Computers of the Future

podcast with. Oh yes, I have two things. The first is that I have to let our listeners know we have a guest producer for this podcast. Yes, Mr Matt Frederick, who you can tell he's a guest producer because before he hit record he said take one. Matt is unaware of the fact that Chris and I always get it in the first take. So Matt, if you would just not even bother next time, Okay. But the second thing that we have to start off this podcast is listener.

I feel like I've been sabotaged here listen all of y'all. So this listener mayo comes from Ivan, and Ivan says, Hello, Jonathan and Chris. I'm a longtime listener and a first time writer, and I think it would be fun to do a podcast on quantum computers. You could answer questions such as, how do quantum computers handle algorithms differently than classical computers? Would I be able to put together my own quantum computer when our quantum computers expected in the

consumer market. I'm guessing about a decade by Ivan. You know what, Ivan, Uh, your definition of fun and my definition of fun may not be the same thing, but we're gonna tackle it anyway, and we're actually gonna broaden it out. We're not just gonna hit quantum computers. We're going to hit computers of the future. Well, this, uh, this podcast is full of holograms and and funky colors and BP noises. Many both ends died to bring us this podcast? How many? Both of them? Yes? Both? All right?

So I guess, um, I guess first we can talk a little bit about sort of the state of computers today. For Chris is just Chris is gone. Chris is gone. All right, I'm gonna keep going with Chris to get it back under control. So classical computers, Uh, we're rapidly approaching the time when most well, I guess most is

probably too too big a word. But some engineers believe we are reaching a critical point in classical computers where we won't be able to get much faster than what we have right now based upon, uh, the traditional method of building microprocessors. I'm guessing you're mentioning in your head at least, yes, I was gonna get to that. So Core's law, this is this all goes back to Moore's law. And if you've listened to our podcast on Moore's law,

you know what we're talking about. If you haven't, right, if you haven't, I suggest going back and listening to it, because it was a pretty good one, as I recall. And uh, but in general, Moore's law it was this this sort of observation that Gordon Moore made back in the sixties. Yeah, he's the co founder of Intel. Yes, I think it was nineteen sixties seven when he made this observation originally, and he observed that over the course of about well, the time varies depending on when you're

looking at Moore's law, but we'll say eighteen months. But over the course of about eighteen months, you would see the number of transistors uh double on a square inch of silicon chip. You would be able to pack more transistors onto that chip. And there were a lot of different reasons for that, but some of it was technological development where you start finding new ways of making smaller transistors.

Part of it was economic because you could find cheaper ways to mass produce transistors on a on a smaller scale and UM as a result, every eighteen months or so, you would see microprocessors get twice as strong as they used to be because you've got twice the number of

components on them. And UH. For years, people have been predicting the end of Moore's law, saying that that has to come to an end because how could we possibly get smaller than what we're looking at now, because right now we have transistors microprocessors out there with transistors that around the nanoscale they're just you know, a few dozen nanometers wide, and that's incredibly tiny, so tiny you can't

see it with a light microscope. UM. But eventually we're going to hit a point where the traditional methods of making these microprocessors aren't going to work because we just can't make something that's small that works with electrons. So at the end of the traditional cycle, which some say is probably within a decade. Um, there are new ways of creating processors that will sort of get around that by making them three dimensional, basically stacking layers on top

of one another, which is you know, cheating. Yeah, that's uh, yeah, it's not really cheating, of course, I mean we're being a little facetious, but it's it would mean that we'd be sticking more with the class sical computer than branching out and trying something really really unusual and different. Um. And there are several different approaches that some engineers are looking at as alternatives to classical computers, things that, if they work out, could be far more powerful and far

faster than anything we've used up to this point. And perhaps, I don't know, eventually power us to the stars where we can make a prime directive and not mess with other people. You know, we made it to the moon with less computing power than a common or sixty four, so you would think that, you know, with an Atari twenty we could at least make it to Mars. So um, moving on the uh. One of the first one we're going to tackle is the one that Ivan was asking

us about, which was quantum computers. So now you're classical computer. It operates using operations on data, using a set of instructions, and everything gets broken down into bits by binary digits one or zero off exactly. So that's it. You've got tons and tons and tons of these bits put together

to make these instructions. Um. You know, a computer might be running uh bits that are or or or figures that are sixty four bits long, which just doesn't sound like a lot, But when you add up all the different combinations that those bits can have, that's a lot. And of course there that's not the upper limit at all. It's just an example. So quantum computers these are different because they don't use bits. They use quantum binary digits or cubits, which I thought they used to measure the

