you start to get structures forming within the universe you'll see that these structures are turning into like a kind of webike structure this is something known as The Cosmic web it's also being observed to exist it's a kind of overarching organization for our universe as a whole on really vast scales we about hundreds of millions of light years here um what I'm showing you on the screen right now is actually dark matter you wouldn't be able to see it if you actually lived insi

de this universe but the dark matter is the thing that is generating this overarching structure for for the universe as a whole if you switch views and look at the visible Universe it looks like this instead um and now you can see this just sort of traces of traces of gas and also these bright specs where galaxies are forming and if I fly through that you can start to see that the those individual specs resolve into billions and billions of stars and if I restart time because we we're still fair

ly early on in the history of our universe here then you can start to see these individual little galaxies build into larger and larger galaxies over time through a process that we call hierarchical merging um and this makes very very specific predictions it's being driven by the dark matter in a simulation and it makes very specific predictions for what we should find with telescopes as I say many of those not all of them but many of those have turned out to be accurate so this is this is the s

tate ofth art essentially we're trying then to test different ideas for what dark matter and dark energy might be and see how see what effect those have on the galaxies that form and alongside that of course we're having to Grapple with all of those subgrid questions that I posed earlier on but I want to say one final thing because I promised to talk about multiverses so I have to get the multiverses in there um I said about the you know the beginning of the universe is still quite difficult uh

period of the universe's history to understand and Quantum effects back then were really important and that means the the amount of insight that we can get from our simulations on classical computers is really quite limited um it's difficult to simulate Quantum processes on a classical computer so most of our simulations are aimed specifically at understanding the later Universe after the quantum effects have kind of Frozen in they've kind of frozen into just some waves which we can then treat i

n in a classical way but what another sort of aspect of the universe that we'd like to understand is its very early history and to do that we're following a program of uh what we call Quantum simulators this is this is funded by the quantum Technologies for fundamental physics Grant uh that ukri um are supporting and here instead of using a classical computer we use a Quantum analog system so this is actually something that you build in a lab I don't build it in a lab because I I you shouldn't l

et me anywhere near a laboratory it's a total catastrophe if you do um the our collaborators at Cambridge University are building this at the Cavendish laboratory this is a picture of of of what it looks like today and what the the experiment actually consists of is cooling things down to create what's called a Bose Einstein condensate where atoms start behaving in a kind of coherent way that's very similar to some of the physics that was going on in the early universe that I was talking talking

about a moment ago and in particular one of the things that we're really interested in understanding is the potential for the early Universe to not really just be a single universe that because of the weirdness of quantum mechanics one thing that can happen to that that roing mess in in the early universe is that it can form little bubbles within it so here is a here is an animation of that happening you can see a a red bubble forming with in a blue background this is a something known as a pha

se transition the early universe may have gone through these phase transitions um and the the red bubbles would essentially then be pocket universes universes of of their own isolated from each other so our entire universe then would fit inside the red bubble that you see on screen but it would have the implication that there could well be universes beyond our own so we we're trying to simulate this particular particular process in the lab what you actually saw just then was a uh simulation of t

he simulation so we this is getting to uh crazy levels but uh we we we actually try to capture the physics that's going into the quantum experiment in the lab we tried to capture some of that on uh standard computers to understand more about the experiment that we're doing so what you just saw is not the experimental results we're not we're not quite there yet but it's a sort of glimpse of what we're hoping the experimental results might look like shedding light on the way these phase transition

s happen and on the way they might give rise to not just one universe but actually multiple universes so I'm aware that I'm now uh very low on time so I'm going to uh just skip through to my final slide which is bit of an advert that if this says uh wet your appetite then I've written a lot about all of this stuff uh recently in in this book so uh please uh if you're if you're interested then do get a copy perhaps even review it on Amazon um but I think now would be a great time to stop and hand

back to Andrea who I think is going to chare some questions thank you very much Andrew that was that was absolutely fascinating and uh what what I found really particularly brilliant about it was linking together um you know scientific simulation in so many ways one of the things just in chemistry that has really transformed chemistry from being one of those things where you're dealing with glassware and so on the whole rise of computational chemistry and in the Life Sciences too all of this co

mputational biology um it's it's it's really brilliant to see the the kind of bread of stuff that uh that is that is going on so we have a number of questions and um for those of you who are online um what you might want to do is to actually vote up some of the questions just so that you know we see what are the things which uh which really Intrigue you the most but I'll take the the question that's been voted up twice and that is there are news articles that mention um that images of the oldest

galaxies captured by jwst John web John Webb I got J James web sorry wrong saint uh James Webb Space Telescope don't look as expected and how do the simulations deal with these things now perhaps if I could just link it to a second question that's there is how do you simulate a universe in which you don't understand what dark energy and dark matter actually are and how they work yeah absolutely so so let's take the first thing first so yes absolutely there's been a lot of very eye-catching pres

s releases out there talking about excuse me talking about um how the galaxies from oh dear the galaxies from James web are problematic um I take those with a very large grain of salt so what is true is that there are some galaxies that have been found by James Webb which is peering back a long way through time so it's it's looking back to some of the earliest galaxies to form long after the plank Cosmic microwave background that I showed you but also long before the present day um and it is abs

olutely true that uh those some of those are larger and brighter than we had anticipated however um the the problem is that the Universe was a very different place back then so it comes down to the sub grid that I was talking about so if you're interested for example in how fast to start form then we have a very good guide to that by looking at the nearby Universe when we look at the nearby Universe we can see roughly how quickly stars are forming and so that's a guide to these subgrid assumptio

