Physicist Reacts to Quantum Computers Explained by Kurzgesagt
what's going on guys welcome back today we're going to be checking out quantum computers so this one should be fun because you know you might have heard that i am making this cal vpn which is going to be the first commercially available quantum proof vpn ever in the world and yeah that's gonna be awesome so if you're interested in that you should come sign up in the description you might be like why is this important well you're gonna find out in this video alright so stick around and i will give you some give you the scoop on what's coming so you've probably heard about quantum computers before they've been in the news a lot over the last five years or so and you've probably heard about how google intel ibm all these companies are trying to build one a useful one that is we already have them they're just kind of crappy at the moment um i think google and ibm they're trying to build superconducting qubit quantum computers which i don't think that's going to work out very well they're having some problem with what's called error correction um we'll explain that i think intel is looking like a like a winner they're trying to build one uh with silicon essentially like uh your phones and computers they have all these semiconductors in them transistors that are made out of silicon and if we could open up uh that industry with the infrastructure already in place uh to you know then just slightly change it for quantum computing hardware well that would be a game changer and that's what intel is trying to do and that is all thanks to a bunch of breakthroughs coming out of australia so i'll tell you all about that because i've been following this for years um the last five years australia has really been leading the quantum computing breakthroughs uh so i can't wait to tell you about that in this video for most of our history human technology consisted of our brains fire and sharp sticks while fire and sharp sticks became power plants and nuclear weapons the biggest upgrade has happened to our brains since the 1960s the power of our brain machines has kept growing exponentially allowing computers to get smaller and more powerful at the same time but this process is about to meet its physical limits computer parts are approaching the size of an atom to understand why this is a problem we have to clear up some basics so computer plants have already made it to the size of the atom we actually created the first atomic scale transistor in 2012 so that is literally a transistor a semiconductor made out of atoms and just the end of last year we created the world first integrated circuit at the atomic scale so a quantum circuit uh yeah i'll tell you more about that soon so we've already done that and you can actually compare our sort of quantum computing trajectory to classical computing and it's actually pretty similar so stay tuned and i'll tell you about that actually maybe it's done now so what is it 1946 we uh created the first transistor and then 1957 we created i'm pretty sure the first integrated circuit so and then five years after that we had the first sort of commercial uh computing product which was a shitty little calculator and then you can compare that to where how we're going with quantum computing so first atomic scale transistor 2012 first um integrated circuit now the end of 2021 two years ahead of schedule um so from there you could expect we'll have some kind of commercial products for quantum computing or at least quantum simulators we'll talk about those later uh in about five years or less so be ready a computer is made up of very simple components doing very simple things representing data the means of processing it and control mechanisms computer chips contain modules which contain logic gates which contain transistors a transistor is the simplest form of a data processor in computers basically a switch that can either block or open the way for information coming through this information is made up of bits which can be set to either zero or one combinations of several bits are used to represent more complex information transitions i think it can be explained even more simply than that so if you have a transistor you can imagine it as an open gate or a closed gate so when it's when it's uh open electricity can't flow right when it's closed electricity can flow and so we're just arbitrarily defining you know an open gate where there's no electricity flowing as zero and when it's closed as one it's very arbitrary the zeros and ones don't really mean anything they what they translate to is electricity flowing no electricity flowing that's why you can only you know sort of that's why everything at the the base level machine code is zeros and ones that's why when we say zeros and ones that's what we mean uh and then in programming you can go higher and higher level to you know where you get to like python but all python is written in lower level languages like c and then down to even machine code which is you get to ones and zeros so it's pretty incredible if you go on a deep dive into how computers work and you'll have to if you want to understand how quantum computers work so now quantum computers you have instead of bits which are the ones and zeros open close circuit you have qubits and you've probably heard how they can be in like a one or zero or a zero and one at the same time right and that is kind of confusing for a lot of people and what like what does that even mean well you know there's many ways to do this one way to do it is by using the spin of atoms or electrons and one way we're going to talk about a lot in this video is using silicon qubits so you can actually as particularly with nuclei of atoms the nuclei have spin just like electrons and we'll talk about what spin is very briefly because you know that needs a whole video and it's extremely complicated but um the spin you know it can either be as we say spin up or spin down which is very arbitrary so we're not going to go any deeper into what spin is in this video but i'll just say it's an intrinsic form of angular momentum um so all particles have it and composite particles hadrons and uh nuclei but it's the analogy of something spinning it doesn't work so let's just leave it at that for now and