 on the people, because there's a lot of good posters, a lot of good presentations. So we had to make a choice. We did have a secret panel, Humberto was on it, but I'm not going to give you the names. No, I'm just kidding. Anyway, so we have that. So we're going to shortly start with the, we'll start with the SPIE prize. The SPIE has sponsored poster prizes. And then the Optical Society has sponsored the talks, talk prizes. So we'll have four of each, and as soon as they're ready. And after that, we're going to turn to our main attraction, which is the ICO, ICTP prize. Where's Roberto? OK. I just want to keep track of you. We also have Filippo Georgi, head of our climate group and a chair of an IPCC committee. OK. So a long time ago, OK. But he's representing the director. The director apologizes. He would normally be here, but right now he's flying to Texas. And so anyway, hi to John. So are we ready? I guess we're ready. What I'd like to do is invite Katharina Svanberg to come down and present the poster prizes. OK. This is Katharina Svanberg, past president of SPIE. It's a great pleasure for me and a great honor to be here. And I have been standing here a few years now and distributed these prizes. And in front of me, I see all these young students. You are the hope for the future. You should remember that we will fade out, but you will come along and take the places of all of us. So I am really, really happy to see you. And please do all the efforts of network with the colleagues here because you will always remember this school. Even when you grow old, you will remember the Winter College in Grinjano outside Trest. I represent SPIE, the International Society for Optics and Photonics, which as an organization like OSA, Optical Society of America, is a strong supporter of this Winter College. And as a matter of fact, these organizations also distribute prizes. And that's what I now would like to do. And I have a bunch of papers here, not only papers, also valuable bills inside. And of course, all of you are very talented, but not all of you can get the prize. But we all should enjoy when people are clever. And I know that all of you are clever, but somebody here, a jury, has chosen a few of you. And that's the ones that I now will call up to the podium here and get a nice certificate, including this little letter, which also hangs on the certificate. I start with a person here whose name is Rostilav Danilo. And you get the 2018 poster award sponsored by SPIE. So please take this, and it's a great honor. So I take up a new certificate, and I read the name Mojtaba Shiroshan. And tell us what country you represent. From Ilay Atosikhan. I'm from Iran. You are from Iran, Mojtaba from Iran. Congratulations. As you know, we are trying as a society to sponsor also the women in optics. And that's a very great pleasure now for me to announce that the two last certificates here, they are females. I start here with Barbara Kassarin, who has also won a prize. Barbara, you come from Italy, from Italy's nation. So we really congratulate you, Barbara. And my last certificate here is the first prize. The three others, they were second prizes. But this is the first prize, and this is given out to Yusra Boisari. Congratulations. And you come from? Morocco. From Morocco. OK. Yeah, so I hope my tie is not too bright for you. Now we're going to give the OSA prizes. So Kari Aptur will do the honors. I'm Kari Aptur. It's my pleasure to be here today to represent the Optical Society. It's also a surprise to me because Anthony Johnson, our past president, was supposed to be here to do this. But unfortunately, he had a travel snafu, shall we say, and he is not here. He sends his regrets, and I will try to stand in for him. As I said, I work for the Optical Society, which, like SPIE, is a global organization serving members and students and professionals around the world. And we are delighted to also sponsor the person prizes here today. Like Katerina, I will start with the three second place prizes for oral presentation. And I'll save the last one for the first prize for oral presentation. So and these are in no particular order. The first goes to Lucian Mandang. The second goes to Nicola Meyer. Did I say your name correctly? The next prize is for Isabel Maria Alvarez-Castano. And the last prize I will present, this is the first place prize for oral presentation. And I believe I can say the first name. I'm not so sure about the second name. Andre Sharpe Baroff. Congratulations to all the winners. And I believe Andre is president of both an OSA and SPI student chapter. So talk to him about what that means. All right, so now we're going to the ICO, ICTB prize. So I'd like to go in the order. I'll bring up Filippo Giorgi to say a few words on behalf of the director, Fernando Cavado. Thanks, Joe. My name is Filippo Giorgi. I'm actually, I don't work in optics. I'm actually a climate modeler. And I work on issues of climate to climate change. I'm here representing a director who could not make it today and sends his regrets. But I'm actually very happy to do this because I was a good friend of Galliano, who really did a lot for the ICTP. Before we start with the ceremony for the prize, I'd like to invite Robert Taramponi, the new president of ICO, to say a few words. Thank you. But I would just say a few words for ICO. ICO, the International Commission for Optics, as you may know, is an organization that actually gathers all the territorial committees from around the world. And also the major international societies, indeed, SPI, OSA, and other international societies are part of ICO. And well, we like to name ourselves the place where the world of optics meets. And this is really what we want to be. And we also want to be the place where optics meets the world for a better world and a more safer and more sustainable world. So we really try to promote optics and photonics technologies all around the world with a special focus on developing countries and on the use of optics and photonics for the benefit of humankind. And among our activities, we have indeed several prizes. One of these is shared together with the ICTP. And just before giving back the word to the ICTP representative, I would just like to mention and thank the committee that has been choosing the winner of the prize. And we have Murad Skal as the chair of the committee. Unfortunately, at the very last moment, he couldn't make it to the college. And together with him, we have here Anna Concertini, Milchod and Ilov, Joni Mela, and Aamadu Vag as the members of this committee. And well, I will wait for the congratulations until we know the name. But I would like to say that we really enjoy to have these prizes. They are the occasion to see young people from countries where they have been working with some difficulties that could nevertheless make the difference in optics and photonics and do really something outstanding. So it's really a prize that ICO likes a lot, I would say, among all the prizes that ICO is giving. And I hope to see many of those who are in this room now candidates for the next year of the prize. Thank you very much, Roberta. So let's see. Okay, this award recognizes the work of young researchers under 40 years of age. They look so young, I have to say, from developing countries who are active in optics and photonics research and have contributed to the promotion of research activities in their own country or in other developing countries. The ICO, the International Commission for Optics and the ICTP established this award in 2000. And in September 2007, they agreed to dedicate the award to the memory and legacy of Professor Galiano Denardo, who greatly contributed to the development of optics research within ICTP and in many developing countries. The award is granted annually for significant contributions in optics and photonics, and the ceremony takes place during the winter college on extreme nonlinear optics. This year, college has seen a significant increase in women scientists, 43% of the participants were women and 23% of the faculty, which are, I think, quite impressive numbers. I think we can come to the winner of the award. So the 2018 prize winner is Professor Urbasi Sinha for her pioneering research in photonic quantum technologies, contributions to cutting edge experimental