Prof. Chris Bishop: The Future of Computers

well thank you very much it's good to see such a good turnout thank you all for coming this is a talk about computers of the future but it's not a talk about tomorrow's laptops or next year's mobile phones it's not a talk about technologies of the near future it's a talk really about the ideas the concepts the science that makes digital technology possible so I'm going to talk about some quite futuristic ideas for new kinds of computers new kinds of computation and I'm going to begin with one area of computer science which I think is is undergoing a very very exciting transformation I think the next 10 years will be tremendously productive and it's actually my own field of research it's the field of machine learning now the idea of machine learning is actually very simple there are lots of problems that we'd like computers to be able to do like recognizing handwriting recognizing faces understanding speech and so on that are so hard that you can't program a computer to solve them directly it's impossible so instead we teach the computer the computer learns by example it learns from experience just the way people learn from experience so I'm going to begin with a demonstration of machine learning now this of course is not an ordinary lecture this is a Christmas lecture which means that you the audience get to play a full and active part in all of the demonstrates in some of the demonstrations and that's true of this first one so this example of machine learning is going to make use of these two cameras here which are pointing out at the audience so if you can bring the feeds up from these cameras on the screen so there's one and there's the other so if you don't eyes just wave at the camera isn't just check that they're all working that you can see yourselves that's excellent ok right what we're going to do now is we're going to train the computer to recognize your movements and then you're going to be controlling a computer game so we're going to play a computer game this evening and the way we're going to do this is you're going to be leaning to the left and right in your seats and you're going to use that to steer a racing car ok we need everybody to join in to make this work alright so we're going to begin with a bit of practice so I want you to all lean to your left to your left as far as you can X that's really good get really friendly with that person sitting next to you fantastic okay now try leaning to the right exit this is going to work really well that's wonderful okay and fantastic what we're going to do is take pictures of you leaning left and right and that will be the training image and we'll use that to teach the system what it looks like when you're leaning and then we'll use that to to actually race the car now the other thing is it's very important that you're all working together if half of you lean to the left and half of your lean to the right then nothing much will happen it won't work so to help you sort of coordinate what I'd like you to do is if you think everybody should be going to the left I want you to shout left okay if you think we're going to the right want a shout right so let's just practice that I want everybody's shout left please mmm not bad okay bit louder this time everybody shout right Oh much better okay so now let's click the training images so I'd like everybody saline to your left as far as you can and we'll capture that image now everybody lean to the right as far as you can we'll capture that image okay set up right now if you look up at the top right there you see a little red bar so if you all lean and to your right for the moment we see the bar go to the right of your lean to the left the bar goes to as a natural - that's working really well okay now we're going to play a racing game so Marcus is going to fire up the game and Marcus is going to control the speed of the racing car you're going to be responsible entirely for steering and remember to shout out nice and loudly so that you're all working together and we'll start off nicely slowly and it looks so easy will speed things up a little bit okay here we go laughs you get good yeah left okay excellent when you give yourselves a big round of applause for that let's pretend this okay so that's machine learning let me show you a sort of a more practical or everyday application of machine learning which has to do with image editing so this is some some work that colleagues of mine have been doing and the goal is to provide very simple mechanisms for people to edit images at home without much skill without using very complex software packages so this is a colleague of mine Antonio and the picture of him playing croquet let's suppose that we wanted to cut Antonio out of the image and paste him into a different image now if we have a sort of pair of electronic scissors it would take a very long time to go all the way around the edge and and it would be very tedious and boring to do that so this is a system called grab cut which makes this a whole lot easier all I have to do is to draw a rectangle around the figure and it does a little bit of computation and you can see this line of marching ants which is the computers attempt to work out the outline of the object now it's um quite a good job but of course it doesn't know about two heads and feet and croquet mallets and things so it hasn't got it quite right so it's very important to have very easy ways to drink little adjustments to this so it's missed his head off so if I just paint on his head a little bit you see it's captured his head and it looks like it's missed his feet and there we go so that's