What makes quantum computers so powerful? What kinds of problems can they solve? How do we actually build one? I discuss these questions and more in today’s Ask a Spaceman!
Support the show: http://www.patreon.com/pmsutter
All episodes: http://www.AskASpaceman.com
Follow on Twitter: http://www.twitter.com/PaulMattSutter
Like on Facebook: http://www.facebook.com/PaulMattSutter
Watch on YouTube: http://www.youtube.com/PaulMSutter
Read a book: http://www.pmsutter/book
Go on an adventure: http://www.AstroTours.co
Keep those questions about space, science, astronomy, astrophysics, physics, and cosmology coming to #AskASpaceman for COMPLETE KNOWLEDGE OF TIME AND SPACE!
Big thanks to my top Patreon supporters this month: Justin G, Matthew K, Chris L, Barbara K, Duncan M, Corey D, Justin Z, Neuterdude, Nate H, Andrew F, Naila, Aaron S, Scott M, Rob H, Lowell, David B, Frank T, Tim R, Alex P, Tom Van S, Mark R, Alan B, Craig B, Richard K, Steve P, Dave L, Chuck C, Stephen M, Maureen R, Stace J, Neil P, lothian53 , COTFM, Stephen S, Ken L, Debra S, Alberto M, Matt C, Ron S, Joe R, Jeremy K, David P, Norm Z, Ulfert B, Robert B, Fr. Bruce W, Catherine R, Nicolai B, Sean M, Edward K, Callan R, Darren W, JJ_Holy, Tracy F, Tom, Sarah K, Bill H, Steven S, Jens O, Ryan L, Ella F, Richard S, Sam R, Thomas K, James C, Jorg D, R Larche, Syamkumar M, John S, Fred S, Homer V, Mark D, Brianna V, Colin B, Bruce A, Steven M, Brent B, Bill E, Jim L, Tim Z, Thomas W, Linda C, Joshua, David W, Aissa F, Tom G, Marc H, Avery P, Scott M, Michael G, Katelyn, Thomas H, and Farshad A!
Music by Jason Grady and Nick Bain. Thanks to Cathy Rinella for editing.
Hosted by Paul M. Sutter, astrophysicist and the one and only Agent to the Stars (http://www.pmsutter.com).
All Episodes | Support | iTunes | Spotify | YouTube
EPISODE TRANSCRIPTION (AUTO-GENERATED)
Let's say you've got a mouse running around your house and it's so annoying, you can hear it skittering in the night. There are little holes in your food. You're pretty sure you didn't eat all of the cheese. It's it's it's making your life a living nightmare. You want to catch that mouse so you go out and buy a cat, and it's a normal, everyday, perfectly functional classical cat. And you take the cat home and you wait. How does the cat find the mouse? Well, it searches room to room, just one room at a time. It goes in the living room, looks around, smells around. Does whatever cats do, doesn't find the mouse, goes to the bathroom. Looks around. No mouse goes to the kitchen. Looks around. No mouse. Just this very procedural approach to finding mice. It's a very slow process. You have no idea where the mouse actually is, and neither does the cat. So it has to search room after room one at a time.
And so you're a patient. You wait a night. No, mouse, You wait another night. No, mouse. You wait like a week, and there's still no mouse. So this is taking too long. This cat is just way too slow. I'm not even sure what it's doing when I'm asleep. So you buy another classical cat, you're gonna You're gonna speed this up, right? If if you have two cats, then they can run in parallel. One cat is searching one room. Another cat is searching another room, and and overall, they should. One of them should be able to find the mouse at any point. And it should take only half as much time because you have twice as many cats. But you wait a night and another night and like, another week and there is still no mouse. Did I mention that you, like, live in a mansion and there are, like, 10,000 rooms in your house to make this analogy worthwhile? Anyway, you have this giant house. There are, like 10,000 rooms. This is taking forever. Two mice are gonna cut it. So you buy eight more cats. Now you I'm laughing at my own metaphor.