research in quantum optics and extensive and multi-faceted outreach activities towards popularization experimental optical science in India. Please, Professor Sinha. I would also like to welcome, okay, let's take a picture, maybe here, I have here another very small thing to say that I would also like to welcome Mr. Aninda Sinha, who is the husband of Professor Urbasi Sinha and winner of this award, I'm told, in 2016. So this is the first time that husband and wife actually win a major award of ICTP. So congratulations to both of you for this big award. At this point, let me just say hi to you and also I'll greet you on behalf of Professor Kvedo, who again could not be here. I think now we have a presentation by Professor Sinha. Now we have the best part. The best part? I don't know how to start the presentation, so thanks very much to everybody. I don't think it's, is it on? Well, it's not quite come up yet. So can you hear me at the back? So first of all, I would like to thank the ICTP and the ICO for this huge honor and it's a privilege and of course it's a very humbling experience because I don't think I've really done a lot but then for the little contributions that I seem to have made, you have recognized me as the recipient of this year's ICTP, ICO, Galliano Donado Award. So I'm very, very honored and thank you very much for this. So the title of my talk is Tale of Three Slits from Superposition to Scalable Quantum Computing. So the idea is that we have Quantum Optics Lab and we are of course working on different types of experiments. But because I don't want to necessarily tell you everything of what we do in a short presentation, I decided to choose two things which happen to be two sides of the same coin. So what we have done is in fact, you know, taken an aperture-based system and on the one hand used it to investigate fundamental aspects of quantum mechanics, fundamental tests, precision tests and on the other hand, we have gone on to use the same aperture-based system towards novel architecture for scalable quantum computing. So this is like, you know, fundamental tests and technology in a similar sort of scenario. So I thought this would make a nice story for this presentation. So this is where I am from. So this is Raman Research Institute. It is in Bangalore in India and I would hope that, you know, this talk would serve as an invitation as well for many of you to come and visit us and see what we do in India. And this is a statue in front of the lab and this is a very indicative statue in the sense that, you know, we tend to do that a lot as experimental physicists. Look up and pray sometimes when the results don't happen. So the statue actually, of course, has been there for several decades and my lab is much newer than that. But then, so this is the institute and what we have is a quantum optics lab, which is also a class 10,000 clean room. So the idea is that, you know, some of you may have been to India and of course there is some amount of dust in the metropolitan cities. So in order to have an optics lab which is towards precision tests of quantum mechanics as well as quantum computing and so on, we wanted to have a very modulated environment. Sorry, am I too loud? Is there an echo? Yeah, yeah. So then we wanted a modulated environment with precise control on temperature and humidity and so we decided to actually build a clean room, which of course as you know is a room which is clean, but then with a definition that we have less than 10,000 dust particles per cubic foot and in usually the environment we'll have a couple of orders of magnitude more. So that is the kind of modulation we have and we do research dedicated to quantum information, computation and communication in our lab. So this is how the lab would look on a good day when we have set it up to take pictures. Otherwise, there's a lot of more fibers and a lot more chaos happening, but this is the good day of the lab and of course these are the areas and these are the most important ingredients. I would say they are the ones who actually make the lab function. So these are the students who work in the lab and so we have eight PhD students and she'll start one shortly and so they form the real backbone of the activities of the lab and we have several other types of students, both short-term and long-term who contribute to the different activities. Well, so Richard Feynman happens to be a very, very big source of inspiration for what we do and as you know he's a Nobel laureate and he's a very, very famous physicist and this is a quote that he had made in 1965 that I think I can safely say that nobody understands quantum mechanics. So this was in 1965 and you know, 2018. I think we have to change the quote a little. So what we'll do is we'll say that, I think I can safely say that nobody completely understands quantum mechanics because although we understand a lot more than what we did in 1965, I think there is a lot more that is still left to understand and actually apply. So that is why this is a very, very interesting area because it's still very new in terms of its applications and technology and there's a lot to do and so then this forms the basis for why we do a lot of what we do. Of course, going back to quantum mechanics, so these are examples of applications that you're all familiar with which uses quantum mechanics at some level or the other. So without quantum mechanics some of these will actually function. And so the question comes as to what is quantum mechanics? So of course we know definition wise is a theory which deals with phenomena at microscopic levels. And then of course comes the beautiful implementations of quantum mechanics in light. Well of course our topic today happens to be photonics and optics. So the wave particle duality is one of the most fascinating aspects of quantum mechanics and the Young's double slit experiment which forms a basis for half of my talk actually is has been voted to be the most beautiful experiment in physics by New York Times around three or four years back. So in some sense there's a lot of mystery and a lot of magic associated with different phenomena in quantum mechanics. So this is an example. So you have this, you know the double slit. So you have these single photons in our case with light, you have single photons which are impinging on these slits and they're giving rise to an interference pattern which is a wave light phenomena. So then this is a beautiful embodiment of particle and wave nature in the same setup. Why else is quantum mechanics? Interesting for a more practical point of view I would say that quantum mechanics is a theory which has probably led to the maximum number of Nobel prizes in one area of physics over the last several decades. So we have of course here Max Planck who got the prize in 1918. Who doesn't know Albert Einstein? He got the Nobel prize in 1921. Then we have the Bohr model of the atom and Niels Bohr in 1922. The De Broglie hypothesis, you know the lambda is H by P. So this is the De Broglie hypothesis and he got the prize in 1929. So and then of course we have Thomas, Daviesch and Germer, they shared it in 1937. So indeed the double slit experiment again can be performed using light as well as matter. And so this is in fact, as we mentioned the most beautiful demonstration of wave particle duality. So now of course we have been sufficiently convinced about the history of quantum mechanics in some sense. But then still one open question remains as to what is quantum. And this is in fact one of the areas that our lab is working on in different problems because we still don't know where to find this classical quantum divide or the boundary. When things become important from a quantum perspective and when classical is sufficient. So for instance, if I were to just run and hit myself against the wall, I think we know what the consequence of that will be. I won't tunnel through it perhaps. So because I'm too massive to do that. So when of course the particles become so small that this lambda, the De Broglie wavelength becomes comparable to whatever it is that I'm investigating. Then the quantum phenomena becomes ubiquitous. And so you can't just describe things using classical. So that is one way of looking at when does quantum become important, so to speak. Modern quantum mechanics, of