Frank Wilczek - Nobel Prize in Physics ► A Beautiful Question: Finding Nature’s Deep Design

Today, however, my topic is quite different. I'm going to talk about what physics will be like in 100 years. Now, of course, I don't know what physics will be like in 100 years, but it's fun to think about. You may think it's quite presumptuous to give a talk like this, and in fact, it is. But I'd like you to know that it's actually quite humble, because the origin of this talk is that Brown University is having its 250th anniversary, or actually, this has now happened. And they asked me to talk about what physics is going to be like in 250 years. So I thought about that for a bit and realized that the closest approximation to a treatment of that subject that was reasonably successful was Isaac Newton's queries at the end of his optics, where he laid out the agenda for what physics might be like in the future. And looking back on that, he didn't say how long it was going to take, but looking back on that, by the end of 100 years, his queries were either being answered or seemed obsolete. So even a genius of the stature of Newton couldn't really see more than 100 years ahead. And so I modestly pared back from 250 years to 100 years, a nice round number. It's humbling to think that if you look back 100 years from today, general relativity was just being invented. Quantum mechanics was in its early days. The idea of the Big Bang, of black holes, of modern quantum mechanics, all those things were unknown. And the idea of gauge symmetries that control the world, the modern standard model, the core theory, none of that was even remotely thinkable 100 years ago. And so thinking 100 years in the future is a dangerous endeavor. On the other hand, it's interesting to notice that if you look back 50 years, it's not so bad. 50 years from now, from 50 years ago, a lot of the revolutions I mentioned were already well underway. The Big Bang had just been discovered. Quantum mechanics was already mature. The ideas of the standard model weren't quite there, but it was close, coming. And if you think 25 years back, actually, a physicist waking up from a 25-year sleep, I think, would only need a few weeks to catch up to the frontiers of what's going on today. So it's not completely unrealistic to think about physics in 100 years. In any case, it's fun. It's mind-stretching. And I hope you'll indulge me for the next 45 minutes or so as we speculate about physics in 100 years. This talk will have two main parts. One is about unifications, which is a grand theme, a great tradition in physics. The other part will be about future technologies that are informed by fundamental understanding in physics. So once when I was giving a public lecture, I was asked, after extolling the glories of possible unification, a question that took me aback and then made me think, why do physicists care about unification? Isn't it just enough to describe how things are? And there are two answers to that. One is that, in the past, thinking about unification of different phenomena, even if the phenomena separately were fairly well understood, has turned out to be enormously fruitful and led to surprising new consequences. A classic example of that is Maxwell's synthesis of electricity and magnetism, which gave birth to electromagnetism, the idea of electromagnetic waves unified with optics and led to a lot of advances in modern physics and technology. Another example that I like to think of because, in a sense, it was even more gratuitous, even less forced by experimental developments, is William Rowan Hamilton. In the early 19th century, just for the fun of it, just for the elegance of it, unified geometric optics with mechanics, got a unified description of how those might work together, and developed the Hamiltonian. And many years later, that turned out to be very close to ideas of wave-particle duality and an essential tool in developing quantum theory. So unification is a great thing. It's also the other thing to say about it, besides that it has been fruitful in the past and has a glorious history, is that it's just pretty. It leads to a better understanding of a more satisfactory understanding of the world when you have a few concepts governing a lot, rather than disconnected concepts. So without further ado, let me describe three important unifications that I think are underway or on the horizon. Oh, I'm sorry, so this was out of order. Here are the success stories. There are more, but this is a partial list of success stories where unification of different areas turned out to be enormously fruitful beyond what might have been anticipated from the separate disciplines. And isn't it pretty? Our current best understanding of matter is based on separate theories for electric, weak, and strong forces, and also gravity, that are only loosely connected, but are based on very similar principles. They're all based on generalizations of the Maxwell equations, the Yang-Mills equations, that support enormous amounts of symmetry. And of course, the principles of quantum mechanics. If you write them down, they all have a strong family resemblance. And so it's suggested, because of that, that maybe they are all aspects of one underlying force that relies on the same principles, but has different aspects. Two special features of our understanding of the standard model, or our core theories, encourage this kind of idea. One is the idea of spontaneous symmetry breaking, which shows that you can have a theory with high symmetry, so symmetry that could unify different forces, where the equations have