cut the object out so I can copy that and then paste it and there's a copy of Antonio a bit smaller there we go alright so that's a that's an example as an example of machine learning in action now things like the racing game the the grab cut example and your cellphone laptop computers on all this wonderful digital technology is all made possible because of an extraordinary piece of engineering and it's called the microprocessor and I've got one of the micro processors here this is one of the current generation of desktop computer processors it's a dual-core processor here it is in its packaging it has about a hundred or so gold contacts which connect it to the rest of the computer and if we look inside that packaging we see the actual processor chip itself and this is it it's a little square of silicon that's the shiny silicon and the processor itself is built into the top surface of this tiny chip of silicon and it's covered a little copper contacts into about 100 copper contacts that go to those gold pins so that's a microprocessor and it's little chips like that that have completely changed our lives every household today I calculate it has at least a hundred of these little computers in all sorts of everyday objects not just your laptop and your cell phone but in every credit card for example underneath those little gold contacts on your chip and pin credit card is a little microprocessor fully general-purpose processor it has 30 times the speed and about a hundred times the memory of the computer that took the apollo astronauts to the moon in the 1960s that's inside every credit card so these chips are everywhere and they are quite extraordinary and they're extraordinary in in lots of different ways let's just have a quick look at the surface circuit of this actual microprocessor that I have here you can see just a tiny fraction of the complexity of this processor there are actually about 400 million components in this device each of which is a hundred times smaller than a bacterium it's remarkable not just for the engineering complexity but also for the speed so I want to look a little bit at the extraordinary speed of micro processes and again to do this I'm going to need your help so what I'm going to do is have a little bit of a test first of all to see how fast you are at doing some multiplications okay so hope you're all wide awake what I'm going to do is bring up some simple arithmetic or calculations as soon as you know the answer just shout it out as loudly as you can okay so this is a speed test everybody ready excellent twelve pretty good a bit slower there hey ready okay this one is a lot harder well I certainly would need a pencil paper it would take me weeks and I'd get the wrong answer anyway but this little tiny chip of silicon can do billions of those calculations every second and get every single one of them correct to sixteen significant figures that's incredible so I can do a calculation like that in a billionth of a second now it's actually very hard to understand just how tiny an interval of time a billion for the second is so I want to do this evening is to try to give you a feeling for what's meant by a billionth of a second so to do this what we're going to do is to take something that's very fast and we're going to time it we're going to time it in billionths of a second okay now computer scientists and physicists and so on refer to a billionth of a second as a nanosecond so going to take something very fast and we'll time it in nanoseconds now the very fast thing that we're going to choose is is an explosion and to make an explosion we need an explosive and the explosive that we're going to use is something called nitrocellulose so it has quite an interesting history it was discovered by a swiss german chemist christian friedrich Sherm bine and he was doing some chemistry in his kitchen one day for some reason and he spilled some nitric acid so he reached for a cloth and he wiped up the nitric acid and it happen to be a cotton cloth and he hung the cloth out to dry on the stove and when it dried it exploded and he discovered nitrocellulose so let me first of all show you some some nitrocellulose now because it's um there's a Christmas lecture I thought would have a little festive example of some nitrocellulose so this this snowman who's going to be sacrificed in the interest of science is made entirely entirely from nitrocellulose so it looks a bit like cotton wool but it behaves slightly differently as you'll see now this won't make very much noise it makes a bit of a whoosh noise you might hear it if you listen carefully it does make quite a bright light you might choose to look slightly to one side when we set this off okay so this is nitrocellulose so this is nitrocellulose burning in the open right yes okay so that's nitrocellulose I'm going to use some nitrocellulose now to create an explosion I'm going to do that using this apparatus so here we have a much smaller quantity of nitrocellulose but this time it's enclosed in a tube and the tube is sealed at this end and it's open at the far end so mu ignite the nitrocellulose it'll create a pressure wave and the pressure wave will travel very quickly down the tube now at this point on the tube we've blocked the tube off with a thin sheet of paper and across the paper and just see there we've painted a line of silver paint so that's electrically conducting and that's connected by wires back to this box of electronics a little bit further on in the tube we've got an identical piece of paper again with wires connected back to the electronics this box of electronics is going to measure the time in nanoseconds between