course, we have Schrodinger, Dirac, Heisenberg, and Max Born. And it goes without saying all of them did win the Nobel Prize in physics at different points of time. So the plan of the talk is that I will discuss two different applications of aperture-based systems, as I mentioned. One in fundamental physics, fundamental tests, and the other in quantum computation technology. So the first work deals with predicting as well as experimentally measuring the correction term that is required in the application of the superposition principle in slit-based interference experiments. So I think you might have all heard about the superposition principle. So we have actually found that it is not correctly applied so far in slit-based interference experiments. And we have found the correction term in theory as well as done the first experiment to measure it. So this is a very, very fundamental piece of work, which is important because superposition forms the basis for quantum computing. And then, of course, we go on to using similar aperture-based systems, but in a very different mode. We are using them towards making novel, higher-dimensional quantum systems. I'll explain what that is shortly. And using the spatial degree of freedom of the single photon towards quantum computing. So these are the two things which we will discuss in the rest of the talk. So going on to the superposition principle. Well, of course, I think we are familiar with what it is. But just the idea is that in the classical world, we can have, if we have two states, then it is either this or that. But in a superposition, when you have a superposition, it can be a certain probability amplitude associated with this and a certain probability amplitude associated with that. So that together forms a superposition state. So there's a little bit of this and a little bit of that. Whatever this and that might be, which, of course, is more concretized in this example, where we are talking about a classical computer, so which is what we are using right now. So that is based on, as you know, classical bits. So this is the on and off state of a transistor. So the on could be the one state and the off could be the zero state. So this together forms the classical bit. So if it is off, then it is off. If it is on, it is on. So that is how we know the classical bits to work. Now if you go on to the concept of a quantum bit or what we call a qubit, right, then comes this interesting picture where the zero and the one form the two poles of a sphere called the block sphere. And then any point on the sphere represents what is called the superposition state. So we have the alpha zero plus beta one state where alpha and beta, of course, satisfy certain normalization conditions. And essentially there is a certain probability associated with zero and a certain with one. So in principle, as it mentions, that a qubit might seem to contain an infinite amount of information because its coordinates can encode an infinite sequence of digits. So of course, we will see what that actually means for quantum computing. Now here we have, you know, this is the spin. I mean, we're just discussing spin here with this arrow. So we have a single spin, that is a single bit. So that could be either up or down, you know, just like the on and off. So that would be the zero state or the one state. So if it is zero, it is zero. If it is one, it is one. So two possibilities here, but not simultaneously. But in a quantum bit, you can have them simultaneously. So you have access to two states already. But of course, the beauty comes when you have more than one. So if you have two spins, then you can have them in 00, 01, 10, or 11, right? So these are the four possibilities which could happen, but not together. But here you can have, you know, the alpha 00 plus beta 11 plus gamma 01 and so on. So that gives you access to two square, which is the four states, right? With three, again, it's three. And here you have two cubed, which is eight. So then you can see that this is an exponential advantage that we are building up here. So with three qubits, we have eight possible states. And then of course, once we go to, let's say, 50 qubits, then we have access to two to the 50. And can anyone here tell me what this P stands for? I mean, are you sleeping already? If not, then what is P? Okay, so this is also not a surprise. I've asked this question many times now. And it has always woken people up, but then not with an answer. So P is supposed to be a pillion. And so we know million, billion, I mean, you know, ICTP, ICO, you know, of trillions and so on. So you're very rich organizations. But I don't think you know what is a pillion because a pillion is beyond our number system. We don't even deal with it when we study numbers. So we go up to a trillion, perhaps, and we stop. So pillion is something which is quite unimaginable. And that is what the number of states we would have when we have 50 qubits. And this is supposed to be the holy grail of quantum computing that we want 50 qubits in a coherent superposition. And that would give us this quantum processor, which will have this exponential advantage over classical. So this is, of course, at the moment, as we know, that there are several technology giants like IBM and Google who are all claiming to reach this number 50 this year. And so 49 is what somebody's saying, 50 is what someone else is saying. And then, of course, it comes with a catch, though, that this is not going to come with what is called error correction. So we have to take it with a grain of salt. We might have a 50 qubit processor, but it will not have the error correction inbuilt, but we are getting there. So quantum computing is not a technology of a distant future. It could be present and very near future. So that is how we have brought it down over the last several decades, right? So there are several devices which are contenders for the qubit. You know, it's a two-level system. So you can have two levels discussed in many different devices. So these are all examples. And again, of course, my favorite topic of the Nobel Prize. So we have David Weinland and Sergei Hiroshi, who have been awarded the Nobel Prize in physics two or three years ago, due to their contributions to the photon and the ion trap technology in quantum technologies. So then it's as new as that. And so a lot is happening in quantum computing right now. So what is the first work that we're doing related to in terms of what I've just talked about so far? First, we will talk about superposition itself. I think we understood why superposition is important for quantum computing, because of course the superposition state is what gives us this exponential speedup. So what have we done for the superposition principle? So here you have, again, the double slit experiment. So we are all familiar with it. I'm saying that I have two slits, A and B. And then what I'm telling you is that I have a source, I have a detector, and the source could be a source of anything. I mean, of course, for the current talk, it could be a source of photons or electromagnetic waves, but it could be also a source of electrons. Any particles or wave, we don't care about it right now, but it's a source and we have a detector. And the slits actually present it with a boundary condition, right? So we are used to solving the potential wave and potential barrier. The slits are just another boundary condition that we can impose on waves which are impinging on it. So now I'm saying that I'm closing slit B, just making slit A open, okay? And I'm saying that the solution to this wave equation, I will call it psi A, because somehow I want to be partial towards Schrodinger equation, but that is just, you know, in the mindset. I have to call it something, I call it psi A, okay? Then I close A, open B. Solution is, let's say, psi B. The question that comes up is that if I open both A and B together, then what is the solution? And the answer that we find in many textbooks, very well-known textbooks of quantum mechanics, as well as optics, and what we have been using as well till, I would say, 2012, is psi A plus psi B. So we say that we have psi A and psi B when we open both slits, and then the sum of the two gives us a solution to this boundary condition. But that's, of course, not true, right? So