so with the queue bits you can store information quantum information as you know ones or zeros but you can also do a combination of the two states you could do like 70 percent uh you know up 30 down or like one or zero the problem is the the the combinations of states they're very uh fragile and it's very hard to get them to stay in those uh combinations for very long so that's a major problem we'll talk about and a bit of an output in the room for you know the hardware approach that google and ibm are taking they're having a real problem with uh keeping a lot of the the quantum state stable that i think the current record is like 100 microseconds which is awful you know if you want to need to do it actually useful calculation you're going to need to be able to keep the you know the qubits in in in these superpositions for longer than that and then you can manipulate these qubits with microwaves or magnetic fields different ways of doing this are combined to create logic gates which still do very simple stuff for example an and gate sends an output of one if all of its inputs are one and an output of zero otherwise combinations of logic gates finally form meaningful modules say for adding two numbers once you can add you can also multiply and once you can multiply you can basically do anything since all basic operations are literally simpler than first grade math you can imagine a computer as a group of seven-year-olds answering really basic math questions a large enough bunch of them could compute anything from astrophysics to zelda however with parts getting tinier and tinier quantum physics are making things tricky in a nutshell a transistor is just an electric switch electricity is electrons moving from one place to another so a switch is a passage that can block so i don't know if you know but yeah in electrical circuits electrons aren't actually flowing so that's a very complicated and convoluted discussion but just keep that in mind electrons from moving in one direction today a typical scale for transistors is 14 nanometers which is about eight times less than the hiv viruses diameter and 500 times smaller than a red blood cells as transistors are shrinking to the size of only a few atoms electrons may just transfer themselves to the other side of a blocked passage via a process called quantum tunneling in the quantum realm physics works quite differently from the predictable ways we're used to and traditional computers just stop making sense we are approaching a real thing so we're going to skip that one today because that is a again a convoluted thing to talk about for our technological progress to solve this problem scientists are trying to use these unusual quantum properties to their advantage by building quantum computers in normal computers bits are the smallest units of information quantum computers use qubits which can also be set to one of two values a cubic can be any two level quantum system such as a spin in a magnetic field or a single photon zero and one of this system's possible states like the photons horizontal or vertical polarization in the quantum world the qubit doesn't have to be in just one of those it can be in any proportions of both states at once this is called superposition but as soon as you test its value say by sending the photon through a filter it has to decide to be either vertically or horizontally polarized so as long as it's unobserved the qubit is in a superposition of probabilities for zero and one and you can't predict which it will be but the instance you measure it it collapses into one of the definite states he's doing a really great job with this is a game changer so all the hype around quantum computers probably started way back in 1994 that's when it really started to ramp up because a guy called peter shaw from mit basically showed that you could factorize a really large number really quickly with a quantum computer and the way it does this is by basically representing the factorizations of uh the number as like quantum waves that can like slosh around simultaneously through the qubits because of their correct ones just pop out so you could break encryption of like uh internet communication with this quite easily as well which is scary and a bunch of other things so it is you know this is why there's so much hype around it so the encryption systems that basically secure the internet currently uh basically rely on uh the fact that classical computers like the one in front of you probably they just can't handle factorizing really large numbers that just they just get overwhelmed very easily and that's why current encryption you know is pretty safe from you know normal hacking there's a big problem though because shaw basically assumed that um you know these qubit these quantum waves would be able to slosh around the qubits for as long as they want uh so you know it can calculate things that take longer even on the common computer but unfortunately that's not the case with uh the hardware google and ibm are building unfortunately they can't slush around for very long at all like a fraction of a second you're talking like 100 microseconds on google's one so google and ibm specifically use qubits made of tiny resonating circuits of superconducting metal etched onto like these little microchips and they did that because at the time when they started constructing them they were the best that was the easiest way to uh manipulate the qubits and to create these chips but that's all changed there's been a paradigm shift as of late with huge breakthroughs coming out of australia of all places and uh we can now do this with silicon and at the atomic scale like literally atom so instead of using um you know metal etched onto you know these microchips or like cavities and you have to manipulate them with microwaves or even magnetism you can now do this in a much more simple way you can actually use something called electric nuclear resonance so we'll talk about that soon because that discovery is just mind-blowing and it really changes everything um but there's also some other huge advantages of using these new breakthroughs to do it on silicon and so there's a there's a couple teams in australia which are kind of competing um one is called silicon quantum computing and their research is