Sorry. Now you have 10 cats, all right? And you're gonna run them in parallel and each one is assigned to 1000 rooms to find this one mouse and oh, it works. Boom! In one night, one of them caught the mouse. The search is done, and you're wondering if any of your neighbors would like a free cat or nine of them. This strategy of buying more cats is how the world's most powerful supercomputers work. Uh, if if you run into a problem that a computer is taking, like, way too long to solve, you just buy another computer and you put it next to it and you let them run in parallel and each one is assigned to a different part of the problem, which is fine. I mean, we get a lot of work done that way. But what if your house is really big and I? I mean, I mean really big. Not 100 rooms, not 10,000 rooms. But what if it was? What if your house that had, like, a quadrillion quadrillion rooms, how many computers would you need to run in parallel in order to search for that one dang mouse in your house with a quadrillion quadrillion rooms, you might be saying, Well, Paul, like, do problems really get that big?
Yes, problems in the real life really do get that big to give you some example of of just how nuts this can get. Let's say I'm, you know, analyzing, studying, simulating some physical system. And I'm studying the physical system and I need to to chop it up into little bits and in each one little bit is is the part I'm gonna study if I'm just looking at, like a one dimensional system like a a string of atoms in a line, you know, and I and I divide up that line, you know that that's not too bad. I can I can have, like 10,000 different parts of the line. I get very, very high resolution, and I can I can really study it, and I just need, like, 10,000 bits of memory or whatever. But if I wanna study a atom, say, on a two dimensional surface and I want that same resolution, I want that same fine grained analysis of it. Well, I need 10,000 on one side and 10,000 on the other. That's squared. That's 10,000 squared. That's a large number.
And what if I want to study like three dimensional volume like I wanna study atoms floating around in a box, and I want that same resolution. Well, that's 10,000 times 10,000 times 10,000. If I want to double the resolution, if I want to go be able to study this or simulate this or compute this at an even higher resolution at an even higher like like a smaller scale, then every time I double the resolution, my data needs increased by a factor of 16, and so it very quickly goes nuts and you can actually study. All this happens all the time in science. You wanna study very high dimensional problems like Like let's say you want to keep track of ocean currents moving through a patch of the ocean. You want to study this, you want to simulate it. You wanna see how the flows go back and forth like you really, really want to model this on a computer? Well, in order to track those ocean currents, you need to keep track of the three spatial dimensions, you know, like the water that is actually in each individual dimension.
And you need to keep track of the three different directions of the velocities. Like if this water is moving, You need to keep track of its velocity and you want to keep track of the temperature and you want to keep track of the pressure. And if we just stopped there, that's eight things that you want to keep track of. If you take this big chunk of ocean that you're going to simulate it and you divide it into a bunch of little, little tiny cells and you're gonna use these cells, these little sections of it to perform your calculation, each cell needs to keep track of eight numbers like the the amount of water in it, the dimension of the water, the velocity of the water, the temperature of the water, the pressure of the water. So every time you double that resolution, every time you try to go down another level in terms of fine grained study of this, you have to increase your memory by a factor of 256. In computing computational science, this is called the Curse of dimensionality, where a lot of physical systems that we want to study very, very, very quickly get exceedingly large.
These kinds of problems crop up all the time these just giant problems where we're trying to understand a physical system and all of a sudden we have to keep track of so many numbers and perform so many calculations that just it goes haywire. And our current solution to these giant problems is to buy more cats. When we run into situations that get too large, we either build a giant supercomputer, which a supercomputer is just a whole bunch of normal computers all lined up next to each other, all tackling different parts of the problem, or we just not bother studying the problem at all. Where there there are so many problems in physics, in chemistry and biology and all these all these areas of science. We're just like, you know what? I would love to study that, but I can't figure out a way to do it with our computing resources. So I guess I don't know the answer. Surely there's a smarter way, So let's go back to our original mouse hunting problem. You got a mouse in your house in your house has, like 10,000 rooms to make this metaphor worth it.
But instead of buying a classical cat, you go to a different store. You go to Schrodinger's house of Kittens and you buy a quantum cat. Now, this quantum cat is very, very strange. Here's how it works. You take your quantum cat home, you place it in a room. It doesn't matter which one. It doesn't move. It doesn't eat. It doesn't blink. It doesn't even use the litter box. It just sits there. You try poking at it. You try giving it a treat. You you know, you try talking to it. It's just not responding. Just sits there. So you go to bed and you're considering returning your quantum cat in the morning. In the middle of the night, you hear a noise, you wake up, you put on your slippers and you walk through the house and you find the cat in your bathroom, which is weird. That's not where you put the cat. You put the cat in the kitchen because that's where you thought the mouse might be. But now the cat is in the bathroom, which is very weird, and you also find it in the hallway and you find it in your bedroom and you find it in the living room and in the kitchen, in the dining room and on the porch and on the roof.