this high symmetry, but their realization in the world has less, because all the stable realizations of the equations have less symmetry than the equations themselves. So we say the symmetry is spontaneously broken. And our current theory of electric and weak interactions, the so-called electroweak theory, builds that idea in from the very beginning. It's a necessary part of our theories of those forces. And its basic idea was dramatically confirmed in recent years, when the Higgs particle, which was a necessary ingredient to making this kind of instability of the equations, was actually discovered. The other principle, so if we imagine a super-duper Higgs mechanism, we could imagine that all the three symmetries of electric, weak, and strong interactions, based on U1, SU2, and SU3 gauge groups, that all of them could be manifestations of spontaneous breaking of some underlying larger symmetry. The other idea that allows that to possibly be implemented and allows quantitative study of that possibility is asymptotic freedom, which has been alluded to. The idea, or the fact, that the effective strength of different kinds of your basic interaction changes depending on at what distance or at what energy you measure it. In quantum mechanics, when you neglect the effect of masses, and if we extrapolate to very high energies, masses become irrelevant. These are dual concepts. And so simplicity, or evolution of this coupling at short distance, also is equivalent to and implies a similar evolution as you study processes at higher and higher energy. Now, if the underlying equations have a larger symmetry than the separate interactions, based on separate gauge symmetries, SU3, SU2, and U1, then by the principles of symmetry, they should all have equal strength. But as we observe them, they don't. However, if the symmetry is spontaneously, we entertain the hypothesis that the symmetry is spontaneously broken at a short distance, or by condensates, similar to the Higgs condensate in the electroweak interaction, which spoils the symmetry when you look at its effects at energies whose magnitude is less than the value of the condensate. Then we can entertain the hypothesis that a similar thing works on a grand scale, that originally there was symmetry between these guys, and the couplings were equal, but the condensate comes in and spoils the equality. And to reconstruct that equality, we have to run the coupling constants back up to high energy and see what the basic interactions would look like if we strip away the complications due to the structure of the vacuum, due to the spontaneous symmetry breaking and the vacuum polarization, technically. So this is a kind of calculation one knows how to do. Nobel Prizes have been awarded for it, so it must be right. And if you plot the inverse coupling strength versus logarithm of the energy, then to a good approximation, the evolution looks like straight lines. And if you know what kind of particles contribute to the vacuum polarization, you can calculate the slope of those lines. And if you make the correction to see how the couplings evolve, make the corrections due to vacuum polarization for all the known couplings, you see that the hypothesis that they unify almost, but not quite, works. Almost, but not quite. The width of these lines indicates the experimental uncertainty and theoretical uncertainties. We start at relatively modest energies here. So this is something like the LHC. And extrapolate boldly to much shorter distances or higher energies. And it almost, but not quite, works. So how are we supposed to react to this? Well, the famous philosopher Karl Popper taught that the goal of science was to produce falsifiable theories. And here we've produced a theory that's not only falsifiable, but actually false. So what could be better? We've done our job. But of course, that's not the way we approach things. If we have beautiful ideas that seem promising, especially if we have the prospects of unity, we should not give up easily. And these guys seem to be wanting to do the right thing. They're just a little bit off. So the different strengths are approaching each other. They just never come to a common focus. So maybe we've left out something. Maybe we should be even more ambitious. And an extension, a bigger unification that you might add to the unification of the different forces would be to unify the forces and substances. The unification of the different forces I mentioned would unify all the different strong colored gluons, the W and Z bosons of the weak interaction and the photon of electromagnetism all into a common description in one as all being carriers of one larger gauge force. And it would, to a very remarkable extent, unite the quarks and leptons of different kinds, the different kinds of matter particles into one common multiplet for experts one representations of the unification group if you do SO10 unification. But it would still leave us with two things, force on the one hand, substance on the other. An idea called supersymmetry, which sort of goes one step beyond wave particle unification. Let me backtrack a little bit. Wave particle unification already blurs the distinction between force and substance. We think of substance as particles, forces as mediated by waves, as in photons and Maxwell theory. But that applies to single particles. And indeed, electrons, single electrons, behave a lot like single photons. They both can be described as either waves or particles. But when you go to two electrons or two photons, it's quite different. Photons are bosons, so they happily occupy