the first piece of paper being broken and the second piece of paper so it'll time that pressure wave down the tube now this is connected by a cable back to the laptop and I think we're going to bring the display from the laptop on this screen so when this is fired this screen will show in nanoseconds how long it takes the pressure to travel from here to here what I think you'll observe is that pressure wave travels extremely quickly now this time the explosion will make quite a loud bang those of you who don't like loud bangs do feel free to take your fingers and just put them in your ears to keep your sitting near the front I shall wear some ear defenders because I'd be standing right next to it and then we'll we'll set this off and we'll measure how long it takes the pressure wave to travel that short distance so this is about a gram of nitrocellulose and it's going in three two one and that looks like it took just just under a million nanoseconds nine hundred and fifty thousand nanoseconds so in the time it took the pressure wave to travel from here to here that tiny chip of silicon could do nine hundred and fifty thousand complex arithmetic or calculations and get every single one of them correct so that's I think that's extremely remarkable now how did computers get to be so fast obviously extraordinarily quick well computers have grown in speed in rather special ways something we call X exponential growth so to understand exponential growth let's have a look at an example taken from biology in this case here we see some bacteria and this is going to be a time-lapse video of these bacteria dividing so if we run the video now we're starting off with I think just four bacteria but every twenty minutes each bacterium splits in two so every twenty minutes the total number of bacteria will double before then eight and sixteen and so on and you see very quickly the numbers are growing hugely takes off kind of slowly and then really accelerates away and that's called exponential growth and if we plot a graph of the number of bacteria or if you like the speed of computers against time we see a curve that looks like this it starts off fairly flat starts to grow and then it really accelerates away in this tremendous fashion and that really is a very dramatic kind of growth the rate at which it's growing is itself also growing exponentially and it's really that exponential growth in the speed of computers that's really driven the digital revolution here's another example of exponential growth which is one of my favorites and this is a chain reaction involving mousetrap so I bought a mousetrap along to show you so here we have a mousetrap and I'm going to arm the mousetrap and then place a ping-pong ball on the trap and now if I take another match another ping-pong ball and drop it on the trap you'll see it sets the trap off so if you have a lot of these mash traps we can set up a source of chain react so let's have a look at that now so if we run this video we have a box this box contains 225 mousetraps and we're dropping one extra thing upon boiling okay that's over pretty quickly so let's just I run a slow-motion version of that so it in goes the ping-pong ball setting off a couple for growing the number growing exponentially and then of course eventually it runs out of ping-pong balls and it it stops so that's another example of exponential growth so really it's exponential growth that is allowed computer technology to take off in the in the very impressive way that it has this exponential growth of computers won't go on forever in fact it may not go on for very much longer and when it stops then we'll the digital revolution stop or will we think of other ways of making fast computers so what I want to look at now is some interesting ideas for ways to make new kinds of computers computers that are much faster than any present-day computer and I don't just mean twice as fast or ten times as fast but could we build computers that are millions or billions of times faster than any computer that we have today well one idea one perhaps crazy idea that people have had for building fast computers is to use chemistry and the reason why chemistry might be a good idea to build fast computers is this if I look at this bottle of water the number of molecules in this bottom bottle of water is about one with about 23 zeros on the end of it so if somehow I could get each little molecule to do a tiny calculation even if it was quite slow with 10 to the power 23 molecules all working together I would have an extremely powerful supercomputer so that sounds like a really exciting idea but there's a bit of a problem and the problem is that chemistry doesn't look as if it would be any good for computation after all computers work by eating lips go round loops they do little calculations over and over again each time doing something great it's the accumulation of billions of very simple steps that add up to wonderful computer graphics or control of an aeroplane or any of these other wonderful applications now chemistry doesn't seem to do that chemistry you go from the start you have a chemical reaction and you arrive at the end so let's just have a look at some sort of standard chemistry just to see to see how that works so here's a piece of chemistry this is a balloon and the balloon contains a mixture of hydrogen gas and oxygen gas so we will set fire to the balloon and what will happen yes what will happen is the hydrogen and the oxygen will combine together to make water and of course release a lot of energy so there'll be a ball of flame and again this time will be quite a loud bang so again if you wish do feel