on to quote, to reduce the suffering, hey hydra, to dive in. other states rupture the leaving over $300000.00 without the electricity. while a lot of say land slides and flooding, devastated this highways. those little vigils indicated that it's a rental range of the true then why high wouldn't colin's a humidity car though from the amazon. i'm an atmospheric blockage close by feet waves. the lines are still have to go monday. i don't expect to hit the link until sunday while that loads up this news hour i've seen read time, sees on the underground. that's up next. we'll be back in about 30 seconds. we'll see what the the

i'm action or time see, and welcome back to going underground in the final show of this season after we've witness genocide in this region to the eyes of scholars, politicians and journalists couldn't algorithm help the world of its moral catastrophe. the mathematical procedure shows algorithm, promises new age solutions to geopolitics, hunger, disease, climate change, even if it made me the door open for cyber attack on the world's nuclear weapon systems. naturally laid. capitalism is investing in it because it leads can honestly, algorithm for the power as physical financial services industry to create will beating investment strategies. political groups meanwhile, could use it to break the encryption on global bank transactions. the formula is shows algorithm its creator is the professor professor of applied mathematics and m i t p to show he joins me now from cambridge, massachusetts, in the usa,

a progressive. sure, thanks so much for coming on. alarming that the, the entire financial system relying on electronic bank payment transactions, nuclear codes, c i a file encryption, ms files. they might be good though for mid sin and yours for pandemic sin diseases. a lot of a lot of stuff is invested in sure was algorithm named off to you. i added simplest . what is yours algorithm? so source algorithm is a way of factoring large numbers using a quantum computer. so quantum computers are these hypothetical devices, which so far are only you only have very preliminary prototypes that show that they will work, but the prototypes are so small, they can't actually do anything useful yet. but if you know,

if they keep on growing the way they have been in the last 20 years, probably in the next 10 or 20 years, they will be able to do useful things. and what shows algorithm is actually, it's a way of factoring large numbers, which needs quantum computers is no rush. i claimed it as a 50 cubit the quantum computer. by the end of the year, there was a russian $116.00 at the end of 2023. ibm says they've got a 1121 cubic cubic being a unit, a like binary for the old for your laptops, binary switches inside the inside your laptop. explain exactly what the quantum computer is because most people think of their own computers. switching on and off and it's loads of these switches would simulate sums that make graphics and games and all the other things that you do on your computer. yeah, so uh,

a to bit is, uh, well, it's basically a quantum binary switch, which means either it's 0 or one. the think about cubits as they know if they have a classical bed, it's either 0 or it's one. and there is no in between or cubits. uh, cubic can be in a superposition of both 0 and $1.00 at the same time. so it's going to sub have the option. well, it's more than a 3rd option. that's a whole continuum of options. and you know, at 1st of all, you say, well, what's the use of it being and supervision, and 01 the same time. what can you do with it? well, the interesting thing is that there are algorithms you can do using cubits that you cannot do on a single computer. and one of these is the one i discovered cause choice algorithm

. and it's for factoring large numbers and 2 primes. ok, so we may have to go through another stage before we fully get it to your algorithm . then because the other day microsoft was saying, you figured out a way of error correction in these quantum computers. because if you're going to get a continuum rather than a discrete 0 and a one of the whole point of a laptops is they have systems in there because there's always a mistake going to happen. and there's a way of correcting it. i thought the whole point of quantum mechanics and you might have to explain that to split experiments as well is the when you identify a particle, you change where it's at. so in your a quantum computer, when you find out where the switch is, you all to where the switch is. yeah,

that's right. so um, when the quantum computers 1st came out, um, you know, the general consensus was that you would never be able to build these things because they were impossible to hit correct arrows on. and no quantum mechanics is such a delicate 3 that there is no way of building a quantum computer that will do gates that are absolutely error free. i mean, we still think this true that you can never build a quantum computer with gates that are free, but what's different is that we figured out a way of correcting the errors. so quantum mechanics, if you measure something, you change it state. so if you major at the state of a bottom, peter, in the middle of a computation, the state will collapse and then the computation will no longer be valid, you valid. so how can you measure errors without correcting power?

and if you don't measure errors, how can you possibly know whether you have an error and correct? so the solution to this was to come up with a way of encoding the state of a quantum computer so that you can measure the error without major in the state that the computer is actually. yeah. and then once you've done that, you know what the error is, you have a distributed computation because you didn't measure the state of the computation, you just measured the state of the error. so now you can correct the error i continue on is usually error related to where this is. well, is not if you do it right, if you do it right, there's a state of computation which is protected by, that's fine, them aircraft and code. and that are the errors which apply to the cubits, but do not change the state of the want them computer. now, if you have

a big enough, there are. so suppose, you know, suppose you have an error correcting code, you can have an error correcting coast allowed to do to correct safe by the errors at once. but if you have one of these and you get 7 errors, then that is going to change the state of your one computer. but if you have a 5 error code and you only get 5 errors, then you're fine. you're correct. there's and keep on with the computation. okay, well back to the algorithm, then we're not advising people who have access to nuclear weapon systems to investigate, to show us how good them, uh, with respect to them. but at the moment, with normal computers, you can't break into the c, i a or whatever, because of cryptography. why? what is, what are the reasons why you can't break into a code,