the same quantum state. In fact, they want to. And so in a laser, it's easy to get photons to all act in the same way. Whereas electrons resolutely refuse to occupy the same quantum state. They obey the Pauli exclusion principle. And so that's responsible for the hardness of matter, that the electrons don't want to go through each other, don't want to occupy the same space. Or the fact that white dwarfs can support themselves against gravity. These are effects of electrons not wanting to occupy the same state. They are fermions. So when you go to two particles, even though the one particle behavior is quite similar, when you go to two particles, they're quite different. The unification of those two very, very different kinds of aspects of reality, substance on the one hand and force on the other, is a very non-trivial thing. It was not known how to achieve anything like that until the early 1970s, when Wess and Zumino discovered something called supersymmetry. Supersymmetry is a possible symmetry of the equations that govern nature that allow one, in the equations, to interchange force and substance, interchange bosons and fermions technically, leaving the content of the equations unchanged. So that's a symmetry. Even though you change very different kinds of particles one into the other, so the equations look quite different, their consequences are the same. And that would be a ultimate unification, including not only unifying the different forces and unifying the different substances, but also unifying force and substance together. To do that, however, we need to modify the equations. We have to include a lot more particles to make equations that have this symmetry. By the way, it's got to be, of course, a spontaneously broken symmetry again, because the world does not exhibit the particles that we actually observe don't exhibit this possibility of changing bosons into fermions without changing the mass of the particles. So you have to break the symmetry. But if you include the particles necessary to implement them and give them a relatively modest mass, then their effects as corrections to evolution of the couplings, their effects in vacuum polarization can be calculated and make a significant change to the calculation I showed you that almost but not quite worked. And if you do it quantitatively, you find that now they do accurately come together. It's quite stunning. And as a bonus, gravity so far has been left out of this discussion. But we can apply the same sort of consideration to considering how gravity behaves at very, very high energy. Gravity notoriously at low energies is much, much weaker than the other forces. If I plotted it on this graph, it would be way outside the known universe. So it doesn't appear actually on this transparency. It's out there somewhere. But if I continue the calculation, because the evolution, the dependence on energy is not logarithmic as it is for this running of couplings due to vacuum polarization, but directly depends on the square of the energy, it's much more dramatic. And gravity unifies with the other ones pretty nearly, too. So this is a bonus. These calculations were all done based on just low energy information about these separate interactions. But it solves a notorious problem, potentially, of why the strength of gravity is so different from the strength of the other forces, just as a bonus. So that's why I love Susie. And I firmly believe that this is going to be a component of physics in 100 years. The coincidence is too great for that to be a coincidence. An exciting prospect is that it may not take so long as 100 years. Experiments at the LHC are opening windows that possibly will lead to the discovery of supersymmetric particles, the extra particles we had to include to make supersymmetry of our equations already might be evident at the LHC. We can hope. This would be a very satisfactory development in our understanding of nature. At present, we have two different things, forces and substances, very reminiscent of the Chinese duality of yang force and yin substance. And supersymmetry would show us that you can flip that upside down and interchange one with the other and still have a valid description of the world. So each would show it. So the existence of each, of both, would be necessary, a necessary implication of the existence of the other. A second grand unification that's occurring today, where we can expect much more over the next 100 years, is the unification of the small and the large. This is a famous picture of the anisotropies. This is what the sky looks like in microwaves. Well, not quite. This is what the sky would look like if you filtered out the average value of the microwave background, which is very accurately uniform, and then turned up the contrast by a factor of 10,000. So these are part in 10,000 anisotropies in the energy of the microwave sky. This is ascribed to tiny seed fluctuations at the time of when photons began to propagate freely, liberated from the dense early plasma of the Big Bang. So this is carrying information from the early history of the universe to us, and tells us that the universe started out much, much more homogeneous, but had tiny fluctuations that later grew by gravitational instability into galaxies, stars, and other structures we observe today. And there's a more or less plausible astrophysical history for how that all happened, given these fluctuations. But where did the fluctuations come from? Well, there's a glorious theory of that, based on the idea that those fluctuations are also vacuum polarization, that what you're seeing