free to put your fingers in your ears and now we look at the chemistry of hydrogen reacting with oxygen so this is hydrogen plus oxygen makes water in 3 2 1 okay so that was a there was a typical chemical reaction it starts at the beginning and it goes to the end hydrogen plus oxygen makes water the reaction stops it's all over if we want to do computation with chemistry we're going to need something that looked more like a typical computation something which has a loop in other words something which oscillates now for a long time everybody thought including professional chemists everybody thought that chemistry couldn't oscillate chemistry goes like the the hydrogen balloon goes the beginning to the end or it sort of reaches a halfway point in which the forward reaction and the backward reaction balances and everything just stops the chemistry doesn't oscillate they thought that was impossible and then in the 1950s a Russian chemist called Boris Berezovsky reaction which actually oscillates which goes forwards and then backwards and he wrote up his public eight he was read that up as a paper and he sent it to the top chemistry journal in in Russia and the editor of the journal reviewed this and rejected the paper because of course chemistry doesn't oscillate and so bell Asaf sent it to another prestigious journal and the editor of that journal also rejected the paper for the same reasons and Benes off apparently became very depressed and fed up and even even ceased to be a scientist after this and so this discovery was lost for about a decade until a graduate student came along their graduate students are great because they don't know that certain things are impossible so they just go ahead anyway and his name was Boris Bela soft and he discovered sorry his name was Anna told zhabotinsky and he discovered Bela soft notes and he reconstructed the reaction and eventually he was able to travel to Vienna where the the reaction was published his work was published and it became very very well known and it's now known as the Bela Savas a buttinsky reaction now the little twist to this tale because a couple of years later to a high school teachers in San Francisco were messing around with the the Bellas often zhabotinsky reaction in the in the school laboratory it sort of after hours and they discovered another kind of oscillating reaction which is now named after them it's called the Briggs rasher reaction so now I'm going to demonstrate the Briggs rasher oscillating chemical action and I think this really is one of the most beautiful pieces of chemistry so this is a magnetic stirrer it's just stirring the liquid in the flask just because I'm lazy so it's doing all the stirring for me so that's that's all that's going on there and in this flask is some water with a little bit of hydrogen peroxide just speed up the stirrer there what I'm going to do now is to add another chemical this is water this also contains potassium iodate and a little bit of sulfuric acid so I'll add a little bit of this second mixture to the flask and nothing much happening so far but I'm now going to add a third mixture this is water containing malonic acid and manganese sulfate and also a little bit of starch so when I add this third liquid I want you to keep a close eye on this flask and we'll see what happens starting off with a clear liquid it's now turning a sort of amber color keep watching it's turn blue so just keep watching we'll see what happens so it started off clear it looked like water to begin with it turned an amber color then it very suddenly turned blue now if we keep watching we'll see that it's becoming paler and paler hopefully in a minute okay it's becoming paler can you see that keep watching mmm okay so chemistry is a subtle and complex these doesn't it what I can just about see but what you should be able to see in a very spectacular way is what what usually happens and it may be maybe the temperature isn't quite right but what usually happens with this reaction is it starts out clear it turns amber it very suddenly turns blue and then it gradually clears back to water like color again and then it turns amber and then it suddenly turns blue again and it is actually doing it I can see it but it's not a very dramatic it's not clearing completely it's still staying partially blue sure whether you can see that on the screen or not perhaps you can okay well that usually is a very spectacular reaction tonight it's it is it is oscillating but it's just it's just not going completely clear but that is an example of an oscillating chemical reaction it is going around in a loop it's going clear then amber then blue then clear then ambulant blue and so on so could we use that sort of chemistry to do computation could we build a computer out of chemistry like that well it turns out we can't we can't because chemical reactions like this are not quite complex enough they're very complex but they're not quite complex enough and mathematically it was proven in fact by a colleague of mine dhulfiqqar delhi who's also a graduate of edinburgh university that system such as this are not sufficiently complex to build general-purpose computers we need to allow a little bit more complexity and one way to do that is is this so this is another example of an oscillating chemical reaction but this time it's a little bit different so we've just run the video what you're looking at here is a glass dish and the camera is sitting above the dish and I'm pouring in a mixture of chemicals has this orange color and just mixing it around and then what I'm going to do is just to leave the dish very very still so this time instead of being