The quantum cat is simultaneously everywhere. It's in every room. There's a copy of that cat. Everywhere in the house, you are convinced you're having a bad dream. So you go back to bed. You wake up in the morning. The quantum cat you find is not in the kitchen, not in the bathroom. It's in the hallway, and that's the only place it is. It is. It's no longer a all over the house. It's just in one place, not the place you put it. You put it in the kitchen. When you went to bed, you woke up in the middle of the night. The cat was everywhere. And then you wake up in the morning, the cat is in the hallway and there's a dead mouse at its feet. Welcome to quantum computing. But to dig deeper into quantum computing, we need to talk a little bit about, you know, like regular computing, because we're we're going to import so much jargon and understanding and concepts from the world of regular computing. Uh, that we need to talk about if we really want to talk about quantum computing, and this show is not called. Ask a computer man, so I won't get into too many details.
But in case you didn't know, your computer is just a giant miniature calculator or miniature giant calculator. It's a calculator. That's all it does. That's literally all it does in the world of computers. Everything and I mean everything gets translated into numbers, but not the familiar ones like eight and Pi. Just two numbers. Zero in one. Everything that you do on a computer, like typing in your word processor or playing a video game or listening to this podcast or watching cat videos on YouTube gets translated into exceedingly long strings of ones and zeros. And it's by doing simple math on those ones and zeros that computers well, you know, compute. And they do all the things that you want them to do. Modern computers are generic. They're not hardwired to do one thing and one thing only. But to solve a variety of math problems, like recording podcasts and watching cat videos. And yes, everything you do gets translated ultimately into math problems on those ones and zeros to do this Your computer has a bunch of different parts as memory, where it stores all the ones and zeros.
And it has a processor which is chock full of something called logic gates that that do interesting things to those ones and zeros like ones and zeros come in. They get shoved together in very interesting ways. They might get compared. They might get canceled out. They might get reversed. Uh, they make new combinations. For example, one kind of logic gate is called an and gate. And it's where, uh, if two zeros come in to the Logic Gate, then only 10 comes out. If a zero in one come in, then also a zero comes out. If a one and a zero come in, it's also a zero. And if a one and a one come in, then a one comes out. It's like it's just like interesting combinations of ones and zeros to tie all this together. You as a human, write some computer program, which we call an algorithm, and I'll be using that word a lot. An algorithm is a detailed list of tasks for what you want the computer to do. So you say like, Hey, Computer, I want you to fill up this part of the memory with a certain pattern of ones and zeros. These this pattern of ones and zeros represents, you know, some math problem trying to solve or some spreadsheet or some some cat hair on the video.
You know, just it represents that. And then I want you to send these ones and zeros through these kinds of logic gates, uh, to get this particular result. And then I want you to put the that answer back in the memory, uh, and and also tell the monitor to display the cat video that's basically a computer I. I know there are probably a lot of computer scientists and engineers in the audience. You'll forgive me for providing an exceedingly high level description of how this works, these ones and zeros that represent like everything that you do on a computer. They have a very specific name. They're called bits. They are the foundation of everything that your computer does. The memory on your computer can store a certain number of bits, and they do this by having little teeny tiny places that either can be empty or filled with electricity representing a zero or a one, and the logic gates operate on those bits and spit out new bits bit by bit. Ha, ha! Your computer gets work done, you know. Simple, clean, efficient, powerful.
You know, it's kind of a revolution in human technology, but like everything else, it has limits. The number of bits that your computer can handle limits the sizes of problem that it can crunch. For example, if you have a chunk of memory that is supposed to represent something, whether it's a number or a letter or a cat. If that chunk of memory is just one bit in size, it's pretty limited to what it can represent. It can represent a zero, or it can represent a one. It can represent two different things. That's it. If you have a chunk of memory that is two bits wide, you can represent four different things because you can have 001001 or 11. That's four different things, so you can represent, like four different numbers or four different letters, and that's it. If you have 10 bits, you can represent up to 1024 different things. If you have 100 bits you can represent, you know a much larger number than that. But the chunk of memory that you're using to represent real world concepts is limited by the size of that chunk, and these chunks can only represent one number at a time.
