One of the reasons that the theory of inflation took root very rapidly in the cosmological community, is not simply that it solved potentially these global issues of how the Universe began it's expansion. But it also, without any pre-planning, gave us a potential answer of where the structure in the universe comes from. Remember, we saw the large scale structure in the galaxy distribution. How it could be intensified by the action of gravity. But there still had to be some small, weak density irregularities. Density, as a function of position in space, had to have some small fluctuations at early times. Where did they come from? Well, in inflation, the whole visible universe inflated out of a domain of subnuclear size. That means it's in the regime where quantum mechanics is a dominant process. So in the same way as we saw that the vacuum energy couldn't be quite zero, because you would have to have no quantum fluctuations. Similarly in the small domain the density can't be precisely constant. And its possible to calculate an inflationary model, the amplitude of these density fluctuations. And one can account quantitatively for the seeds of structure that are needed to make the large scale structure in the galaxy distribution, and the fluctuations in the microwave background. So, that's a tremendous leap to think that everything in the universe, all the structure including ourselves ultimately, we are all amplified quantum fluctuations. And it really is astonishing to be able to have a theory of the origin of pretty well everything that goes back to such fundamental pieces of physics. Now we can see why inflation contains a multiverse. Because this process of a small domain being blown up by a large vacuum energy, can happen many times in different places. So one's picture is of a set of seeds, each of which blow up into a large bubble. Observers sit within tiny domains somewhere within these bubbles. So the bubbles function as separate universes. Which can have, potentially, different physical properties. This is the way in which it's conjectured that the vacuum problem might be solved. Supposing, that the scalar field potential was actually more complicated than I showed before. Supposing it had many different minima in it. Each of these would correspond to a potential level of the dark energy. Mostly, this level could be extremely high, but just occasionally you could get one that sits very close to zero. Satisfying the observed constraints on dark energy that we have. Fascinatingly in, in recent years string theory has actually required, for reasons of fundamental physics, nothing to do with cosmology, that such a picture should exist. And they now will say that there should be something like 10 to the 500th powers of these little minima, Vacuum States. And therefore, it's possible within this multiverse to find one in which, inevitably, the energy density of the vacuum would be low enough to match what we see around us. So, the stage is now set once again, for anthropic selection to come in. There are all these different members of the inflationary multiverse, most of which have extremely high values of the vacuum density, it seems possible. Why don't we live in one of those? The answer potentially was given by Steven Weinberg in the 1980s. Weinberg, this is the same Steven Weinberg who won the Nobel Prize in the 1970s for setting up the standard model of physics. So, he's an eminent authority in a wide area of science. He was puzzling over the fundamental arguments as to why the vacuum energy should be much higher than it seems to be. Of course in the 1980s we only had upper limits, rather than actual measurements. But even so, those upper limits were very restrictive compared to a natural scale for the vacuum energy. And he made the anthropic argument. To do with the growth of structure under... under gravity. If you plot the amplitude of large scale structure, (I don't need to say quite how I am defining that) versus time, it grows under the action of gravity. But at some point, it levels off. Where does that leveling off happen? The answer is, where the dark energy begins to dominate. So, as long as the universe is dominated by ordinary matter, gravity can cause structures in it to clump. But if the dark energy dominates, everything expands in too fast and dilute a way for this process to continue. And the larger the dark energy is, the earlier this freeze out will happen. So Weinberg realized that a large value of dark energy would suppress the formation of structure. This, first of all, allowed him to understand why we weren't living in a universe where the vacuum density was large; because there would be nothing there to observe it. Moreover, he predicted at this time that the dark energy density would be seen to be nonzero. Because although there's an argument that means it can't be too large, once you get beyond a point, a certain point, structure can form and then the dark energy density doesn't matter. So there's nothing that drives the value to be exactly zero, it just has to be below a threshold. And he expected, therefore, that since large values were natural, it would be about as large as it could be consistent with that threshold, and this is more or less what we saw in the 1990s. So, this is an intoxicating package of ideas. It's claiming to be able to solve all known problems of cosmology. Wouldn't we be in a good position if we could actually test it, because it could just be a fairy story. So, how to we test inflation? The best test that's known about, is to go back to this general idea that structure in the universe arose, was seeded, by quantum mechanical fluctuations. Because, if that mechanism operates, it shouldn't just modulate the density of matter in the universe, it should affect any field. In particular, it should affect the gravitational field itself. So the inflationary theory predicts so that there should