continuously stirred the chemicals are forming a very thin layer now what we've done is we've speeded up the film by a factor of 64 and now we'll see what happens you see that starting to form these regions of blue color if we keep watching you'll see what's happening is again this is oscillating so if you look at any particular place in this dish at any particular point is oscillating between orange and blue and orange and blue but as well as forming oscillations in time it's also forming oscillations in space and whereas in the flask the flask may contain perhaps twenty chemical species and that limits how much information we can store here we can store as much information as we like because we can store different pieces of information at different places in the dish and people have even used chemical reactions like this to build simple logic gates there's a version of the reaction that's sensitive to light and then they print a pattern of light onto the chemicals to set up an initial condition then the chemical reaction runs and these oscillations act like logic gates and perform simple computations and again it's been shown that systems like this are sufficiently complex to do general-purpose computation the problem of course is that in this example it had to speed up the video by a factor of 64 otherwise we'd all fall asleep with boredom and that wasn't the goal the goal was to produce really fast computers not something very slow so although we could build computers that way it's not going to help us with our goal of building very fast computers but there's another way in which we can take the chemistry of oscillating reactions and generalize it to something that can be used to do general-purpose computation and that's something that's been around for a very long time and it's this is DNA and so today people are looking at using DNA to build computers but not for general-purpose computation but for some very different reason and it's this DNA as we know is a sequence of bases shown by these colored slices and the particular choice and order of those bases stores information so DNA is a digital information storage system and that's been known for a very long time since the time of Crick and Watson but what's been realized very recently is that DNA is much more than simply a passive store of information the DNA creates proteins and the proteins in turn affect the extent to which DNA expresses other proteins and so on so the complex biochemistry of a living cell involves lots of feedback loops lots of complex oscillations and today I think one of the most exciting frontiers of computer science is the use of ideas from computation mathematics that was developed to understand the structure of computer software and using that mathematics to understand the biology of DNA and the chemistry of living cells and just to give you an example of where this may be leading some time in 2002 Shapiro and colleagues at the vitamin Institute in Israel published a paper in Nature in which they demonstrated in a test tube a DNA computer which is able to detect when a cell is becoming cancerous and can release a sequence of DNA which can destroy the cell so offering the potential possibility of a new route to curing cancer and that's made possible because of the theory of computation so that's hugely exciting and today the interaction the intersection between biology and computation is one of the most exciting areas of the field but it didn't achieve our goal which was to produce an extremely fast computer so let's revisit that goal and think is there another way in which we could build a computer that's very much faster than anything we have today and one idea comes from physics so let me show you a very curious and interesting piece of physics so let's imagine we do the following experiment we'll take a light source a laser and we'll shine it at a screen and the screen has two slits now let's start by covering up one of those slits and then behind the behind this this this wall will put a a screen and the laser light will go through the lower slit and will form a patch on that spree so nothing very surprising there if we cover up the other slit instead and of course we get a a single patch of light in a different place again nothing very surprising however if we uncover both slits at the same time something very interesting happens instead of getting two patches of light as you might expect we got a whole series of patches these are called interference fringes and the reason they occur is follows the light is passing through both slits and it's arriving at the screen on the right-hand side now light is a wave and if the waves from the two slits are going up and down together they add up and produce a bright region but in other places they're going in opposite directions and they cancel out and we get a dark region so those dark and light stripes arise because the light has passed through both slits at the same time okay so that's quite interesting but now something very bizarre happens let's suppose we do the experiment again but this time we turn down the intensity of the laser we make it a very very dim laser now light also consists of particles light comes a little chunks called photons you can't cut a photon in half it's the smallest piece of light that you could have a single photon if we make our laser very very very dim it's only going to emit one photon at a time okay so I'm going to fire one photon at a time but these two slits and here's the screen and this is this is what we're going to see and if we just run this video now so this video has again been speeded up this takes about half an hour to