as you may remember, the superposition principle is applicable only when the boundary condition remains the same. So if I have a solution A, and then I have a solution B, and for the same boundary condition, then I say the linear superposition of A and B is also a solution. That is how we define this in differential equation solutions and so on. But here, of course, these are three different boundary conditions. Slit A open is one. Slit B open is one. A and B open together is a third one. So I'm not really allowed to add A and B and call it a solution to both A and B together. And so this is something which we have been using. And of course, what we feel is that now, of course, as we explain it to you, it's obvious to everyone that this shouldn't be the case, but then why are we doing this for several years? It's probably because the correction that will come to the superposition principle, application in slit interference experiments will be very small. And that is why it's probably okay to ignore, which we do a lot in physics, right? We have this assumption that, okay, we can ignore something which is small. And so that is how perhaps we have been doing this so far. So now, Yabuki in 1986 actually worked out a path integral picture of this. You know of the Feynman path integral formalism. So it's basically a technique which is used to find the probability amplitude of let's say if you have A and B as two points of going from A to B. So it gives you the sum of all possible paths is the total probability amplitude. So that is in a nutshell the path integral. So if you now have both AB open, then there is the possibility of this sort of a path. So please bear in mind, this is a Feynman path. It may not be a path real trajectory in a position in space time. So it's not that it's going backward in time or anything like that. This is a Feynman path. And all paths are possible in the path integral. So this is also a possible path. Perhaps a sub leading path as opposed to the more normal ones. But this will only come when both A and B are open. If they're individually open, then it won't happen. So these sort of paths will make a contribution which will be ignored if we just make that assumption that A plus B is AB. And so then he had done this for the double slit. But then the problem was that his calculation was extremely theory oriented in the sense that he took point slits and plank length. So essentially everything was very idealized. So in some sense, the assumptions tended to be a little bit incorrect. And so we cannot just use that towards an experiment. So my goal with any theory that I would do would be to also do the experiment associated with it, which is what of course we will lead to. And so then this was not very experimentally friendly. It was of course a beautiful observation. So what we went on to do in 2014 is actually quantify the effect due to these subdominant paths in the path integral and find a quantity which will be zero if the superposition principle is correct as it stands, you know, the application and would be non-zero if the superposition principle requires a correction term. And so then we propose a precision experiment which is a null experiment. And I'm sure you've heard of this. So null experiment is one of the most difficult things to do because you have to measure something to be a zero or a non-zero. And as you know, a zero is not just a number here. It is essentially a bound. So any experiment will come with lots of errors. So these errors will contribute to a certain bound for the quantity. So if you measure the value to be less than that bound, then it is clearly zero. You have to measure it higher than the error bound which is what creates the trouble with null experiments. But then this is what we proposed here in this paper. And now it seems my presentation is not going forward. Okay, it is, yeah. So then, so this was the idea that you know, using path integral and I'll just skip straight to the, so this is path integral. I'll go to this video. So the idea is now you have a triple slit setup and just to summarize what I've been saying. So there would be paths of this kind which would cross the slit plane once which would extremize the action, as you know. And so these would be the classically dominant parts, right? And so we call them in short classical parts, okay? And then there would be parts of this kind and this is just an illustrative example which will cross the slit plane more than once. And so these will not extremize the action and we call them, you know, classically subdominant or in short non-classical parts. So these parts are not quantum parts or anything. So this is just classically subdominant. And so these contributions would actually give the correction term to the superposition principle which is what we have quantified. Now why triple, I mean, you know, again, question. So you must have noted that the example I took was a triple slit experiment. But of course I've been talking about the double slit for some time now. So why did I take the example of a triple slit? Now this is a very important question because in the double slit also the same thing will obviously happen, that there will be a correction term. But what will the correction be in? It'll be in the interference pattern that you observe, you know, the beautiful double slit interference pattern. So with or without the correction term, you will observe an interference pattern. And because this is subdominant, the difference between the two interference patterns will be so small that it'll be very difficult to establish that it exists, this difference exists beyond the error bound. Because when you do an experiment, as you know, you will measure an interference pattern with error bars. And these error bars might actually agree with both the theories. So then it's not a very clean thing to do in an experimental scenario. But the triple slit is actually something which gives us this quantity kappa, which is zero, as I mentioned, which is zero if the correction is not required, and non-zero if it is required. So of course then it is much, much cleaner because of course if you measure a non-zero you're already proving that this is required. It's not a non-zero one versus a non-zero two, as in the case of the double slit. So this is the difference between double and triple, makes the triple very important. And of course what we are doing is using interference as our tool here. So this is Rafael Sorkin who proposed what is called the sum over histories approach to quantum mechanics. But then again I'll not discuss that in too much detail, but go on to the triple slit and why it's important. So if you have the single slit open, then of course what you get is just the, probability is just simply, the intensity is just the probability, so it's proportional. So there's nothing very non-trivial which happens with the single slit here. But if you have two slits open, then the interference, the first order interference as we call it, is the probability of both open minus the singles. So that of course is non-zero in quantum mechanics and that is why we know that we have, we talked about the wave particle duality and so on. So we have this wave nature manifested and that is why this actually is a proof in some sense of this is consistent with quantum mechanics. But now if we go over to the second order interference which is the triple slit term, that would be P of, I mean the probability when all three slits are open minus the sum of the doubles plus the singles. So it's just plus minus plus. So it is going in a very systematic way, the definition. And so now if I say P of all three slits open is mod psi A plus psi B plus psi C whole square which of course is just the bond rule as you know, the probability being mod psi square is the bond rule. And then I have said that the wave function is just the sum of the three which is of course the assumption that I'm challenging. But then if I make that assumption and do the little algebra, then I ABC turns out to be zero. So this actually assumes both the naive application of the superposition principle as well as the bond rule for probabilities. So now over the last few years several experiments have actually happened for the last 10 years using aperture systems and diffraction ratings and so