say that edward snowden released to the world, that we were all being bugged, or whatever he had access to the documents. to break into those judgments, you'd have to require the length of the life of the universe. so something to be able to break into it. why is that cryptography? it's so strong. but why is it's a week in the face of your algorithm? so i guess before around 1975 or so if you know, if you had, if you wanted to predict the secret and you wanted to communicate by so good code to somebody else you needed to have gotten together with the other person and exchange the secret key and if you both have the secret key, then you can, one person can encode the message using the secret key and the other person can decode it. it's what, what side of it and telegram and facebook messenger and all the social media apps

used today is that, i think, yeah, but, but they, um, but you know, it's, it's really cumbersome and awkward and expensive to exchange secret keyes, but everyone you're communicating with. so 3 people, i guess we're best should, we're an edelman. i figured out how to exchange messages without using secret keyes that both people know and these are called public key crypto systems. and they are basically the basis of most communication today. i don't know that better they're used for protecting nuclear secrets and things like that. but i know banks how many um, you know, why are they so secure this process?

okay, so there's so secure because if you suppose you have to prime numbers like 11 and the 17, i can take eleven's and 17 and multiply them together and get $187.00. but suppose i give you 100 g, so it's much more complicated to take 187 and back to it and learn it. so 11 times 17, actually with 187. it's not so complicated. but if you had a 100 digit product of primes, it's much more complicated had if you have a 300 digit product of primes, the fastest computers in the world don't know how to do it efficiently. so and what the rest should be or in hillman did was they came up with a code so that in order to encode something,

you can multiply 2 number 2 prime numbers together and get their product and use that as the basis of the criptos system. but to decrypt the system without knowing the 2 factors that gave rise to this prime, you need to factor with this large number into primes, and that is basically impossible for computers today. so this is why our as a is use so much now i show that if you have a quantum computer, you could factor numbers much more efficiently than on a classical computer. so the number of steps on a quantum computer port back bring a large number is essentially not much more than the number of steps it takes to take 2 numbers to get to take 2 numbers and modify them together to get this large number. so that means that we hope you have upon them computers

as a public key cryptography is no longer secure. and before anyone thinks that this is just a break into a secret codes, even if it is in the public interest. the factoring in the primes of big numbers is an important concept in the whole of life on a, even genomes, dna proteins and mentions. i think you got that a little bit wrong. so factoring large numbers into primes is really, is not very important. the scheme of things. what is important is that quantum computers can solve one of mechanics, problems very quickly. and you may not think that's important. but when you realize

that, you know, chemistry is a quantum mechanical problem. designing drugs is a quantum mechanical problem. building new superconductors that start conducting a larger temperatures than ever before is a quantum mechanical problem. and if you could solve those quantum mechanical problems, you could really make great strides in our knowledge and useful um technology and other things provides a visual. i'll stop you there. more from the award winning computers understand divisor of shows algorithm off of this, right? the

the the, [000:00:00;00] the the welcome back to going underground. i'm still here with the award winning can be designed as an advisor, show us albert and professor peter shore. present. sure we were talking about,

or you correcting me about the fact that it's going to mechanics not too much was algorithm. and the factoring a prime is that good solver. so many problems in health care and who knows? who knows what else of funding on basic science in your country, the united states is the lowest, isn't the lowest for 2 decades, presumably in favor weapons hunting or something. is that because policy makers don't understand the potential of quantum mechanics and they don't understand what possibilities quoted mechanics offers or they're trying to cover it up. know, i'm sure they know trying to cover though, is it very expensive to research? actually, quantum computing has a ton of money for it right now it's and that of course is presumably because it's useful for um, one of mechanics is also useful for cryptography. what doesn't have

a lot of money is that more basic basic science? so i think they think of quantum computing as applied science, which is actually it's still pretty far from applied, but it gets classified that way by the policy makers. so if you're trying to, um, why don't know, understand the biological pathways and molecules that might be hard to get funding for that. of course, that is something that could profoundly affect medicine. so yeah, the policy makers are putting more money into more applied science and less money into the very basic science, which it may lead to a deficit of knowledge about basic science, which could be very, you know, and that this knowledge could really be useful many years from the future. yeah,

billions of dollars going into that into that end of it. not at the beginning and just try and explain do the labor and like myself, why or how one would write a program for a quantum computer because they probably code is out there thinking of how you write a little piece of code for a laptop. how do you, how do you write those basic computer program? harnessing quantum mechanics in a quantum computer? well, it's not so horribly different from reading programs for laptops. so what a platinum computer does is that applies. want them gates which are really um you know, i guess control tosses if this then that yeah, things like that. so to cubits. and the basic um