arc but is it turns out there actually data. See I thought it was an awesome nineteen eighties arcade game where you jumped around on a pyramid. Oh, you're right, that's exactly what I was going to say. So cubits, water, cubits, Well, it's it's a special kind of bit really um quantum that we're gonna have to go into quantum theory. I really didn't want to have to go into quantum theory because that's more of a science topic than a computer topic. Also is likely to cause the staff to go get

them up because my brain will explode. Yeah, I may have an aneurysm before I finish this this this next description. So quantum theory. Quantum theory is really looking at systems that are really, really, really small. We're talking on the sub atomic level um. You can also use it to describe some really really big systems too, but we won't get into that. So on this small, small level, things do not behave the way they do in our macro world. Um things that make sense to us because it's on

a classical physics kind of a scheme. They don't that the rules don't apply on the quantum scheme. So to us it may seem like things are breaking the laws of physics. They're not. It's just they're following a different set of laws. One of these laws that will come into uh importance with the quantum peters is that you can have a quantum element inhabits several states at the same time. And I'm not talking like states like Idaho

and Montana. I wasn't going to make that. I'm just I just was gonna head you off at the past just in case. So bits have two states zero and one, or on and off if you prefer. Now, a quantum bit can be both a zero and a one at the same time and all points in between. It can

inhabit all of those states. Now, what does this mean from a computing standpoint, Well, that means that when you're making a calculation, instead of having to run one set of bits to do one calculation, and then a different set of bits to do a different calculation, you could set one set of cubits and do all possible calculations within that that you know realm of of bits that you're using, and it can handle a lot of calculations all at the same time instead of doing you know,

one thing at a time so very quickly. For example, one of the one of the big possible uses of a quantum computer would be to decrypt information and find out what people are actually saying about you behind your back. Because cryptography, a lot of the cryptography we depend upon today um uses It uses factoring. And you take two really really big prime numbers, you multiply them together, you

get another number. This becomes the basis of your cryptography and only by knowing the two prime numbers that were used to generate that big number, are you able to decrypt the information? Now, for a classical computer to find the two largest prime factors of a large number, I mean we're talking huge numbers here, it can take years, like millions of years in some cases for a classical computer to decrypt or to to find those two factors.

A quantum computer, assuming that you have one that's powerful enough that can run cubits, could find it in a fraction of that time. So that would totally freak out. I was gonna say that that's little, uh mind, Yeah, it's um, which is when you get into quantum cryptography, which again is going beyond the realms of this podcast. I really really can't get into it because, honestly, people,

I barely have a grasp on this concept. I've listened to so many professors and scientists talk about quantum computers quantum physics, and it seems like a common element is that they all will eventually admit to not really being able to grasp everything about quantum theory. This is pretty heavy stuff. Yeah, and like I said, a lot of it seems to contradict what we know. For example, you wouldn't normally say that an object can be multiple things

all at the same time. Um, and there's another element to quantum computing that's kind of tricky. Why why don't we have quantum computers now? If we understand quantum computing, aren't there quantum computers out on the market now? There are quantum computers being worked on in labs um. There was one that was reportedly up to sixteen cubits a couple of years ago. But why aren't we seeing them now? Well, one of the reasons is because it's really hard to

keep a quantum computer in working order. Um. There are a couple of different reasons for this. The elements tend to have a habit of interacting with things around their environment as opposed to each other. So then you get corrupt data because from what from what I understand, everything stays the way it is as long as nothing touches it.

But since we're talking about very very tiny things and things into other things, you have to be able to isolate everything and not have it interact with anything other than what it's supposed to interact with. Otherwise your your results are not trustworthy. That seems problematic There's also the the the the old principle of if you observe it, you change the observed. You know this. This principle often mentioned as part of the whole shrot Injurer's cat problem.

Are you familiar with singers? A familiar with for those who are alive maybe until you open the box? Um so Schrodinger's cat. This is a classic quantum physics problem. Uh the The idea being that you have a cat shut in a box. There is a canister of poisonous gas that will release sometime between say, five minutes and twenty five minutes, and there's no way of predicting. It's just gonna it's gonna pop open randomly sometime between five

and twenty five minutes. If you open up that box within twelve minutes and observe the cat, it will either be alive or dead. But before you open up the box and observe it, it is, according to this principle, both alive and dead at the same time. It only becomes one or the other for sure when you open it and observe it, you have changed it. The reason behind this is it gets really kind of complex. But

if you were to try and observe quantum particles. Just by the act of observing them, by hitting them with a photon of light, you have changed the behavior of that quantum particle. Therefore, it is no longer doing what it used to do, and your measurement doesn't really matter anymore. You're not measuring what it was what it had been doing. You're measuring what it's doing right now after you've hit it with light. Quantum computers have a similar problem. You

try and observe them. They've become classic computers, and you've just threw into your quantum computer. Alright, So you have to find a way to measure the results in such a way that does not disturb the cubits themselves. And also all your results are coming out in sort of a probability as opposed to this is definitely the answer. You might get seven percent chance that this is your answer. There's a twenty percent chance that this is your answer.