on, trying to measure this quantity I three, the normalized version is called Kappa. And none of them have actually measured a non-zero so far. But the whole community has been focused on measuring this non-zero as a test for the bond rule. Because of course, you know, if the bond rule happens to have a correction, then everything as we know it will cease to be the way we know it. So it's a huge thing to happen if it does happen. So people have been focusing on trying to get a non-zero and if they do then they would challenge the bond rule. And then that would lead to generalization of quantum mechanics and perhaps unification of generativity and so on and so forth. So then this is the big game. But then of course what people have missed is that, you know, the application of the superposition principle itself is not complete in these experiments. So if I applied correctly, this quantity will be non-zero anyway. And so measuring a non-zero i3 does not necessarily imply falsification of bond rule. But then it could simply be the boundary condition effect which we are using incorrectly so far. And so this is what we found and proposed that first you better check this before you challenge everything as we know it because that's a bigger challenge to face. And of course as an experimentalist if I actually measured something which violates bond rule, I won't know what to check it against. So I'll be at a complete loss on what to do next. So then this is the first experiment that in fact I was involved in which was a triple slit experiment towards measuring this Sorkin parameter as we call it. And of course none of them has measured non-zero so far. So as just to summarize, so what this means is a non-zero contribution to kappa will imply that the naive application of superposition principle needs to be corrected and non-zero kappa does not necessarily imply falsification of bond rule. So this was the first non-zero estimate of kappa and then this is the path integral language and if there is anyone here who likes the path integral formalism for perhaps people who do quantum field theory then they would actually see this as an experimental validation of the full scope of the path integral formalism because here we are actually validating those subdominant parts which we tend to ignore in most calculations. So this is another take on the story. So this was our numerical simulation. So this blue curve that you see is kappa as a function of detector position and the red is the interference pattern from parameters which were taken from my first experiment here. That's very important because I had done the first experiment to measure this kappa and I found something and now I found a theory which says that it was meant to be non-zero. So I need to check whether my first experiment is still valid and of course with a very, very heavy heart I was checking this and it turned out to be very, very valid because of course the kappa that I would need, the non-zero kappa that this would bring is 10 to the minus five level whereas what we had measured earlier was 10 to the minus two. So it wasn't precise enough to measure this non-zero anyway. So whatever we measured was a zero with that bound and so then we also went on to do the simulation for electrons using parameters from this paper here and again kappa is a non-zero, modulating non-zero of the order of 10 to the minus eight. Then we went on to actually find an analytic form for kappa with certain assumptions because I wanted to do the experiment to measure kappa. So if I had a form which is dependent on certain parameters it would help me in designing such an experiment and so that was, I will not go through the details of the various assumptions but this yellow box is very important because it tells you what is the maximum contribution that this non-zero can have in a certain experiment. So that is proportional to lambda to the three by two by d to the half times w. What is lambda? Lambda is the wavelength of the photon, electron, whatever it is that you're doing. D is the interslit distance and w is the slit width. So now this tells you what to do. I mean, if you want to design an experiment where you really want to see this non-zero kappa then you better deal with something which is a very large wavelength. So then of course everything will crank up because of course up to a certain extent you can control the errors but then beyond that if your kappa is already expected to be a certain large number then it helps because you can't control errors beyond what you can control. I mean, as you know, there's only so much you can do and the rest is instrumental. So then this would help you in designing such an experiment and of course, you know, the numerical and analytic form matched very well. This is also a very important graph for me because the blue line that you see is our analytic form. The black dots are the, you know, from this paper. Where is the, yeah, so there is this paper here which dealt with finite difference time domain simulations of Maxwell's equation and so they predicted this black line for kappa. And so then the blue and the black match very well, of course. And so this tells me that, you know, so those guys they had done it using three and a half days of computation time on a supercomputer and 6.3 terabytes of memory and our formula of course worked on this beautiful machine sitting in front of me using Mathematica. So this is something that is also a very instructive thing because it tells you that you don't have to always do the hardest thing to begin with and because you may not have the resources and we come to the concept of, you know, in our country we may not have access to supercomputers just like that. But then it works fine. The formula is beautiful, gives a nice match and at least tells me whether to be worried about this contribution or not in a certain experimental scenario. So which is what we went on to do. Apologies to the theorists in the audience. But then of course, you know, this is another quote that I really like by Richard Feynman. It says that it doesn't matter how beautiful your theory is, it doesn't matter how smart you are. If it doesn't agree with experiment, well, you know. So then the idea is that if I have some theory which is experimentally testable, then I should do that. And of course, optics is the best to test several things and that is why I think most of the fundamental experiments in quantum mechanics, whether it's Bell inequality or, you know, entanglement, everything has been tested using optics testpad to begin with. And then we have done other systems. And so this is, you know, as I mentioned, the first experiment was to test one rule which was in fact done using single photons, heralded photon source. And so this was the result. So essentially we bounded kappa to 10 to the minus two. So this was the first bound on this quantity. And so it was zero up to 10 to the minus two. Then different groups took up different versions of the experiment. So they did this, you know, this grating version where the better, you know, the bound was better this way. It was 10 to the minus four. So of course, you want to make it better and better with different versions. So that finally you can come to a version which is an actual non-zero. That's the hope. And then the NMR implementation also was somewhat 10 to the minus three. And then we have this new paper in Nature Communications by the Boyd group where they actually claim to have found the, you know, these looped trajectories as they call them, which we call the non-classical path. So they claim that they have found an experimental verification of our theory in the triple slit scenario. So they have also used a single photon source. But what they've done is that the slits they have manufactured in a clever way. They have used a gold film and made slits there. And so what happens is gold actually leads to what are called surface plasmas. So then this near field effect leads to surface plasmas which leads to an enhancement of this quantity, which, you know, this quantity kappa. And so this is their way of getting this enhancement. Of course, one has to understand that this is a material-induced enhancement. Whereas kappa, as we have found and know it, is a geometry-based parameter. So they have not used the geometry