be relic gravitational waves. By which I mean, waves that are the analogue to gravity as a, as light is to electricity and magnetism. Fluctuating lines of force in the gravitational field; time dependent gravitational forces. So these gravitational waves have never been seen directly in the laboratory. So, how are we going to detect them? Well the answer, potentially, came from a Soviet physicist called Starobinsky, in 1979. He realized that gravitational waves would be generated during inflation. And moreover, if you think about the picture. Here is observational cosmology, with us looking back at the last scattering surface, to the CMB. Remember this is t = 400,000 years after, well not after the Big Bang, but after the end of inflation where the apparent Big Bang is. But gravitational radiation would be generated at much earlier times, actually during inflation itself. So from somewhere out here. So gravitational waves would allow us to see to earlier times in the universe. And they propagate through a sphere of last scattering. So, the very material that we're looking at that generates the photons and the microwave background, would be affected by these gravitational waves passing through it. And so, without going into the technical details, the microwave background in principle can reveal the presence of gravitational waves passing from earlier times seeing right back to the epoch of inflation. The idea then of searching for evidence of primordial gravitational waves from inflation using the microwave background's properties, have been around since 1990s. And then in 2014, in particular on this date, an experiment called BICEP2 announced that it had seen gravitational waves consistent with the properties of the waves expected from inflation. If that claim is true, then it's no exaggeration to say that this is perhaps the most important date, from a scientific point of view, in human history. The inflationary universe, remember, looks something like this. If the BICEP2 detection was correct, we can ask, we can learn a great deal about the properties at this point where inflation is taking place. We know that the size of the universe now, divided by the size of the universe during inflation, was equal to roughly 10 to the 55. To put it in context, using your eyesight to using an electron microscope, improves the resolution by roughly 10 powers of 10. That doesn't remotely compare with an increase in fidelity of viewing the small universe of 55 powers of 10. We know that the rate of inflation, so we know that what you would call the doubling time, the time during which the universe increases in size by a factor of 2, would be 10 to the minus 38 seconds. So we know, we can observe, the entire properties of this initial expansion that launched the entire universe. It's amazing, it's too good to be true. Well, maybe it is. Because at the time that this is being filmed, there's a lively debate as to whether the signal seen by the BICEP2 experiment could result from nothing more profound than a little bit of foreground dust that sits between the stars and the Milky Way. Because of course we're in the Milky Way, we observe the distant universe through it. That foreground has to be corrected. It's not clear whether yet that's been done precisely enough to make these claims. So what a fascinating time to live in. We have a complete picture potentially of how the universe began, how structure in it originated, why we're here. Why we observe strange properties of the universe such as low vacuum density. It all hinges on the theory of inflation being correct. We have an observation that may or may not validate that, and we don't yet know whether it's true. So the next year or so, while this question is thrashed out, will undoubtedly be one of the most exciting periods that there's been in cosmology. So the frontiers of modern cosmology really deal with an astonishing variety of ideas: the inflationary multiverse, anthropic selection within them. And yet, from a philosophical point of view, it should be quite disturbing; because in order to explain observations in the universe around us, we've had to introduce a hypothesis of many other universes, which are in principal completely unobservable. Is this science at all? Science should be constrained by observations. We confront theories with observations. The theories either match the observations or they fall. If we can never carry out this test, how can we test the theory? Well, the best that could be said is that science always has some limitations. To take a simple example, in astrophysics the temperature of the center of the sun, you can look up in any textbook, is about 15 million degrees. No one doubts that this is the correct figure, because it's calculated on the basis of physics that's tested, validated, many times over, in other contexts. So, if we eventually reach the point of having complete confidence in a theoretical framework, we should be able to trust the predictions that it makes. Even though nobody will ever go to the center of the sun with a thermometer, we just know that this is the right number. Could it be that cosmology could reach the same state? If potentially, the BICEP results are shown to be correct, or eventually we do detect primordial gravity waves, maybe in the same way, we should have the courage to accept the other consequences of the theory of inflation. And these will include the existence of a near infinity of other universes. Quite like our own in some ways, but with different values of the vacuum energy. In most cases, so extreme that these are completely uninhabitable. They are utterly sterile. And only a very few special members of the multiverse ensemble have conditions that are suitable for life. And we, we turn out to be lucky enough to live in one of them. Because, according to the anthropic principal, that's the only part of the ensemble we can observe.