And if I have a if I have a chunk of memory that is 10 bits wide, I can represent any number between zero and 1024. But I can only represent one of those numbers, like these bits some of them will be on. Some of them will be off. That is one particular state of the memory, and that represents all thing. And so to represent more than one thing at a time, which we kind of want computers to do. We have very, very, very large memories. They're divided into lots and lots and lots of little chunks, so we have a bunch of chunks of memory. Each chunk of memory can represent one value in the width of the chunk of the memory, determines how many different values I can pick from, but ultimately that chunk has to pick one kind of value to represent. The same goes for the logic Gates. If I only have one logic gate or like three, yes, technically, I can get all sorts of computations done, but it's gonna take a long time.
So the more logic gates I have, the more operations I can do. The more I can get work done, the faster I can churn through the bits that are in memory and combine them with my algorithm to get interesting answers and store them back in memory. So we like processors with lots of logic gates, and we like memory with lots of bits. But what we find when we're trying to solve these exceedingly large problems is that you run out of memory because you can't hold all the numbers that you need to at the same time. And or your computer runs really slowly because you can only have so many logic gates available to do all the number crunching. And you have to wait for the computation to run through all the possibilities. So traditional computers, as powerful as they are, do have limitations. But quantum computers are, well, different. A quantum computer is computer that takes advantage of the weird and wonderful and wacky world of quantum mechanics, And I find it increasingly hilarious that in the meandering journey through space and time, that is this podcast that we've never dug into the fundamentals of quantum mechanics and yet continued to do show after show on the more advanced stuff.
It's like we skipped Quantum 101 and went right to Super Advanced Grad school quantum topics. I know I'm going to do a series on Quantum Mechanics and lay it out. I just don't know when, which is probably keeping in the spirit of quantum mechanics. But anyway, there are two properties of subatomic systems that make them especially interesting for solving some seriously tough computational problems. And these two properties are superposition and entanglement. Now you you'll remember Quantum mechanics doesn't apply. Like up here in the macroscopic world. The the rules that we figured out of quantum mechanics simply don't apply up here. This is a different world. This is the classical world up here. But when you zoom down into the level of atoms and subatomic particles, they simply play by different rules. These rules do not make sense. These rules are hard to describe these rules are nuts, and yet they are the rules Nonetheless. One of those rules is this idea of superposition.
Superposition is what allows. Or I should say, it's the fact that subatomic systems, subatomic particles, subatomic states, subatomic setups, subatomic, whatever don't just live in one state at a time. Instead, they exist in all possible states at once simultaneously until you make an observation. This is going to be absolutely critical for quantum computers because we're going to transform a bit into a cubit. A quantum bit a bit is binary. It's either zero or one. You can pick either one, but you can only pick one at a time. And that's just what it's gonna be if that bit is one. It's just going to be one. Every time I look at it, every time I check every time it interacts with another bed, it's just gonna be one. And same for zero. But a cubit is different. It's something new. There are various ways to describe a cubit to describe this superposition, you know, technically fundamentally, quantum states are just different than classical states.
It's a new beast. It's a new creature. It has its own rules and wins and preferences for online cat videos. It is absolutely nothing like a normal classical state. It's its own thing. But for the discussion on quantum computers, in this case, I'm going to bend a little and not just stick to the math. And I'm going to describe a cubit as like a normal computer bit. But cooler a cubit contains both the zero in one state at the same time simultaneous. It exists in both states, which means that when you do math on it, like sending the cubit through a quantum logic gate, you perform the math on both the zero and the one simultaneously. The transformation of bit to cubit is like having a super duper giant supercomputer on your desk. If you have a problem that requires a lot of work or requires a lot of memory, instead of building lots of computers to do the work in parallel, you just have a single quantum computer that solves all of your problems simultaneously.
For example, one bit one bit can be a zero or a one. You get one choice of values. One cubit is a zero and a one simultaneously encodes both values in its superposition state, which means you get twice the information out of it or you can store twice as much information. You know, if if I want to replicate this with bits, I would need two bits. I need one over here to represent the zero and 11 over here to represent the one. But with the cubit I can have both at the same time stored in the same cubit. If I have two cubits, I can store four states simultaneously. I can store the 001001 and 11 all four combinations. I can store them all simultaneously. If I want to do that with traditional bits, I would need four pairs of bits, one pair of bits to represent the 00 another pair of bits to represent 10, another pair of bits to represent the 01 and then another pair of bits to represent the 11 with just two cubits, two quantum bits.