much to their advantage, but it's the material. And then they went on to do this. So, you know, this is the kappa as a function of detector position. So these points here, these round circles that you see, this is the experimental data. Of course, you know, this is the theory curve. So what is interesting, however, is that they have been able to only measure the negative values of kappa. They don't have the precision to somehow measure the positives or anywhere in the slope. So the theory, of course, is perfectly modulating as expected. But the experimental data only fits the negative values for some reason. So, indeed, we could do better than that, and which is what we went on to do. So this is our experiment. So, you know, after all the discussion about the modulated environment in my lab, I'm discussing the first experiment which happened outside the lab. And this was in an open field, as you can see. This is a field in RRI, which is a volleyball court. And so then these are my three slots. And this is an antenna, which is a source of electromagnetic waves, antenna as a detector. So I told you to remember that yellow box and the formula. So if I increase lambda really drastically, then there's a possibility of getting a larger contribution to kappa, which is what we went on to do. Drastic being, of course, really drastic. So we went to centimeter wavelength, you know. So instead of the nanometer that we are used to with single photons, we decided to go up by several orders of magnitude. And then this is the final version of the experiment in this nice and colorful tint. So this is 100 kilometers from our institute. It's an observatory, which is used for radio astronomy-based experiments. So it's a quiet zone. So it does not have any interference from gigahertz frequencies. And it also has a corn field here, which, you know, the corn grew and was cut twice during the experiment because it took two and a half years. We enjoyed the corn. And the corn actually is a very good absorber of gigahertz frequency for some reason. So when the corn was actually there, the data was better because it would absorb any stray. So it's a very romantic description of an experiment, which I can go on about. But then this is what we ended up doing. And these are the experimental parameters. So as you can see, much larger than what we have been used to. So five centimeter wavelength, a slot with 10, 13 centimeter inch slot distance. And of course we wanted to be in the front of a regime. So we had to have distances in meters. So the slot and the source and the detector were separated by three meters and four meters and so on. Of course, you can't go on forever because then the signal drops and you don't measure much. But then you can go up to eight meters and still measure a good signal. And so this is the schematic of the experiment. And so then, you know, this is the graph that you would expect if only the so-called classical parts contribute to kappa. And this is our graph. So this is the theory, the blue, and the red circles with the error bars. This forms the experiment. So this is the beautifully modulating kappa as a function of detector position. And I would venture to say that the theory and experiment match rather well for a precision experiment done in an open field. So then of course, so this is what we got. And then we also changed the distances just to prove that it does work as a function of changing distance and so on. And then of course, still there would be skeptics who would say, maybe I'm missing out on some error contribution. And this is all just an error that I didn't think about. So then of course, we have this final experiment, which we call the baffle experiment. This word baffle was actually suggested by Professor Tony Leggett, who was very, very fascinated by this entire superposition story. So we have these absorbers, which we have placed perpendicular to the slit plane. So in some sense, the contribution from these hugging parts, which forms the maximum contribution to the kappa parameter, would then not come into being. So now what we have done, we have increased the baffle size. And as a function of that, we would expect that this contribution to kappa would decrease, which we have actually found experimentally as well. And then beyond a certain size, of course kappa becomes comparable to errors, and it becomes 0. So what we have is a tunable experiment, whereby we can heighten the effect of these non-classical parts, and then use a tunability parameter in this baffle to actually remove the effect altogether. So we can remove it and bring it back. And this is something that of course was impossible with the photon experiment. This is because of the macroscopic nature that we could actually do it. And of course, our experiment also has implications on precision cosmology. Because even in cosmology, what they do is they use these arrays of dipole antennas as a receiver system for their signals from the epoch of whatever, reionization or recombination. And those are very weak signals, as you might expect. And so they try to increase this signal by interference effects. And in order to do that, when they do the simulation, sometimes they just add the contributions instead of taking into account this correction term. But then we found this graph from their parameters. And this is our contribution to kappa from the cosmology parameters. And it is, again, a beautifully modulating function. And this is of the order of the same order, tensor minus 2, which we have already measured. So then precision cosmology experiments should take into account such correction terms. So this is what we have concluded. So basically, we have the first observation, so to speak, of these sub-dominant parts and slit-based interference in a classical domain on top of that. Tunability parameter makes this unambiguous, because we can kill it and bring it back, so to speak. And path integral is, of course, an overarching framework and transcends the classical quantum boundaries. And our experiment serves as a test pad for future experimental design for bond rule tests. Because of course, if you can make this go up, then you can also design an experiment which kappa is really small by, again, making use of that length scale. So if you really want to do this bond rule test, then you have to have contributions from these boundary effects really diminished, which also we know how, because of our analytic form. And of course, yeah, so this is something which we really feel the last point, because it's a very simple experiment, so to speak. And according to us, it's rather beautiful, because it gives us a precision experiment which serves as the first non-zero observation of this quantity and the correction to superposition. So this is a fundamental test and an experiment dedicated to that. From that, I actually would like to spend the next five or seven minutes on the other application, which is on scalable quantum computing. So this is the other side of the coin. So you have to do have, this gives us the understanding of the system, but the application is towards a very high-end technology, which is a quantum computing architecture that we are building. So that brings us to what is a Q-trit. Now, of course, I discussed what is a Q-bit and so on. That is a two-level system with certain conditions. A Q-trit is actually a three-level system. So in Q-bits, you have 0 and 1 as the two basis states. So in Q-trit, you can have 0, 1, and 2. And the superposition is then alpha 0 plus beta 1 plus gamma 2. So this forms a Q-trit. And indeed, what we have done is we have used the three slits as the three possible spatial degrees of freedom of a single photon. And this is how we define our Q-trit, because of course, as you know, a photon, the most popular degree of freedom is polarization. So you have the h and the v, and that forms a natural Q-bit. But if I want to go to a higher dimensional system, then I need something else. And so we have used the spatial degree. And in principle, this could scale up to a very large number. So this is our Q-trit. And Q-trits have certain advantages. They are known to be more resistant to noise than Q-bits in quantum cryptography. There are proofs that they are more secure against digital attacks. And also they have some foundational advantages. But then of course, the most interesting advantage which easily is appreciable is this 2 to the 50 comparison. So we discussed about 50 Q-bits, and we want 50 Q-bits. But we don't yet have it, right? That's because it's a problem. It's like it's almost sociological. So when you have a large number of people in a room, and you put more and more, they don't like to be in the same room beyond a point. Because it leads to some sort of antagonism or whatever, right? So the Q-bits are like that. So beyond a point, they don't want to be in a coherent superposition. So that's why we haven't reached 50 as of now. But then what do we do? One thing we can do is keep on trying, which of course the community is doing. The other is, instead of 2 to the 50, I change 2 itself. So of course, if I change 2 to a higher number, the 50 will drop. So 2 cubed is 8. 