do this experiment so each of those little dots that you see is a single photon arriving at the screen having been fired at those two slits now if we wait long enough we see something very strange happens the photons don't land uniformly on this screen they land in form of those stripes so those interference fringes are being built up again and that must mean that those individual photons have actually gone through Bo slits so in some bizarre way a single particle of light has managed to pass through both of those slits at the same time okay thanks very much so we've seen that a photon a parting of light can somehow pass through two slits and arrive at the screen that photon was in two places at the same time and that's a very bizarre effect from quantum physics we call it superposition it's not just photons that can be in this superposition so here's a model of a molecule called a carbon 60 or a buckyball exist of 60 carbon atoms linked together and they're in a shape a bit like a football a molecules as big as this have been put into quantum superposition states in which a single molecule is in two different places at the same time that's a very strange effect that's not something we normally see in the everyday world but down at the quantum level inside individual atoms or even in the case of moderately big molecules like this it is possible for something to be in two places at the same time so can we take that very strange piece of physics and use it to build a really fast computer well here's how we might try to do that so all you to imagine a coin now a coin of course has heads on one side and hopefully tails on the other side a coin lying on the table is either heads up or its tails up but if you could imagine a quantum coin that a quantum coin could be in a superposition in which it's both heads up and tails up at the same time okay supposing we could make such a quantum coin what could we do with it well one thing we might want to know about a coin is whether it's a fair coin so a fair coin is heads one side and tails the other okay if somebody are trying to cheat you they might try to make a bet with you using an unfair coin the coin that we say heads both sides for instance so they're really four possibilities it could be heads and tails or it could be tails and heads one give your head some heads or it could be tails and Tails let's try to take our coin and see if it's fair we're going to make a measurement on the coin to see whether one side is heads and the other side is tails if it is we'll call that a fair coin now this may not sound like computation but actually making a measurement like this is a little computational step mathematically it is a piece of computation that we're doing here now with ordinary coins classical coins coins that behave the way conventional computers behave if you like we'd have to make two measurements we'd have to measure one side of the coin then we'd have to turn it over and measure the other side so we've measured both sides of the coin so we see if they were the same in which case it's an unfair coin or if they're different it's a fair coin but if it's a quantum coin we could put it into this superposition where it's both one side up and the other side up at the same time and we can make a single measurement which would tell us whether the two sides were the same because if you like we can see both sides at once and so in a single measurement we can ask the question are the two sides the same or are they different so a conventional calculation needs to make two measurements whereas in the quantum world we can do just one so this quantum computer if you like would be twice the speed of the classical computer because it needs one step instead of two now at this point you may think we're getting something for nothing we're not really because the quantum computer would tell us whether the two sides were the same or different supposing it said the two sides were the same we still don't know if there are two heads or two tails we need to make a second measurement to find out if it's two heads or two tails we still need to make two measurements in order to get information about both sides of the coin and if it's a fair coin one step of a quantum computer will tell us it's fair but it won't tell us which side is the heads on which side is the tails we still need to make a second measurement to find that out so we still need to make two measurements to get all the information so we're not getting something for nothing what we are getting is a way of answering certain very specific kinds of questions very quickly in a way that a conventional computer cannot do so I showed you that a quantum computer might be twice as fast as a conventional computer that's not very exciting because after all wait 18 months the conventional computers will be twice as fast but let's just think about what happens if we scale this up now as I'm sure you all know everything in the digital world is stored as ones and zeros all the information in your computer all the data and so on is stored as great long strings of ones and zeros and each of those is called a bit or a binary digit so a binary digit is either a 0 or a 1 now we've seen that for a single binary digit something which can be in two states the quantum computer would be twice as fast as the is the conventional computer there are two possibilities let's consider two bits now two bits can be in four possible States as you've seen here naught naught naught 1 1 naught and 1 1 so if a conventional computer wants to process all of those possible states it's going to need 4 steps there are four possibilities that we'll have to