there is a magnified version, magnified by a process of cosmic expansion at a rapid rate, sometimes called inflation, of the fluctuations that we know are happening everywhere and all the time in space and time, due to the basic nature of quantum mechanics. What you see here is the fluctuations that are calculated in the energy density of gluon fields. If your eyes could resolve distances of 10 to the minus 14 centimeters, and times of order 10 to the minus 24 seconds, this is the kind of thing you would see. We can say that with some confidence, because the equations, I mean, the calculations that give this kind of picture of the dynamics of what's happening have many other consequences we can check, including the vacuum polarization I showed you before, and the confinement of quarks. This is actually, of course, since you don't see it, there's some artistic creativity involved in conveying what you would see if your eyes were different from what they are. But what you see here is just the distribution in energy of gluon fields as it fluctuates due to spontaneous activity in the vacuum, zero point motion, or vacuum polarization, or virtual particles. They're all names for the same thing. Where you see these hot colors, that's where the energy density is the largest. And then you see cooler colors, where it's smaller. And of course, there's a cutoff applied where you leave things empty. That's not because there's no fluctuations, but because you apply a cutoff. It's necessary to do that in order to make the whole thing visible. Otherwise, it would just be a haze. So it's an amazing possibility that these fluctuations could be the same as these fluctuations. There's a lot of circumstantial evidence that points in that direction. But at present, we don't know what it is that's doing the fluctuation. We have a name for it. It's the inflaton. But it's not closely integrated into our fundamental theories of nature. We don't know. It's certainly not the gluon fields that are depicted here. But it might be some other field that appears in unified theories. And it's a great goal of physics over the next 100 years to put flesh on the bones of these speculations and nail down what it is that's doing the fluctuation, if this is the right picture. There are other mysteries about the connection between the large and the small that are very promising, and I think, I certainly hope, will be resolved over the next 100 years. One outstanding mystery is the nature of the dark matter that astronomers have identified for us. There seems to be a lot. If you view the motion of distant galaxies, there seems to be more acceleration than can be ascribed to the visible kinds of matter, the matters we know and love and are made out of and that we understand very well in our core theories. There's some other kind of thing that is five times as abundant as measured by mass. So we call this dark matter. We don't know what it is. We have candidates. It's probably axions. But the experimental verdict is not yet in. And that's an ongoing frontier. And I certainly hope in the next 100 years, in fact, I hope in the next 10 years, we'll know what that dark matter is. And then another unification is understanding what this is. This is a picture of a complicated network of connections. This is what you and I are fundamentally. We're, we think, as physical objects. We're neurons, our memories, our personalities, are all somehow our thoughts, are somehow functions of electrical activity in these nets. How does it work? Nobody knows. So we know what the physical laws are to adequate accuracy, I think, almost certainly, for addressing these questions. The laws of quantum electrodynamics, of quantum chromodynamics, and even if necessary, the weak interactions have been tested so accurately and under such extreme conditions that I think it's overwhelmingly plausible that the fundamental equations that govern those neural nets as physical objects are known to sufficient accuracy. But there's the great challenge of understanding how those basic physical laws give rise to the experience of the world that we have, and in general, to complex phenomena. So let me turn now from the discussion of unifications to discussions of technologies. Arthur C. Clarke, the science fiction author, famously said that any sufficiently advanced technology is indistinguishable from magic, to which I'd like to add an addendum or a scolium, a comment, which is that nature's technology is highly advanced. And I'll let you construct the syllogism. So let me discuss technology on several different scales and in several different contexts. First, making things on the microscopic level. As I mentioned already, we almost certainly know the laws necessary for all practical applications with a very wide definition of what practical means. So specifically, we know the laws, as Dirac already claimed in the 1930s, adequate to describe all of chemistry. So really, physicists should be aiming to put chemists out of business. We should be aiming to replace chemical laboratories by big banks of computers running the equations that we know govern chemistry. That would be much easier, much less smelly, much more flexible than actually having to find the substances, refine them, and put them through various indignities to figure out useful designs. This kind of process has happened in aerodynamics. At MIT, you can find vast, enormous wind tunnels that were once used to test airplane components and possible designs of aircraft. Those were marvelous things, not quite as large as the LHC, but pretty darn big, and marvels of