I can represent all four states simultaneously. With four cubits, I can represent 16 states simultaneously with 10 cubits, only 10 cubits. I can represent 1024 states simultaneously with 100 cubits. This number is so large, I can't even, like translate it. So I'm just going to read off the digits. Like if you have 100 cubits, The number of states that you can represent simultaneously is equal to 126765060022822940149670320! 5376. However many trillions trillion trillions, that is, with just 100 cubits, I can represent all of those states simultaneously. Imagine how much memory it would take to be able to simultaneously represent all those possible states with traditional bits. And that's why quantum computers are so intriguing. Because of this property of superposition, they can simultaneously find all possible solutions.
You don't need to wait. You don't need to divide up your memory. You don't need to take turns. You just shove it all into the quantum computer, and it does all the work at once. Quantum computers can work out absolutely ginormous problems with only a few dozen cubits, which is cool. But that doesn't get you the answer, right. This this sounds really awesome, like wow, I can just store so much information on cubits instead of bits, and I can do math in parallel, right? I send one cubit through a quantum logic gate. I'm operating on both its zero state and its one state simultaneously. That's cool. If I put 10,000 possibilities into a quantum computer, it's going to run 10,000 calculations at the same time. That's nice. Really nice, but I don't want 10,000 answers. I want one answer. I want the right answer.
I don't want I'm not interested in all the wrong answers. I'm interested in the correct one. The real trick of quantum computing is to get the machine to re only return the answer you want. Not the 9999 answers. You don't as an example with our quantum cat, you don't care where the mouse isn't. You want to know where the mouse is, and that's where the second property of quantum mechanics comes in. And that's entanglement. Entanglement means that the cubits can talk to each other, work with each other. It means that they're not separate. Quantum states each live in their own independent lives. They're all sharing a common state. You have one state or set of states that encompasses all the cubits on your quantum computer. At the same time, if you manipulate one cubit because of entanglement, you end up manipulating the rest.
For example, with our quantum cat, if it's existing in all rooms simultaneously, if you go to poke at one of those cats, the other cats will feel it. And so the game of quantum computing looks like this. You set up your problem. You use superposition to represent all possible solutions to your problem. Using a bunch of cubits, you use special quantum logic gates and more on what makes them quantum in a little bit. To navigate these cubits very carefully, the cubits travel through the quantum logic gates. Math gets done, but you do it in a very special way so that the entanglement forces all the wrong and useless answers to go away. And the only thing that pops out at the end of the computation is the correct or useful answer. It sounds like lunacy, but it's not the only thing in the universe that's crazy. For example, there's patreon patreon dot com slash PM Sutter, where Some people are crazy enough to support this show, and I love them for it. That's patreon dot com slash PM Sutter To keep this going, I really do appreciate it, but this is what you get in the world of quantum mechanics, which is pure lunacy, like nothing makes sense.
Like everything I've just said, like, sounds like nonsense, and it's probably very confusing. But welcome to quantum mechanics. It just is. There aren't that many known quantum algorithms. In other words, we only have so many examples of being able to set up this kind of superposition and entanglement and quantum logic gates to make the algorithm work and make the algorithm worth it. But where it works, it works. Quantum computing was one of those curiosities of theoretical physics for over a decade until 1994 when mathematician Peter Shore figured out a very, very, very, very, very, very, very, very, very, very useful quantum algorithm. I won't get into the gory details of that quantum algorithm, but the algorithm has to do with factorizing very, very large numbers, like if you give a computer a number and it has to figure out what numbers multiply together to get that number. It turns out this is a very difficult problem for classical computers because essentially they have to search through almost all possible combinations like, Hey, computer.
Uh, what are the factors of 20? And it's like, OK, give me a second. Well, let me see One times. Two no one times three No. One times four No. One times five. No. And then eventually get like it'll get to four times five. Yes, four times six No. Four times seven. No, I'm watering it down here, but it it essentially takes forever. But quantum algorithms do this so much better because they are able to process all possible combinations at the same time. So instead of going, OK, folks, let's try one times. One no. One times two. No. One times three and checking loop after loop after loop after loop after loop. Instead, it sends all possibilities all possible combinations through at the same time, and then it uses the design of the algorithm, uses entanglement to pop out the correct answer. Because no matter what you don't get to see, we don't get to see superposition.