ns that we have to make to uh get the simulations to work at all um but if you if you base those subgrade assumptions on what you know about the current the present day Universe then you're not taking into account that the early Universe was a very different place and in particular the chemistry Andrea you'll love this the chemistry of the early Universe was very different and chemistry actually matters Believe It or Not of course it does because because to to to to form a star you need to cool

it down you need to cool the gas down so it can collapse and become really small before the nuclear fusion you know that that pressurizes it then nuclear fusion lights up and then it can become hot again but there's this intermediate stage where it has to become very very cool and that the way that that happens really depends on the chemical makeup of of the gas that you have because the things that are doing the cooling are completely different depending on what uh chemical elements are there s

o the the most likely explanation for what's going on with James web is that we have overe extrapolated about the way that cooling works in the very early Universe it's a slightly boring explanation but it's by far to my mind the most plausible explanation um of course we don't know that for sure I mean it it may be that once this has all been looked into in detail we still go actually you know what we can't we still can't account for how those galaxies got to be so bright so quickly um but if I

had to bet on it I would bet that that that's going to be the answer not some huge Revelation in cosmology as some of the news articles suggest um okay for dark matter for dark matter and dark energy we we guess and um we we have some pretty good guesses dark matter in particular the fact that we haven't detected it yet here on Earth tells you that whatever it's made out of is very very bad at bouncing off any regular materials right so it it's what we call its cross-section is it's got to be v

ery very small and that in it in an ironic way that makes it actually quite easy to simulate because it does very little other than pull things around through gravity um so it it's quite sort of forgiving in a sense that you you don't really need to know very much about it you just need to assume there is some extra stuff out there it pulls things around by gravity but it doesn't make its presence felt through other forces and that's that's sort of the core assumption that gets baked into these

simulations perfect that's great thank you um the one one question which I think is uh is verging on the conspiracist but it's great is if we can upload the conditions of the universe into a computer to simulate theorize about it uh does that mean that it's conceivable that we are in fact inside of simulation that the universe itself is some kind of simulation and that we are within it yes so I I meant to talk about this but I mistimed it so it's a great it's great and there there's um you know

a range of people have put forward that idea ranging from perfectly serious um uh philosophers like Nick Bostrom for example all the way through to um perhaps slightly less serious people like Elon Musk uh who had you know quite seriously put forward the idea that well maybe we live inside a virtual reality uh how would we know and I'm very skep IAL of this idea and very cautious about this idea because I think there's there's two issues with it the first issue is that it is such a wild extrapol

ation from from where we are today you know that our best virtual realities that we can create in cosmology have around 10 to the 14 bits in them bits is the unit of storage in a in a computer it's about 10 to the 14 uh whereas in reality as far as we can tell that the information in reality would need a well it would be a quantum computer but it would need about 10 to the 100 124 bits um so there really you know there's a wild wild gap between what we're able to do right now and what we're talk

ing about when we talk about could we be inside a simulation now could that Gap ever be bridged well actually the answer is no because you would need to use you can prove that you would need to use the entirety of all the material in the universe to build a computer capable of simulating the universe so you know it's not actually going to be possible ever to do that um now you can you can layer ifs and buts on top of that but I think at the heart of it it is such a wild extrapolation that um it'

s a mistake to take this seriously as though it's a scientific proposition really it's not saying very much more than you know you go back to I don't know decart or or Plato like Shadows on the wall of the cave yes there may be some layer of reality that's inaccessible to us but that's not really in the remit of Science and I think it's a mistake to uh say that anything we're doing in science right now points in that direction so we only have about two and a half minutes left so let me just thro

w some quickies at you and one is you know how do you enthuse young people living in urban centers about astronomy with no nightscape due to light pollution this is from someone who's in a dark Valley in the South Pacific which is brilliant um but coupled with that what impact do you think the vast number of new satellites being put in or by people like you just mentioned him Mr musk and Mr Bezos and so on uh will have for enthusiasm for astronomy and cosmology well I think um it's It's Tricky i

t's the the first um is the answer to the first part of the question um but there are schemes that take people from Urban centers and take them out to see the beauty of the night sky and um that of course is an expensive and involved thing to do and it's very hard to reach large numbers of people in that way but I think ultimately that can have a profound impact on people um and and short of that of course you can also go and talk to them and show them images um but it doesn't of course it doesn

't have quite the same impact uh in terms of polluting the night sky with satellites and so on I I doubt I doubt it going to have any negative effect on people's enthusiasm for astronomy but it's certainly a huge headache for people who are trying to take professional observations of the night sky it's incredibly difficult already to uh take a take a long exposure image of the night sky and not have one of these damn things come in front of your camera and make a complete mess of it um it's caus

ing a lot of problems thank you um now I was going to take two question questions about simulation but I think we're really really close to time now and so maybe what we should do is first of all to thank Andrew for such a great and thought-provoking uh presentation which has covered such a wide range of areas to thank all of you for attending and just to point out that uh on Thursday we've got a really important uh talk to Mark International women's day and that's um asking the question or maki

ng the statement why we need to teach everyone about reproductive health and this is a really big question in the current political climate um and so I hope that you will uh drop in on another one of these lunch hour lectures here from UCL so thank you very much for coming um I've certainly learned a lot and uh Andrew thank you again it's been a real pleasure I'm just sorry that we haven't been able to have lunch together today but uh we'll do that uh sometime soon goodbye everyone Take Care tha

nk you for coming along