look at one at a time it will take 4 steps in a quantum computer we could take those 4 possibilities and put them into a superposition so our two quantum bits can be in each of those four possibilities at the same time and so our quantum computer can do in one step what a conventional computer would take four steps to do that's with 2 bits with 3 bits there are eight possibilities so every time we add another bit to our quantum computer we double the number of possibilities and therefore we double the amount by which a quantum computer can be faster than a conventional computer so if we had just 10 bits then there are a thousand and 24 possibilities to raise to the power 10 so our quantum computer be a thousand times faster than our conventional computer and if we have just 300 bits with just that's you know about the number of people in the room this evening with 300 bits this is how many possibilities there are is this number okay that's a big number that is bigger than the number of atoms in the universe so that's the amount by which a quantum computer could be faster than a conventional computer and that is a huge number the earth is mostly made of silicon if you took all the silicon in the planet Earth and turn it into silicon chips it would be but nothing compared to that sort of speed so people are very very excited to try and build quantum computers because with that sort of speed-up we would be able to answer computational questions we'd be able to do computational calculations which no conventional computer will ever be able to do now there is again a little snag you've noticed that quantum computers aren't on sale in the shops yet and the reason has to do with that superposition that quantum superposition is a very fragile thing it's very easily disturbed and so far people have only managed to build quite small quantum computers in fact as of today one of the most complex quantum computers built is is this this is a model of a molecule of per fluro butadiene I'll iron complex I'm sure you recognize it immediately and this has just seven quantum bits and this molecule was used to show that the number 15 is equal to 3 times 5 now that doesn't sound very exciting but it is very exciting it's very exciting for two reasons the first reason is that somebody has actually built a fully working quantum computer that's done a calculation end to end a real calculation that's one reason why it's exciting the other reason why it's exciting is that if you can take numbers and work out their factors then you're able to solve one of the biggest unsolved problems in mathematics which is how to find the factors of large numbers very efficiently and that's important because all of internet security all of our software for encrypting information and so on depends on the fact that conventional computers cannot do these factorization calculations very quickly so it would take a a normal computer at the age of the universe to factor the product of two very large prime numbers whereas a quantum computer would be able to do that in a few seconds or a few minutes and so it would have a very profound impact for things like quantum quantum security for for the security of digital information and there are other applications for quantum computers as well such as searching through enormous databases at very high speed or very importantly simulating other quantum systems so there's a lot of interest in building quantum computers and the challenge is to scale up from seven quantum bits up to a few hundred or a few thousand and this is still largely an unsolved challenge but lots of people are working on this and there are lots and lots of different approaches so I'm going to finish by showing you one phenomenon which is being used as a possible way to build quantum computers it's called superconductivity so this is a piece of ceramic material very special materials called yttrium barium copper oxide there's what's called a high-temperature superconductor now high-temperature here is a sort of relative term we're actually going to make this extremely cold when it becomes extremely cold something while the special will happen which is it will lose all of its electrical resistance so to make it very cold I'm going to use liquid nitrogen so this is nitrogen which is at a temperature of minus 196 degrees centigrade so nitrogen as you know is about four-fifths of the air we breathe is nitrogen we cool it down to minus 196 degrees it becomes a liquid and I'm now going to pour some of this liquid onto the ceramic material so what you can see there is the nitrogen boiling away at minus 196 degrees and it's boiling because it's taking the heat out of the superconductor and cooling it down so when that stops boiling it will have reached a temperature of minus 196 degrees okay that's going to take just a minute or two so while we're waiting for that there's another demonstration here which contains an identical piece of that ceramic material so again I'm just going to cover this in liquid nitrogen and and we'll wait for that to cool down that's going to take a moment or two so just to fill in the time I thought we'd just have a little bit of fun with some liquid nitrogen and it's quite a fun demonstration what I have here is some hot water so in the pour some hot water that's near near boiling water and I'm now going to pour liquid nitrogen at minus 196 degrees on to the water we should about plus 80 degrees and this is what it looks like this is my mad scientist impression okay so these are still boiling away but in a minute or so these species of superconductor will have cooled down to minus 196 degrees at which point they