We know that superposition happens because it's our only way of describing the behavior of subatomic particles. But we don't get to observe these cubits existing in all possible states. When we open it up and look at it. Then the superposition falls apart, the entanglement falls apart and everything collapses to use a choice of words into a particular state. So even though these cubits can represent and do represent all possible values simultaneously, when we go to actually finish the calculation and say, Hey, computer, what's the answer? They will collapse and they will look like normal bits. And so what we want our quantum algorithms to do is that when we go to actually ask that question Hey, computer, what is the answer? Then the quantum bits collapse into bits. They choose one or zero, and they choose it in just the right way so that the correct answer pops out. It sounds like it shouldn't work, but it does.
It sounds like alchemy. It's like quantum alchemy, but it works. It works because math, because we understand how superposition works, we understand how entanglement works, and these algorithms are able to take advantage of those properties. Why is factoring large numbers useful? Why did people pay attention in 1998 4? Because systems of encryption like, say, making the connection to your bank secure rely on the fact that factoring large numbers is super hard. I won't get into the details, so you'll just have to trust me. So all if all of a sudden you could snap your quantum fingers and pop out a factorization of a very large number, a bunch of our encryption techniques would vanish. As you can imagine, this, you know, made all people a lot more interested in quantum computers. They're like, 00, not just for the nerds anymore. I see this quantum computing thing almost seems too good to be true. It's It's like a magical computer that can solve some of our most difficult problems and make our computing life way easier.
It's almost too good to be true. And here's the part you say. OK, Paul, what's the catch? There's a few catches. First sketch quantum logic gates. They have to be special. They can't be like the logic gates that are used in regular computers and the quantum lingo. They must be what's called reversible. In other words, they can't spoil the superposition in the entanglement. If they do, if if when you're actually performing these operations on the cubits and passing them through quantum logic gates, the gates themselves, the operations, if they spoil the superposition. If they spoil the entanglement, then the algorithm can't keep track of the superposition in all these special quantum states anymore, and you can't be guaranteed of the right answer. So every step that our quantum cat makes it must be reversible. And that's how we guarantee that the superposition and entanglement are preserved. It means you can run the logic gate in any direction, which is different than a normal computer gate.
Second catch is is this isn't really useful for all problems. Anything a quantum computer can solve a regular computer can solve, too, with enough time, and we don't have, at least at the moment, and we may never have general purpose. Quantum computers, like the the computer in your phone or your laptop is a general purpose computer. You can pretty much throw any kind of problem at it, and it will solve. It may need a really, really long time to do it, but you can do it in principle, quantum computers may be limited to only a specific set of tasks like technically, anything you can do on a regular computer you can do on a quantum computer. We don't have those actually built a general purpose quantum computer. And besides, quantum computers are only really worth it for a certain limited number of tasks like quantum cats are good at catching mice, but not the greatest at cooking dinner. They're not always better than a classical computer. So factoring factorizing searching big databases. Uh, there are a few problems like that that regular computers have a really, really hard time with.
And and quantum cures are easy peasy for them, and that's why they're very interested. But they're not gonna replace normal computers, and they're not gonna do everything that normal computers do. Like I said, there are only a few dozen known quantum algorithms around 50 or so known quantum algorithms that are worth it. They're actually more powerful than their classical counterparts. Another catch is quantum devils. There's like noise. If you're running your quantum computer, anything that spoils the superposition, anything that spoils the entanglement will mess you up. Quantum states with these superposition entanglement are very, very fragile. If you make your device too warm or you hit it or you shake it, or sometimes even if you look at it sideways, the entanglement, the superposition falls apart. And when it does, the answer that comes out at the end is just junk. It's just nonsense. The the algorithm failed. So, for example, in the story at the beginning of the episode where we found the cat in the middle of the night and the cat was in every room, if I poked at one of the cats, then the superposition would fall apart, the entanglement would fall apart and the cat would not be able to find the mouse.
Another catch is probability. There are two wonderful features of quantum mechanics that we're taking advantage of superposition and entanglement. But there's a third property of quantum mechanics that we wish would just go away. And that's the fact that you're never quite guaranteed. The answer ever in quantum mechanics probabilities is just how the quantum world works. You're never 100% sure you have the right answer, you know, it could be 99% of the time your cat does catch the mouse, but that 1% the cat is gonna be in the wrong room, and there's absolutely nothing you can do about it. So to solve this with quantum computers, they employ various schemes for error correcting. Or you just run the algorithm a whole bunch and then see which answer comes out the most. But it's not necessarily easy. Another catch is that any measurement like I keep harping on destroys the quantum state. You can't check on the status of your computation in the middle of the computation. You can't check on the cat's progress in the middle of the night.