technology in their time and very useful. But now they're not used at all. They're totally abandoned, and they've been replaced by things like this. You calculate airline designs. You don't use wind tunnels anymore. And I think in 100 years, we can look forward to the same sort of thing happening for chemistry. This will allow great flexibility in considering new materials, great creativity in building interesting molecules. So here's an example of something that has been calculated very accurately and has been used to predict properties. The electronic structure of carbon-60 buckyballs, which have a lot of symmetry, so you can actually make a lot of progress solving the equations and get marvelous predictions for how the electrons are distributed and for the behavior. And in general, the structure of graphene, which was a substance that existed theoretically since the 1940s, but was just discovered experimentally how to produce it in the 21st century, and very quickly got a Nobel Prize, and now is being developed into technology. You could make a lot of progress very rapidly once the substance was actually produced, because its properties could be calculated theoretically with great confidence. And you can start to dream about putting these things, these sheets together in creative ways into making molecular sieves and structures in three dimensions. That's going ahead. Sorry, the animation got kind of mixed up, but this is the head of a needle, the eye of a needle, you see. And featured there is a skateboarder. So the future is coming where you can make these molecular designs and a micro world of useful devices that will be really, really small. At present, three-dimensional computers and microelectronics, as impressive as it is, is based on things that are two-dimensional, fundamentally, chips that you've seen them laid out. They're two-dimensional. They're easily damaged. They have to be produced in special clean rooms and in very special facilities that cost billions of dollars to fund. And if they're damaged, they don't repair themselves. You can ruin your computer by putting it in a strong magnetic field or by dropping it really hard. None of those are fundamental limitations. And in fact, none of those are properties of biological systems that do computations. So we can look forward to future technologies where computers are three-dimensional, fault-tolerant, and self-repairing, and not only computers, but machines in general that are self-assembling and self-replicating. This leads into interesting possibilities for macroscopic engineering. If you can build things that are self-replicating, you can have exponential growth and think about really, really large structures. At present, our technology is getting us into trouble with different limitations from pollution, not enough energy to go around with an expanding human population and increasing living standards. A vision of science fiction writers, and famously propounded also by Freeman Dyson, is the idea of the Dyson sphere, where you would surround a whole star by energy-gathering devices that would capture all its energy. At present, on Earth, human energy consumption, as large as it is and as troublesome as it's been, is about one part in 40,000 of the energy that's raining down on us all the time from the sun. So there's a lot that's left on the table. I don't think we'll achieve the Dyson sphere within 100 years, but we will necessarily. And I think we're on the way to gathering much more of the sun's energy and running the world on fusion power, namely the fusion energy of the sun. And the technologies that I mentioned in the micro world, making things three-dimensional, new kinds of architectures that are three-dimensional, fault-tolerant, and self-repairing will enable the kinds of visions that Olaf Stapleton and other science fiction writers pioneered, the idea of turning whole brains, whole planets, or at least mountains or asteroids, into gigantic brains. Another possibility that I find entrancing and I myself have been working on is the possibility of enhancing the human sensorium. It's kind of humiliating to human vanity to consider our vision relative to the vision of the mantis shrimp. Our vision is trichromatic. It's limited to this small band of electromagnetic radiation, even though there's a lot more out there. And even within that band, we take only three averages to get our color vision. But in principle, there's a whole continuum of energy out there, and it encodes information about the properties of substances and objects that are very useful, potentially. In fact, there are creatures, in particular, the champion of the animal world is apparently this humble mantis shrimp. It occurs in various varieties. It's a very successful species, as you might imagine. The varieties have, instead of three, between 12 and 16 different kinds of receptors that also extend into the ultraviolet and into the infrared. We shouldn't be satisfied with, given all the resources of modern technology and the fact that we understand that there's more information out there to get, with the three receptors, we should think about ways of getting some more. Now, there are lots of tricks with modern technology that could enable this. But let me just give you a taste of it very simply with something that doesn't require new technology, but just a little bit of trickery in using existing technology. So this is the famous Nishihara test for colorblindness. The most common kind of colorblindness is so-called red-green colorblindness. It can be simulated pretty effectively by taking an