There's a twelve percent chance that this is your answer. So not necessarily something you want when you want to find out what the temperature is outside. I was gonna say it sounds a lot like the computer models that the meteorologists use, because they say, well, on this computer

I'm getting this, and that computer I'm getting that. So an answer to your other questions, ivan Um, I found an article in Nature that suggested that, well it was by it was quoting Andrew Stein of the University of Oxford in the UK um and uh, quantum computers may not really hit the consumer market. They may be more niche products because of the way they do computation. I mean, it's not like we're going to be going to quantum

Facebook and quantum Twitter to do our quantum email. Um. They're they're really sort of high high end computing needs.

Probably not until we're eating all our meals in pill form, right, But around is when he expects to to see that happen, So we we should see them more prominently, assuming that engineers can get beyond the problems of you know, the more quantum logic gates you add to a computer, the more difficult it is to control the the cubits, and therefore the more difficult it is to get reliable results. You have to get past that problem first before you can actually build a quantum computer of really of of

any meaningful power. Well, at least, quantum computing isn't what I originally thought it was, which was you know, taking your laptop into Sedan. Um. Wow. Wow. Anyhow, so high speed stuff has an opening. UM. The the the other part of your question. Can I build one? Now? I mean, if you're maybe if you're at the Stanford Research Institute or something, if you're if you're on one of these

these projects that, yeah, you might be. You're not going to pick up the parts at best by that, No, it's not gonna be one of those things that you order out of Popular Mechanics or anything like that. All right, so we can move on to two different futuristic computers. I was thinking of the DNA computers. DNA computers, okay, yeah, so the ox i ribo nucleic acid computers they run

on good old fashioned DNA. UM. This is another one of the those computers that could potentially replace classical computers, at least in research institutes, just like just like quantum computers could. UM. Again, you're talking about computers that can perform calculations on a parallel kind of scheme where they're they're running multiple applications all at the same time, multiple

um uh computations. I guess you could say, UM and it runs on DNA, and one of the things that it really has going for it is that DNA is kind of cheap because there's a lot of it around. Turns out, what do you know, spitting this cup all right, we've got enough computing power for the next five years. Um. Yeah,

it's kind of cool. And there are a lot of different teams that are working on DNA computers and they're really looking at molecular biology as a way of, uh of of advancing computer science to levels that we can only kind of dream of right now. Um, again, this is stuff that is kind of in the research phase. It's it's fairly recent. Um. The the original idea of the DNA computer, I would say probably dates back to early nineties, So quantum computers actually we're theaterrized back in

the early eighties. But uh, we're still in that very early stage where people are looking at ways where they can harness d N A and use that as a coding mechanism for um for computational problems. Do you have anything to add to that? Not to that topic, I was going to bring up some uh, some computers at the very near future, and I was thinking that, you know, based on some of the other topics we've discussed on

the podcast. I think quantum or quantum computing may not be what we see on our desktop in two years, But what we see on our desktop in two years is probably going to be very small, like portable, and may not even have a hard drive in it because everything is moving to the web. I mean memory and uh, memory and storage space are basically you know, very very very cheap at this point, and I think that's just going to encourage more companies to offer cloud computing for

storage and software as a service. Um, you're likely to see netbooks and tablets and uh, you know even you know, cell phone convergence devices. Hey, I use your favorite word, um, you know so and in the very near future, people, a lot of people, including a person sitting across me, think that thinks that desktops are going away. I think they might become the realm of enterprise, you know, uh,

method you'll still see him. I thinking in school labs and and uh and corporate offices, right, I don't think. And apparently our producer thinks he's going to have a desktop computer for a very long time because he's miming it. Um. Either that or he's itching in some spot. Well, he uses a Mac and that doesn't count. Everyone knows about my anti max bias. Your so the the yeah, we should have like the text stuff drinking game. You drink every time Jonathan says he has an anti mac bias.

