 So you put that copy over here, you take a spectrum, you get the spectrum, it doesn't look very great, but you end up seeing these peaks. It doesn't look like a peak button, just a big arrow bar. You see these peaks. So you get some spectrum with a unique shape. And then what you do is that, you keep the spectrum, now you switch from the weapon that you know is real to the one that you are as a candidate, you are not sure it's real. You take a second spectrum, and now you compare the spectrum with each other. But if the spectrum matches, and if the foil is the same, the means of weapons have to be the same. So this is like the physical analog one comparison is think of Chexons. You have a file, you run a Chexon model, and you get a hash. If you have two files, you run a Chexon model, and if the hash matches, it means the files have to be exactly the same. I mean, most of the same. But the important thing is the following. Not only you can verify that the files are the same, but from the hash, you can never reconstruct the content of the file. One way function. That's right. It's almost one way function. Of course it's one way function. There are some tricks which allow you to go back. One way function is actually... Basically, think of this as the file. Think of this as the Chexon, and this is the hash. So from this, you can do comparisons, but you can never reconstruct all the samples. Another way to think about this, this is actually the formula to explain this, this is their signal, this is the part that depends on the weapon itself, this is the part that depends on the point itself. Think of this as one equation with multiple unknowns. So remember from mathematics, from school mathematics, if you have x plus y equals 10, and your teacher asks you, what is x equal to? I think x can be anything. x can be from minus infinity to plus infinity. There's not enough information. One equation with 12 unknowns. So this is one equation with two unknowns. But importantly, the second unknown is kept fixed, which is what allows you to do these comparisons, while finding out nothing about the other... Are they? Yes, sir. A question, I guess this is a question that Becks could be asked. Has anyone ever found a fake warhead? People have never done this. This is still to be known in the future. People have never applied this to actual warheads. This is for the possibility where you have inspection regime, where Americans expect to show up in Russia, and Russians bring out warheads that are actually fake. They put some crap there, keeping the good warheads in the back. The purpose of this thing is to precisely catch this kind of... This is what I say, a hoaxing scenario, that's what they do. So how do we do our research? How do we study whether this is even possible, just because it sort of looks good as nothing is going to work? It depends on the physics, what the cross-sections are, etc. But we're not going to go into much detail. You'll be nice to listen to real warheads, but the MIT will not have warheads later on. What is the possibility to break the foil? So the foil is made by the posts, and inspectors are allowed to have visual access to the foil. But they are not allowed to look inside the foil. They are not allowed to know anything about multiple data. Almost everything, they have. But if you break the foil, if you drill a hole, then you have the access into the radiation from insects. That's right, yeah. But you are not allowed to do that. So the inspector will never be allowed to temper, or change, or take, or analyze the foil. The only thing the inspector will be allowed to do is visually be sure that nobody moved the foil between those two measurements. That's very important. So what we do is that we have done a series of simulations. In physics, it's what's great, is that if you understand the underlying physics, which we do, you can do experimentation, you can do lots of simulations. So we won't take our little simulations, we'll take the approximate design of warheads from open literature. And we're on simulations where we shoot these photos, these x-ray I talked about, and we put the foil over here, and we'll look at the signal. We try to essentially say, okay, let's take this weapon, and let's change it a little bit. How much will our signal change? Will we be able to see the difference, or not? Maybe this process is such that we will just never see the difference. Just because it exists does not mean it's going to work. Actually, we had to, one of the cool things was computational power, which we didn't have. But at this day and age, you can go to Amazon, and you can rent something like 10,000 cores, use them for an hour, pay $100, $200, and then shut them down. So something that in the past only national laboratories and major facilities could only allow themselves, now a mere mortal can go, can use, pay their $100, $200, and it is incredibly popular for doing these things. Without this, we're going to be very hard to do. So we did a few simulations of different geometries, and what we did was that we tried to take the original weapon, and we tried to change it in different parts, and ask ourselves, does our signal change enough for us to be able to see the change, and how long we have to measure to be able to see it, which is very important for operational qualification. So this is the actual simulated spectrum from a genuine template, for a real-world template. The red one is the signal from a warhead where we replaced the plutonium, which is very valuable, with uranium-238, which is, you can almost buy it on a string. And the question was that will this spectrum be different enough for us to tell the difference? So if you look, you can see there's a clear discrepancy between all these peaks. You can see that the peaks corresponding to the template are much larger than the peaks that correspond to 238. So the sharpening is that, yes, you're going to see this very easily. And we estimate that it takes something like order of 10 minutes. But then we did much more sophisticated talks rather than replacing plutonium with deflate uranium. We also did attempted replacements where we replaced plutonium with different type of plutonium. For a