out of university of sydney there that's a really interesting paradigm as well i'll talk about that but my the one i think that's going to win this race at least the hardware that we're going to use to get there uh they are basically using uh they they put nuclei of atoms into um they entangle them with the electrons of atoms and you can do this with phosphorus and um then you take that electron to another atom and you can entangle that with uh another nucleus and they recently showed like a really recent nature paper like i think a couple weeks ago or a month um that you can get error fidelities like uh of up to like 99.95 so that's that's incredible basically error-free free and the error correction they can do now basically uh finds the errors quicker than they arise that is a game changer because google and ibm currently they're just plagued with this error correction problem and it seems like they're kind of struggling so it's really looking like silicon electric uh nuclear resonance is probably going to be the way to go um and then also out of australia michelle simons who runs that silicon quantum computing she uh literally as well like a month ago just um created the first ever uh atomic scale quantum integrated circuit so that which is just mind-blowing um she was also the first her team in 2012 to create the first atomic scale transistor made out of atoms and so this integrated circuit um they basically created a quantum simulator which is different to a quantum computer because you can't reprogram it and so they modeled um what was the molecule i think it was carbon and so it basically emulated what carbon does but you can't reprogram it unfortunately that's the next stage that silicon quantum computing the company they're aiming to have a useful quantum computer by i think 2028 so approximately five years um and i think they can do it and they have a really interesting lab which took i think like 20 years to make it's like this building within a building that's kind of like floating uh to you know to get rid of all the noise and you know movement of the building because as they build these chips made out of silicon atom by atom you know you can imagine that's a very delicate process uh it's just incredible so that silicon quantum computing they're going to be using uh quantum dots i believe uh where you literally are going to be using the qubits will literally be atoms uh they'll just be you know electrons so it's just mind-blowing the sort of fidelity we're approaching and i really do think uh useful quantum computers are going to be built with silicon because once you get less than one percent error rates with the tech uh that basically opens up the semiconductor industry which is you know there's a lot of infrastructure for it already um and then we can easily you know transition to making well not easily it'll it's very hard but to make our chips for quantum computers four classical bits can be in one of two to the power of four different configurations at a time that's 16 possible combinations out of which you can use just one four qubits in superposition however can be in all of those 16 combinations at once this number grows exponentially with each extra qubit 20 of them can already store a million values in parallel a really weird and unintuitive property qubits can have is entanglement a close connection that makes each of the qubits react to a change in the other state instantaneously no matter how far they are apart this means that when measuring just one entangled qubit you can directly deduce properties of its partners without having to look qubit manipulation is a mind bender as well a normal logic gate gets a simple set of inputs and produces one definite output a quantum gate manipulates an input of superpositions rotates probabilities and produces another superposition as its output so a quantum computer sets up some qubits applies quantum gates to entangle them and manipulate probabilities and finally measures the outcome collapsing superpositions to an actual sequence of zeros and ones what this means is that you get the entire lot of calculations that are possible with your setup all done at the same time oh ultimate no they're not done at the same time quantum computers don't calculate everything at once that's a little bit incorrect i just want to emphasize uh you guys should be very very skeptical of what you read about quantum computers on the internet because so much of it is just absolutely wrong for instance like you see so many places claiming that quantum computers calculate a bunch of things at the same time no they do not you'll see that you know with just a small number of qubits will have more you'll be able to store more data than atoms in the universe that's not true um you'll need quite a few for that and uh well like basically if it doesn't involve transforming an np problem into a p problem uh i would be a little bit skeptical and that you know there's only limited cases of that as well you know sometimes it works theoretically so you know just be skeptical of what you read about quantum computers a lot of people in software and computer science talk about it but unfortunately they don't know the physics behind it and that that is a bit of a bottleneck you can only measure one of the results and it will only probably be the one you want so you may have to double check and try again but by cleverly exploiting superposition and entanglement this can be exponentially more efficient than would ever be possible on a normal computer so while quantum computers will probably not replace our home computers in some areas they are vastly superior one of them is database searching to find something in a database a normal computer may have to test every single one of its entries quantum algorithms need only the square root of that time which for large databases is a huge difference the most famous use of quantum computers is running i.t security right now your browsing email and banking data is being kept secure by an encryption system in which you give everyone a public key to encode messages only you can decode the problem is that this public key can actually be used to calculate your secret private key luckily doing the necessary math on any normal computer would literally take years of trial