3 squared is 9. So 2 cubed trits would give me a similar complexity of the Hilbert space as 3 cubed bits. So in some sense, this number 50 will drop to a much smaller number if I dealt with a cubed trit. Because I've changed the base itself of the problem. And so this is an example that I like to take. Here I have a football match. And of course, it's ICTP. So let's say it's between Italy and Germany or something. And then of course, there's only one result we want, which is perhaps a win for Italy. But there could be three more unfortunate loss. Or there could be maybe a draw or something. I mean, we go on and on, nothing happens. And then it could be a rain or something. And so it's abandoned. So four possibilities I can think of in a football match. And so if I want to represent these four possibilities using a qubit, I cannot do that. Because a qubit only has two states. So I can represent the win and the loss with one qubit. And then the rain and the draw with the other qubit. But if I had a cube quad, which is a four-dimensional system, then all the results can be declared using one system and not two and so on. So that way, I can actually pack in a lot more information in a higher-dimensional system, which is what we are doing in our lab. So of course, we also are working on entanglement. And we need entangled states between Alice and Bob for our quantum cryptography applications and so on. And so what we have is a novel Qtrit architecture using the spatial degree of freedom of the photon. And what we have used is a technique called pump beam modulation, which I'll discuss in a minute. And it leads to highly efficient, robust, and scalable architecture. These are all very important. Because if it's not scalable, then I'm just stuck at three. But I wanted to go higher. Because as I go higher, the exponent drops, which is what I'm interested in, after all. And then, of course, at the moment, our current test bed is of bipartite Qtritz. Bipartite is two-state system. So we have two Qtritz, which are perfectly correlated. We have shown a high degree of spatial correlation between them. And then, of course, recent results, which are not yet out, certify and quantify entanglement in the system, and prove that we have a maximally entangled state. So that is, of course, great. But now that we have a maximally entangled state in this novel system, we will be able to use this towards several possibilities in higher-dimensional applications, so whether it is teleportation, whether it is quantum keys. So we are going to play with different applications of Qtrit, Qtrit, entanglement, now that we have shown it. So this is the idea, and so we have lots of references for this, which I will skip now. We had shown, using a single Qtrit, that we can play the R-on-Vaidman quantum game. So quantum game is something which is of interest to many people. So this shows that a quantum system has a certain advantage over a classical system, and so on. And the quantum one will have a win percentage, which is, in this case, 87. And the classical is the 50% limit. So we had a much better person. Of course, we'll never get 100 in an experiment, as you know. But then we are much closer to 150. So this was one implementation we had done earlier. So what is this pump-beam modulation? Just a brief description. So this is what is called spontaneous parametric down conversion, which forms the basis for a lot of what we do in the lab, which is how we make our photons, and so on. So here you can see that if you have a certain polarized photon, it basically gives rise to two photons, the down conversion process, which could be sort of correlated in polarization. So if one is h, the other is v, and so on. And so the idea is that you can get a coincidence speak when you do a certain time correlation measurement between these photons. But here we are restricted to qubits so far, right? Now what has been shown, however, is that the single photons, they carry the information of the spatial profile of the laser beam that generates it. So this paper had shown that. They had done it for a different purpose. They just wanted to transfer some information, nothing to do with quantum computing. But then this information is carried through. It's almost like children carry certain traits from their parents. But in this case, it carries the entire spatial profile, the photon, because it's not quite human. So in some sense, this is of great advantage to us, what we have done. So I'll just come to that. But here this is another way of doing the same thing. You can place these aperture systems after the photons have been created. And then, of course, they would carry that information and so on. But then this has nothing to do with the production level. It's after they have been produced. But what we have gone on to do is, in fact, modulate the pump beam itself with this aperture system. So as we say here, keeping slits in front of the generated photons, we can only deal with a subspace. But then when we need to generate the photons themselves with those three levels, we want to change the pump. So we have actually modulated the pump. So here, I mean, I'll just go over quickly to the, OK, this is how the experiment would look in the lab, perhaps not very illuminating, so to speak. This is much better. So you have this laser beam. And what I do is, I place a slit system in front of the laser beam. So the laser beam is known to have a Gaussian profile, as you know. And so when I place the slit system, this pump then carries the slit profile with it. And by using an imaging technique, I can then have that incident on the non-linear crystal. So now the pump beam itself has a triple slit profile. So the two daughter photons, as we call them, that come out, also carry the triple slit profile. And so in this way, I've actually generated two Q-trits directly from the down conversion process by modulating the pump beam. And this is a very new technique, which hasn't been done before. And so these are the conditions for our experiment. Essentially, we have one of the detectors fixed, and the other one moves. And by that, we can find out whether we have the perfect correlation or not. So if you have a Gaussian, it gets perfectly transferred to the daughter photons. But then of course, of interest to us would be the slit itself. So this is the pump profile at the crystal. And this is the single photon profile at the detector. So I mean, of course, there'll be a little bit of broadening, because the photon is going through the air and so on. But then this is indeed a very faithful transfer of the profile to the daughter photons. So then this is the idea. So now if I place the detector in one position on one side, then the correlation graph that I can plot on the other will show a peak at exactly the position of the twin photon. So and nowhere else. So you can only have the peak at. So if the first photon is in slit A position on one side, then the correlation graph will peak on the slit A position on the other side. So we have perfect 1, 1, 2, 2, 3, 3 type correlation. And so this is how it happens in experiment, as you can see. So you have, so these are the experimented. So this is the singles profile. And then one of the photons is here. The correlation peaks here. When it is here, it peaks here. And when it's third, it peaks. So then it's actually a very high correlation percentage that we found of around 0.96 using the Pearson correlation coefficient, which has