will have lost all their electrical resistance which means if an electrical current flows through them it will flow forever now one of the properties of a piece of material has no electrical resistance is that it repels magnetic field if you try to push a magnetic field into a piece of superconductor the changing magnetic field induces a little currents inside the superconductor and those currents produce a magnetic field themselves which cancels out the field that you're trying to push in and because those currents flow forever because there's no electrical resistance the magnetic field can't get inside so a piece of superconductor will repel magnetic field now the usual sort of demonstration that is to take a piece of superconductor and then take a little magnet and show that you can levitate the magnet above the superconductor is very pretty demonstration but it's all about this big okay so what we've done is we turned it upside down to make a much bigger demonstration so this is a ring of steel and glued to the surface are lots of powerful little magnets arranged north south north south and so we should be able to take this piece of superconductor and levitate it above the magnetic field so hopefully this is cooled down enough let's see if this works you get a close-up shot okay so that's one of about 20 or so different physical phenomena that are being explored as a way to produce a large-scale quantum computer and sometime perhaps the next 10 or 20 years somebody will actually succeed in building a quantum computer with hundreds of quantum bits and they will then be able to solve new kinds of computation new kinds of calculation that will never be possible with a conventional computer so I'm pretty much done I've got one final demonstration before I do I just want to leave you with one thought about technology and about computers which is this we're all very familiar with technology we use gadgets everyday we use computers everyday take a moment though to think about the science that makes all this possible these gadgets don't just fall out of the sky they're made possible because of the mathematics because of the physics because of the computer science the idea is the intellectual ideas that actually make these technologies possible and that field of computer science is a very challenging and a fascinating intellectual discipline and it's one which in many respects is still in its infancy so I think many of the most exciting developments in computer science whether it be the intersection with with with biology and DNA and trying to cure cancer whether it's exotic quantum effects like superconductivity and trying to build a quantum computer I think many of those interesting development still lie in the future so I'll leave you with that thought and also with this final demonstration which makes use of another property of superconductivity so I said that the superconductor repels magnetic field and indeed you've seen that it does but this is what's called a type 2 super conductor and a type 2 superconductor does something else if a magnetic field is passing through the superconductor when it becomes superconducting it traps the field so the difference with this demonstration is that over here I cool the material down till it became superconducting and then I placed it on the magnetic field this tin can has a powerful magnet attached to the top and the magnetic field was already passing through the superconductor before I added the liquid nitrogen so hopefully what's happened is that magnetic field has become trapped in the superconductor now if that's happened what I should be able to do is to lower this support if I'm lucky there we go that's super conductivity okay thank you very much thank you very much I think we've we've been royally entertained and royally educated bye-bye Chris I think we should let him off now and I'll call on Tandy L to present the prize from the patterns of the questions that have been put I have my doubts as to whether many of you misuse your Saturday night's by watching Camelot Merlin and King Arthur because I have news for Merlin that he's not as formidable of magicians as Chris Bishop Chris I couldn't help wondering what your predecessors of yesteryear Christopher Longet Higgins my friend Donald Michie Robin Noma who still comes back and indeed though the present vice-chancellor when he was a young lecturer Tim O'Shea would have made of I think the exponential growth in magic it was a perforce I'm also curious as to what your blood from the discipline of giving those astonishing Christmas lectures I think all the saw realized that they must been well sculpted and there was one heck of a lot of work behind them well this is a very formidable contribution to the public understanding of SARS and as a general conclusion I think it's pretty justified in saying that the acceptance by a majority of people of what is happening this fortnight in Copenhagen would not have had the public support that is crucial had it not been for the efforts of very many people to promote the public understanding of science your contribution has been huge but it is a great privilege to produce not only a gift from the University but also a medal it's a beautiful medal I dread to think of the alchemy because objective Chris there you are thank you very much indeed this production is copyright the University of Edinburgh

Channel: The University of Edinburgh

Views: 108,901

Rating: 4.8843284 out of 5

Keywords: chris, bishop, future, computers, tam, dalyell, prize, lecture, informatics, edinburgh, university

Id: Iqg90Cq5Xfc

Length: 51min 47sec (3107 seconds)

Published: Fri Dec 18 2009