and error you're talking millions to billions of years by the way some computer with exponential speed up could do it in a breeze another really exciting that's why uh yeah rsa ecc elliptic curve encryption uh and bitcoin they're all at risk so if you're into cryptocurrencies bitcoin all that sort of jazz it's probably time to start worrying because a lot of people don't realize that ah powerful quantum computers will actually be able to kind of ruin cryptocurrencies in general so it's actually pretty straightforward as to why cryptocurrencies are kind of at risk if these powerful quantum computers already exist and for when they're coming in the future and if you know hackers or governments just want to steal data currently and you know they can retroactively break the encryption and that's because every time you make a transaction with a cryptocurrency you're exposing your public key right and there's nothing wrong with that like everyone knows this but if you have a powerful enough kind of computer you will potentially be able to uh with classical computers it will take billions of years millions to billions um but with the public key you'll actually be able to probably derive the private key so if you have the private key well then you can do whatever you want you can make transactions and so you can kind of see quantum computers are a real problem for cryptocurrencies in general um so a lot of people don't seem to think you know we need to worry at all but it's like why wouldn't we start preparing for that if it's kind of coming you know what i mean and again it might already be here so let's quickly talk about why these crappy quantum computers haven't been out of basically break um public private key cryptography yeah so that's basically because we haven't made quantum circuits big enough yet it's coming though um and so i think a good source to i think microsoft published a paper which said you know about 4 000 um cubits a circuit will be able to break um rsa encryption and then to break elliptic curve uh encryption i think it was like 2500 qubits which we're not there yet you know that what's the record like you know within the just over 100 qubits and you've got to also remember that microsoft paper that was accounting for you know ideal circumstance in physics was always ideal idealists and so these cubits are like in perfect uh harmony and it's just not reality uh you'd need a lot more pubits than that to really do it in reality um and that's we can do that when you have enough queue bits you can account for the the realistic problems via error correction and this happens in normal computers as well and classical error correction is actually pretty easy because you can basically just copy the bits um so if you know if there's an error you you have a copy quantum computers you can't copy the qubits um because there's something called the no cloning theorem and so the way you do it is by basically having a lot more qubits traditionally with this quantum error correction but now with what this new research out of australia is showing is uh you can actually get less than one percent uh error rates with entangled qubits if you use uh the the paradigm that these researchers using my prediction is we're gonna have useful quantum computers within the next five years that's why i've created cal vpn i think it's very important because as i keep hopping on about i'll say it one more time you know there could be hackers around the world particularly governments china could be stealing data very sensitive data your data uh to you know with the intention of decrypting it hacking it when these computers exist because i'll explain why they're going to be able to hack them using something called shaw's algorithm so we already know it's possible so why wouldn't they be doing this i'm pretty techno-optimistic uh in general and especially with quantum computers from knowing the stuff that i know with the research that i've read uh and i really do think that we might be able to break rsa encryption within the next five years and this i don't want to you know put my tinfoil hat on too much and worry you guys but governments really could have like they could be well ahead in the quantum computing race and they could actually have very sophisticated quantum computers already and now if you're saying hang on a second i'd like you're going you're getting a bit crazy here well keep in mind that governments have been ahead uh with technology before take for instance the internet darpa a lot of the infrastructure was developed in secret by darpa um so that's the internet you know quantum computers is probably pretty comparable to that you know paradigm shift of when the internet and the world wide web became a thing so i would not be surprised if uh america china does have very secret you know pro black technology um and if they do well obviously they're not going to tell anyone you know they're going to secretly spy on each other i think this is actually quite likely and i think the time to start worrying is right now and i think if you go listen to experts talk about this they say the same stuff you know so i don't want you to think i'm going off on the the wild end of speculation this is a the time to worry is now and so this is where cal vpn comes in it will protect you guys because there is um so let's talk about it okay so what the the way it protects you is it's like a normal vpn right but it has these post quantum cryptographic algorithms in it so important thing to note here is uh post quantum cryptography is very different to quantum cryptography all right they sound nearly the same so post condom cryptography just mathematical problems there's no special hardware you know testing this stuff that's what cal vpn utilizes whereas quantum cryptography it uses the properties of quantum mechanics to uh basically you know prove that you can't uh hack it so you use these quantum properties to create a shared key between two parties and you can basically provably show that you know they could they no one saw the key because if they did the the communications data would have been destroyed by way of quantum mechanics you know you've probably heard about uh superposition and stuff you know just observing the wave function collapses it stuff like