RGB display, like on this computer, and simply setting R and G both equal to R plus G divided by 2. Then the difference between red and green is obliterated. And this picture turns into this one. And you can see the basis of the Nishihara. And this is a pretty good simulation, although not perfect, of red-green colorblindness in the sense that people who have red-green colorblindness will not be able to tell the difference between these two pictures, or in general, between pictures that you apply this process of replacing R and G by their average 2. So the point of the test is that normal vision people will see an 8 here, and colorblind people will see nothing, because it'll look like the same thing. So instead of three-dimensional vision, instead of three receptors, they have two, roughly speaking. You can restore that information, though, with a simple trick. A simple trick is to put in time-dependent modulation that restores the information, opening up a new channel. And when you do that, the extra information that distinguishes the 8 leaps right out. So this is a very simple example of a much wider idea that you can restore the potential extra information that's in the electromagnetic spectrum and allow humans to see some of the things that mantis shrimps do by technological tricks. Ramified minds, that is, minds that extend and branch out in directions, I think is already happening in a sense. The younger people in the audience will think nothing of what, to the older of us, seems like a miracle, the possibility of having Skype conversations all over the world with people and seeing them and interacting with them in real time. This is, if you think about it, a spatially expanding sensorium. We can sense what's happening in China immediately. We can sense what's happening elsewhere. This is still in its early days, dramatic as it is. One can certainly imagine richer feedback from distant places happening all the time. And so you'd have an enriched spatial sensorium with distant sensors. If part of your family lives on the West Coast and part on the East Coast in the United States, you could be in both places at once, in effect, if you have very good communications. You could be experiencing what's going on in both places. Distributed intelligence, of course, is another easier, even more common thing that we take for granted, that we can gather information from all over the world on the internet. That information is available to us. A frontier that's only beginning, but I think in 100 years will be on a par with the others, is expanding fields of action. You can imagine having not just sensors in different places, but also actuators that allow you to do things in other places. This, I think, is likely to be the future of space travel, these kinds of ideas. Putting humans in space is very, very difficult. Our bodies are not meant to go there. They're not designed for it. If you've seen the picture of the Martian, you see an idealized version of the difficulties of survival on Mars. Actually, it's much more difficult. And we can do many of the same things, I think, with our minds and technology without hauling our bodies all over space. A fascinating possibility is quantum minds, quantum computers. These are currently on the drawing board, computers that use the full resources of quantum mechanics in a way that present computers don't. Present computers have a state at any one time that's described by lots of ones and zeros. And the operation of the computer at a very high level of abstraction is just to go from one state of ones and zeros to another state by a program. And that can represent information. But in quantum mechanics, states are not just single arrays of ones and zeros, but amplitudes for potentially many, many states of ones and zeros all at once. So you can do information processing in many, many channels at once by operating on these wave functions. Now, at present, quantum computers are in a primitive state. They are used for things like factoring the number 15 in interesting ways. But the potential, the physical potential is there for doing new kinds of computations. And there are problems known for which quantum computers are more efficient than classical computers. Among those problems are problems of cryptography. So agencies like the National Security Agency are very, very interested in quantum computers, which means it gets funded, which means it will develop over the next 100 years. And I think by the end of 100 years, very likely we'll have very capable quantum computers. And we'll start to think about quantum AI. And these will be new kinds of intelligence that are truly alien, truly mind-boggling. They will have capabilities like, because they can be in several states with different amplitudes all at once, those states can be contradictory. They can have quite different properties. And these minds will have them both embodied at once and interacting with each other. The Schrodinger cat will be both alive and dead. And also, because quantum evolution is unitary, it can be run both forwards and backwards. If you have access to the circuits, there are interesting possibilities for manipulating time. Quantum mind will be quite an interesting thing to think about in quantum AI. So I hope I've showed you that physics has a glorious future, both in terms of reaching new levels of basic understanding in unification, but also in exploiting the things we already know in creative ways to make new technologies that are like magic, truly. OK, so thank you. And I hope we can have some questions and discussion. Thank you.