been around for more than 100 years. And it's a correlation function between two distributions, which we can define. So it's been used for a long time, but not necessarily for the purpose of quantum systems. And so this is what the experimental results we found. And it's a very high correlation. So then of course, till now we have a novel bipartite cutrate system, which is highly correlated in the spatial degree of freedom. But what next? I mean, while there are spatial correlations, how do we know if these correlations are classical or quantum? That's very important. And so we can perform what is called an entanglement witness measurement to certify or not certify the entanglement. But what we went on to do is kind of better. This paper by McConaughey et al. They showed the use of the Pearson correlation coefficient as an efficient measure for entanglement in bipartite pure qubit systems. So in this second work here, which will be uploaded on the archive shortly, we have shown that we can effectively use this Pearson correlation coefficient to both certify and quantify entanglement in bipartite pure and mixed cutrate systems. So we have actually extended this to higher dimensions. And in this experimental work here, which we are very fascinated by, this recent experiments show that indeed we can use this beautifully in the experiment. And we have almost maximally entangled cutrates in our lab. So right, so we have generated this. And it has an advantage over previous systems. It has robustness. It has scalability. And then of course, as I said before, now, of course, we have entanglement. That is a big thing for us. So we will use this as a resource in different cutrate-based QIP protocols. So this is basically, I think, almost the semi-end of the talk. So I would like to spend the next minute saying that this is not all we do. This is just a little pick that I made to just discuss with you today. There are lots of other experiments that are going on in the lab. And one of them is on quantum key distribution. And in fact, recently, we have started a flagship project in collaboration with the Indian Space Research Organization, ISRO, which is towards quantum experiments using satellite technology. And of course, the idea is that we will have a ground station in India and a ground station somewhere else. And we will establish a secure key between these two using a satellite as a trusted node. And of course, the fact that China has launched this satellite last year and the first results are coming out is great for us. Because of course, that proves that it's not science, fiction, it's science. And so that, of course, got us the funding. And so then now we have started this experiment with a lot of enthusiasm. And our collaborators on this would be IQC in Canada. So we would like to establish a key between Canada and India to begin with. And so I would like to end this talk with this, you know, with this rather nostalgic postcard. So the idea is that, you know, of course, this talk is very special to me. And of course, this event is very special because, you know, it's like a little recognition, which always helps. I mean, of course, we do science because we love discoveries and so on. But then sometimes when someone tells you that, oh, OK, you're doing all right, that is actually a very nice impetus. And so once such impetus came in 2010, when we got a letter from the son of Max Bond after doing the first experiment to test Bond rule, he's Gustav Bond. He's 97 years old as of today. And he was, of course, not very young even when he wrote the letter. And so then he tells us, you know, that in 1926, when he was five, you know, he actually heard about this Bond rule from his dad. And he's very gratified that we did this experiment to verify and test Bond rule. And he wants us to keep him, you know, informed about what we do and so on. So this is very, very nice because, you know, it's not an award, but it's actually very rewarding to receive a letter from someone called Bond, thankfully not Max Bond, because that would be too spooky, spooky action or a distance. So it's good that the person is, of course, alive and called Bond. And, you know, so we have framed this because this is something we need once in a while to go on doing what we do. And with that, I would like to thank you for your attention. It was very clear, I mean, or extremely boring, right? One of the two. Hi, I'm a postdoc here. Can you please elaborate why you go to triple slit, double slit is not good. You touched up on that, but I didn't get it complete. Yeah, so maybe I'll elaborate on that. So the double slit is fine, it's beautiful. I don't say it's not good, but the point is that when I have this, when I don't have this correction term, I get an interference pattern and you know how it looks, maxima then minima and so on, drops. When I add this correction term, I would also get an interference pattern. And because this correction is coming from subdominant contributions. So then clearly, I mean, in fact, we have of course done this simulation and we find that the two interference patterns look like they're overlapping with each other because the contribution is small. So if you do a subtraction, then you will of course get a non-zero function as a function of distance and so on. So now, of course, the motivation is always to do this experiment. So if you do an experiment, you will get an interference pattern and you will have these small data points with error bars on them because of the fluctuations in photon numbers and so on. So it'll always have errors. So this error bar will actually explain both the theories because the contribution is small. So then it will explain the one with the correction and the one without. So you can never say whether you have very, what you've verified in an experiment, in the double slit case, if you want to actually do an experiment to prove the existence of this correction term. So that is why you need to go to anything more than two, three being the smallest number greater than two. And so then there you can have this correction which will make a quantity non-zero as opposed to it being zero when this correction is not there. And so then that makes it a much better distinction between the presence and absence than a non-zero one, non-zero two and a non-zero three. So that is why we go on to triple or even higher actually if necessary. Any other, please you have Moushinde in the lecture. The Q-trits is more restaurant noise than the Q-bits. I didn't get the idea. Yeah, so there's of course, it's a theoretical work which has been done by other people. So they have actually measured the decoherence effects in Q-bits and Q-trits theoretically, not for our system per se. So it's essentially a theory work which shows that they have been able to show that, you know, Q-trits because they have three levels, they have more, they have less effect due to the decoherence than Q-bits. So it's a purely theory work which we can use as a motivation for building Q-trits. But then of course we haven't verified anything in our own system. We are far from that of course right now. Other questions. I think that at some point you reach a limit. I mean if you increase the number of states that you combine that you may reach a limit where it's not advantageous but it somehow creates some problems. See, I mean, you know, I mean, one thing which I didn't really mention was that, you know, of course the advantage of using three over two is easily understood because the number of systems goes down. But then still it's not that there is a, you know, still the community is trying the Q-bits, right? But then the Q-trit is fine but as you go to let's say a 10 dimensional system or something then of course you will have errors associated with increasing this number even. So then you have to come up with a competitive mechanism whereby, you know, this error is much smaller than the error which comes when you have more Q-bits in superposition. So of course, you know, at the moment we have not reached that stage but that's why it's fascinating to explore, right? Yeah. There are no further questions or comments. Again, congratulations. Thank you very much. Thank you. Thank you very much. Thank you. Thank you very much. Thank you. Have a delicious party and future work. Thank you.