You have to just have stay in bed and cross your fingers. It works. This presents a lot of challenges for computation because you want to make sure you're like doing everything right and everything's hunky dory because you don't want to waste your time and you wanna be sure that you have a right answer. But the only way to check to make sure you're still in that juicy, quantum, superposition entangled state is to look at it, which immediately destroys your juicy quantum mechanical superposition and entangled state algorithms. Get around this or work with it by pausing deliberately deconstructing everything, checking to make sure everything's kosher and then resetting the quantum state. But that slows things down. It makes things more difficult. And then the last hurdle is you actually need to build the thing. Building these things is kind of hard. There are various technologies out there. I'm not gonna get into in too much detail. The most popular is called the Ion Trapped, which uses electric and magnetic fields to trap charge particles. You use these to, uh, prepare a quantum state and I'm just kind of waving my hands around generally instead of actually explaining this.
And then also, you need to keep everything super cool. Because if it gets too warm, uh, you lose track of the quantum superposition entanglement and your quantum computer is, you know, just a really, really expensive doorstop. How far have we actually come? We've come somewhere around 65 cubits. Uh, this number changes all the time. Uh, and this is for the setup I've described with quantum bits and quantum logic. Gates 65 cubits. Sounds like a lot like that's an ungodly number of possible states. But these quantum computers don't have a lot of logic gates in them, and they're pretty unstable to noise, so you're not always guaranteed of the right answer. So it's like it's not as powerful as you might think, because real world considerations You run an algorithm on this quantum computer, it's gonna take longer than you might have thought, and you're not necessarily guaranteed of the answer. One of the issues they run into with quantum computers is when you're trying to run a really big problem and run it through these logic gates.
As it runs, it warms up, and then eventually, if it reaches too much of a critical threshold, the thing falls apart and you get a junk answer out. And so the trick with quantum computers is trying to keep them running while staying chilled so that you can actually get an answer out. Quantum algorithm. Designers and quantum computer builders are searching for something, and are you ready for perhaps one of the dorkiest terms I've ever encountered in my life? We call it It's called Quantum Supremacy. Quantum supremacy is building a quantum computer that can actually run a real world algorithm much, much better than even the world's most powerful supercomputer and not, and not just, like a little bit better. Like like five times faster, like solving a problem that we simply can't with a supercomputer because it's too large and it would take too long again. I'm sorry. I'm not in charge of naming things. It's called quantum supremacy. There it is. Uh, Google has recently announced it claimed that they have reached quantum supremacy, but other people are debating it. Quantum computing is one of those technologies that always seem, uh, 10 years away, like, Oh, just give it.
Just give us another decade of research and we'll build this quantum computer. I mean, quantum computers are gonna come. I think we'll crack the engineering. The technology. We've made huge strides already, especially in theory, and we're kind of waiting for the tech to catch up. But quantum computers, Yes, they will be a revolution of sorts, but they're not gonna replace regular computers. Instead, they're gonna complement regular computers. They're gonna solve the kinds of problems that classical computers have a really, really tough time with. So there will be a niche for quantum computing, but you're not gonna listen to this show on a quantum computer unless you want to listen to all episodes simultaneous. That's let's not get into that, thanks to Kirk on Patreon and at Mihail E on Twitter for the questions that led to today's episode and thank you to my top Patreon supporters this month. Matthew K, Justin G, Justin Z, Kevin O, Duncan M Cody Barbara Kay, Robert MNH and F Chris L, Cameron Naa aone, Tom Scott M and Rob H.
That's patreon dot com slash PM Sutter to keep these episodes coming Thank you so much for listening. I really do appreciate it. Hit me up with more questions. Hashtag ask a spaceman. Ask us spaceman at gmail dot com. Ask the spaceman dot com Just like search. Ask us, Spaceman. You'll figure it out, I'm sure. Also be sure to leave a review on iTunes. I really do appreciate it and thank you again for listening, and I swear I will. I will do a series on the foundations of quantum mechanics one of these days. In the meantime, I'll see you next time for more complete knowledge of time and space. Prime Day is coming July 11 and 12 with two days of epic deals exclusively for prime members, you'll feel like a winner. Number three is amazing deals from electronics to decor. It's on prime day.