bomb, we did something that's called weapon-grade plutonium, which is very rich in one isotope and doesn't have the other isotope. And there's other type of plutonium which are not that way, are much easier to find. You can get them out of a reactor. So questions, can we catch those type of boxes? So we tried all these different processes, different steps. So here's like a freedom, you know, there's lots of information here, but if you just look at all these peaks over here, I mean you clearly see that for a given energy these things are not overlapping with each other, as they should if the two things were the same. And you can also do statistical tests to find out how many sigma difference there is for people who are not in the statistics. Think of that, if you see something that's a five sigma difference, it's basically you're asking the question how likely is this to be a result which will always happen. Five sigma means that it's 0.001%. So if you see a five sigma discrepancy it means that they are definitely different. So we tried different scenarios and for each one of them, again it's hard to see because of background, but we saw much higher than five sigma discrepancy. We published our paper in physical and procedural academic sciences last summer. We also published something that's called geometric boxes, which I won't talk about. It turns out that we can also do to allow them to hoax the system. But we showed that if we do simple rotations of the system and do multiple measurements we'll catch those to us. This is one of our students. This is actually now we're also doing some experimentation. This is my student Jason Bobrick from Canada. So we have a 2.5mm accelerator with a beam that's coming above and it gets bent by a small battery bank that there's a radiator. We get a photon beam. This is a uranium target right over here. And right behind this is the semiconductor detector that we're asking about. It's a high-purity germane detector. It has actually mechanical coolers that you can grab over here to operate there. We're actually taking spectra from this and we're trying to understand that we cannot see the physics that we want to be seeing. So this is some of the actual data. This is not the collection of this data, finally. So these are these three main peaks if you read into 35. And the interesting thing is that you can also see the second piece that come us almost like, I don't know, like twins. Every peak has another peak, which is 45KB off. It has to do with the fact that you can have decays from a particular excited state to the ground state or you can have a decay to the first excited state. So you'll have a emission of a photon which is carried 45KB short, 45KB less. That is also in this algorithm. So conclusion, because I think I'm running full-time, so we leave essentially a nuclear age. It's not as bad as just to be during cold war where we had 70,000 warheads, but we still have enormous number of warheads. And this is not just Russia and America's problems. This is everyone's problems. If something bad happens, everyone is going to be in huge trouble. There's enormous stockpiles of nuclear weapons in the war. There's always the risk of accidental nuclear war, which is something that is really the ultimate threat. And if this thing happens again, it's not just threat to America and Russia. It happens to us, to Azeris, to Turks, to Georgians, Persians. Nuclear winter does not tell differences in religion, at least in genetics. So we essentially need to enable much more ambitious arms reduction treaties to solve this problem. And one of the barriers towards these treaties, not only one, the main one is political, but the other one is technological. Can we come up as physicists, mathematicians, engineers? Can we come up with some kind of applications of what we do, which will make those treaties easier? We are working with one of these ideas, and we think that it has true sensitivity to this type of focusing. It does protect information, and this is something that we'll work with. That said, I'm ready for your questions. I think, is there any chance I'm sure there's probably I think there's any chance that you are pumping x-rays into a you know, some of those covers could cause some funny things to happen. So, how did I ask? People do ask this question very often. The question is that can you heat up explosives for them to benefit? Exactly. So we don't think that there's going to be enough heat. For the specific examples, we actually, for our people who need the calculation, find out that you heat it up by 0.5 degrees. So it's not appropriate. But that's general for all of these things. Yes. What are the problems you get? Like verification of authentication. Well, without photography, in real life, how do you verify? You just have a passport, see your pictures, and then you authenticate. So in photography, verification, usually in photography, we have like software and law, there's nothing just, and then your verification is like you have reference information, then you just verify if this guy knows the password, then he's verified, right? Except it's incomplete, right? Because what if someone else has a passport too? So it's like this. Well, this, yeah, photography, you cannot avoid that situation. It is possible. If somebody, it can be authenticated, you cannot do anything with that. Now, in this case, you are speaking about kind of verification about some, you have to verify, of course, by using physical methods. That's right. It has some information there some properties, so if we go back to this, what you really find, so really what you are verifying, you are verifying that this weapon and this weapon produce the same signal. That's what you actually are measuring. Okay, the same signal, it means, so if the signals are the same, it means that they have the same, same makeup, same composition. Because in the passwords, this is what you know in your mind. That's right. And this is what is physically there. So, is the anonymous password is that, let's say you enter your password, but let's say you are worried about privacy. From a password, the other side can never figure out what is your age or something you want to protect.