Drink every time the word convergence comes up. Um. Cloud computing would be another one, you guys would be tanked by now. And then there's something else magnetic ram oh yeah, yeah, things that are going to improve the day to day

performance and portability of computing. Well, that was another one of the one of the elements of DNA computers is that DNA, of course, is incredibly tiny, and you can pack um enough DNA into a cubic centimeter of space too to get I think it's something like something ridiculous like ten terra flops of of processing speed, which is pretty freaking fast. Um it's very powerful computer, especially when you consider that's one cubic centimeter. That's not necessarily all

that you would have. UM. I was going to talk about one other kind of possible future computer, optical computers or photonic computers. Yeah. These are computers that instead of using electrons as uh, the method of conveying information. That's the way classical computers do convey information. If you weren't aware, Um, it's a hole through electrons. It's there. Aren't actually hamsters running around inside your computer, no matter how old it might be. I mean, unless you turn your computer into

a hamster farm, which I guess you could do. No. Photonic computers use light instead of electrons, so little beams of light to turn on and off. And you know, it's just like bits. You've got two different states. You've got on and you've got off. And of course light travels pretty darn fast. In fact, it travels faster than just about anything else we can think of. So you're talking about a high speed computing system that could potentially

leave classical computers behind. Again, you have to be able to build a an optical core, an optical CPU at the center of this computer. Um. I've read about a few different experiments that have tried to do this. Most of them involve um cooling the CPU to a temperature not that much warmer than absolute zero, which is kind of impractical for the home. It seems reasonably impractical. Yeah,

we're probably fairly expensive. Yes, when you're when you're talking about maybe a degree or two above the temperature of deep space. That's not necessarily something most of us can achieve nor would want in our home. A honey, where do we put the liquid nitrogen? Right? Actually, you probably need liquid helium. I think nitrogen would only get you down to yeah, because liquid helium would be what they use over at the Large Hadron Collider. But yeah, optical computers.

That could be another another future computer that will replace classical computers, at least in the research firm area. So again, not necessarily something that we're gonna see on our desktop at home we're in our Xbox games or anything like that, but it would be it would be I got you know, you'd be like, man, it takes almost point zero zero zero eight seconds for Firefox to load. I can't believe how slow it is. You know, we'd still complain of it would be instantaneous to us, but it would be

just slightly less instantaneous than it should be. So uh well, I mean that's those were the three biggies that I wanted to hit. Were um, quantum, DNA, and optical. Uh did you have anything else? Add? Not really awesome? Then we can wrap this up and of course, We've already done our listener mail and I'm not going to torture you with a second one. So everyone, if you want to learn more, we have articles at how stuff works dot com about all this sort of stuff, including quantum computers,

quantum cryptography, DNA computers. So if you really want to learn more about the future of computing, visit the website. We go into a lot more detail there and linked to other really good resources as you can, And if you really want to explore this and learn more, I highly recommended. The topics are so deep and and dense it's hard to really tackle them in the podcast format, especially since we don't have any real visual aids that we can throw in there either. So I do recommend

visiting the website if you're interested. And as for us, well, I guess we will talk to you again really soon. Actually didn't mention the address, Oh the email address, all right, I need to go back and do that. Then, yeah, what No, I can't work like this. I can't work like oh fine, We're just going and take one alright, alright, listen, everybody listen. That was just to fake out all the people who aren't really tech stuff fans. So all you real tech stuff fans who have stuck around, you didn't

hit stop on your iPods. You're awesome. You're way better than those losers who already stopped the podcast. So if you want to tell us how awesome you are, you can do it by emailing us, and our email address is tech stuff at how stuff works dot com, and we will talk to you what I just wanted to point out if you've got we sometimes we get technical problems.

People write us in with with actual tech support questions rather than how do quantum computers work, which is a much easier question to answer than what the heck did I do to my hard drive? If you actually have a technical support problem, you should probably get in touch with somebody who can help you with that in a

more expedient fashion than we can. Also, we may or may not know what's going on with your particular system, so I would advise you know, finding your brother or person who normally does your tech support and get them to help with those kinds of questions, and we'll handle the how stuff works type of questions. Now, I should point out that all you awesome people who stuck around, clearly you wouldn't have a problem to write down about in the first place. It's all those people earlier on

who would be writing in. But of course they wouldn't do it anyway, because I didn't say that even them, did I. All right, well, then that was a super command. Take that, Fred Levin. Alright, then I'm gonna go and take a nap now, And for the rest of you guys, we'll talk to you again really soon. For more on this and thousands of other topics, visit how stuff Works dot com and be sure to check out the new tech stuff blog now on the house Stuff Works homepage.

Brought to you by the reinvented two thousand twelve camera. It's ready, are you