that okay um so again kelvin relies on post-quantum cryptography mathematical equations which uh they should be resistant to all known attacks even if the governments have these powerful quantum computers in their hands they won't be able to get their get their hands on your data if you use cal vpn so post quantum cryptographic algorithms are just classical algorithms that can basically withstand uh powerful quantum computers and it doesn't rely on any like hardware any complicated quantum hardware uh no quantum properties it's just classical stuff so it really does work and it can also protect you from you know just powerful computers and super computers if you know what i'm talking about you might also wonder why we don't just double the key size well that has to do with the fact how quantum computers work they don't just brute force their way through problems no no so it's that doesn't really do anything but these algorithms that will protect you from quantum computers actually come directly from the submissions to nist which is the us uh national institute of standards and technology and they basically very recently within the last month have announced uh four uh algorithms which will set the stage for uh standard encryption for post quantum encryption that is um and there's still four other algorithms that they're still looking at to see if they'll work but this has been what's what is it like six year a six year long competition uh to find some algorithms that will be able to protect against quantum computers and so basically our software includes these cryptographic algorithms that will protect you and encrypt your data to a degree where you know even if quantum computers are stealing data right now they will never be able to decrypt them so if you're interested in the specific post quantum algorithms and you want to check them out yourselves go look up crystals kyber uh crystals delithium uh falcon and sphinx they're the ones that have been currently selected by nist to to work essentially uh and it's awesome we might talk briefly about them in the dedicated videos that i make they're just quite complicated the math is you know not straightforward and the terminology as well there's a lot to explain so nist have really been trying to promote like a sense of urgency to upgrade encryption standards in everything on the internet uh all software to promote a sense of urgency i remember reading that they cited a uh a study by the quantum economic development consortium which basically found that uh it took a large a large tech enterprise like five years to transition to advanced encryption standards so you know if you're currently using a vpn like uh you know nordvpn surfshark all of those big ones which i've worked with in the past you know in sponsorships on this channel well that's the sort of big tech enterprise which will take a long time to upgrade their encryption standards why because it's so bloody complicated and they probably don't know how you know you they have to hire people to do this and it basically involves changing their software like pretty substantially uh so that's a big problem so if you're already using a vpn i highly recommend you know looking into this further for yourselves and deciding if you want to jump on the kel vpn bandwagon because it's going to take them a long time like even nordvpn it's going to take them a long time to bring in these post quantum cryptographic uh encryption standards i would almost bet you know any sum of money it's going to take them a long time and we can already offer protection against them and like i keep saying you know it's no secret that america has developed black technology in the past and they've been ahead in with technology than say universities so it is something we need to worry about now and it's particularly if you're you know interested in cryptocurrencies if you've got bitcoin you should definitely consider this is simulations simulations of the quantum world are very intense on resources and even for bigger structures such as molecules they often lack accuracy so why not simulate quantum physics with actual quantum physics quantum simulations could provide new insights on proteins that might revolutionize medicine right now we don't know if quantum computers will be just a very specialized tool or a big revolution for humanity we have no idea where the limits of technology are and there's only one i think there's a lot of things in principle which we actually do kind of know what quantum computers will be able to do um and it that those things will revolutionize humanity we really have no idea to the extent at which quantum computers could revolutionize everything like society itself and to give you a little bit of an insight into this like uh when after 1994 when uh we discovered you know these quantum algorithms we basically uh realized that in principle quantum computers would be able to do things that would just be like impossible on classical computers and it seemed like researchers at the time got really excited by that uh and they expected to find a lot of different quantum algorithms that would be able to you know do incredible things and people have described it as a bit of a bummer trajectory because like i said people were expecting to find a huge range of things you could do with these quantum algorithms but nope we basically realized that quantum computers were only going to offer like a really big speed up to a very narrow single class of problems uh within a standard set called np if you've never heard of np problems you should go look into those but they're basically ones that have efficiently verifiable solutions so like factoring right uh and so that was that's been the case for like three decades bit of a bummer but very recently we've we've found a uh an entirely new problem that quantum computers are going to be able to solve exponentially faster than classical computers so that's really exciting but we're not sure if it's like a standalone problem or if it's part of a new whole sort of you know set of problems we just don't know uh it basically involves calculating like really complex inputs of a mathematical process uh from only its jumbled outputs so it's just a completely new new problem so that's very exciting
2022-07-28 20:29