you know, and the basic control pulses are like, i guess, control not gates, which say if this is a one negate this other bit too bad, kind of this is a 0. leave it alone. and things like rotate this cubit to buy 90 degrees and some basis and and you can rotate things by 90 degrees because, well, because cubits are continuous so they don't just have 2 options. rotating it by $180.00 degrees would be and they're getting it so well they get by 90 agrees is the square root of negating it, which is something that doesn't exist on classical computers. so what you need to do is you need to come up with a way of giving these elementary commands that make the computer do what you want

to do. and, you know, just like classical computers, you have programming languages which take the sequence of steps that you want, but what it would be to actually do and break them up into elementary gates. and i guess the difference is that there are some constructions, one computing which don't exist in classical computing. you make it sounds so simple. why is it so difficult to construct a even a simple quantum confusion, then what are they made off silicon? well, there are at least 3 or 4 different things they're made of. so one way you can do one of the computing, so for compete for a q, but you need, uh you need a 2 state system which is either st on or off or spin up or spin

down or something like that. so in i end traps, the basic element is an ion and you use the electronic state of an ion and one state is a 0 and another state is a one. and you manipulate these states in 2 ways. one way is by shining lasers on them and that will change the state of a single ion. and then the other way is if you have 2, i ends in the high end trap. there is a vibrational mode of the ion trap were all behind. so moving back and forth at once. and what you can do is you can shine the laser to the right frequency, the couples, your electronic state of the ion, to this back and forth motion. and then couples this back and forth motion to a different ion. and that makes the 2 ions interact, and that gives you

a gate between these 2. i have these 2 cubits. so that's ion traps. and you know, there, you will have these little silicon chips and you have these little groove in them. and in the grooves, you arrange electromagnetic fields to called your ions. and another um you know, another similar method is neutral, adams, where you have a very similar to high end traps, but you have neutral animals. so they are not charged. and again, they're held in traps and you can use lasers to manipulate their quantum states. and there's a completely different method which is superconducting cubits. and there you have

a chip with some superconducting material or i guess the positive on it to sort of silicon chip with the supreme dot to mutare 0 and the super conducting material. you have little integrated circuit elements called trance minds. and a 0. there is a magnetic flux going around clockwise kind of one, there's a magnetic bucks going around the counter clockwise and manipulate these by how basically electronics. okay. well, i mean is it, it seems relatively copper and simple yet mckinsey, very controversial. i usually, consultancy firm says it only about 5000 gallons of computers by 2030. so what, why is it going to take so long to get even a basic cubit computer like the one you just described, the maybe only 10. so which is maybe you can do much with that on the market.

yeah, well, i mean, 1st, if there's only 10 cubits only a 100 cubits, or maybe even only a 1000 cubits. you really cannot do anything with your quantum computer because like a walk through thousands relate to is it right? yeah, it's not even as good as a poppy calculator. but um, because to do i mean for a while we thought that without error correction you still might be able to do something on the quantum computer. but it's looking more and more like to do anything useful not quite a bit because we do need care oppression. and 22 error correction of the quantum computer as you need to build um, i guess curve reckoning codes and find both telling me the operations on them and the best ones we know about you something like a 100 physical cubits for each logical cubit. so suppose you wanted to do a calculation with a 1000 cubits. well,

these would have to be logical cubits. so you'd need a $100000.00 visible cubits to do it. and i mean, we're haven't even gotten a 1000 cubits on a completely yet. so how many really you need and to scale up to a 100000 probably going to take at least 10 years and probably longer temperatures algorithm, it requires one uh, in the 10s of millions. i think it's right to shows algorithm. well, choice of actually there is a recent discovery that says you can do shows algorithm with fewer then cubits than we thought before. and this is no or the regular came up with this. but it still is going to take an enormous number of cubits to even to biggest inflation of choice algorithm. and

as you know, obviously the biggest are in these vendors, the pentagon over there and military. uh i m i t, uh famously uh you know, no, i'm jumps, case worker was funded by the military. ironically, a, do you think there are the dangers that the defense you're going to be using? shows algorithm together with the what do you own must because been wanting about profit driven a i a to create a little power of a tiny elite. um, well i'd be more worried about the to the just the a i then shows out with them because it's not clear yet whether i mean. so pete, a lot of people are studying want of a i. but it's not at all clear to me that quantum computers do anything to

improve a i and you know, there's lots of papers, bridges about this, but so far i haven't seen any real convincing demonstrations that why don't we do is computer is going to improve a guy, so it could take a while before this happens, just the finally will all these uh, quantum computers be able to help us solve theoretical uh, solutions, doubt, standing mathematical problems that the still haven't been solved. and they've been set, say a 100 years ago. i think it's possible that maybe one or 2 of these problems could be solved by one computer, but i think most of them are not going to be amenable to quantum computers. and you just need a very clever mathematician to come along with find a solution. but will the mathematicians in,

in their work um so the secret rather, you know, a little know and secret about quantum computers is it, for most problems, quantum computers are absolutely no help at all. so you just have to have the right problem to have it solved by a math which isn't with the quantum computer. and probably one of these outstanding open problems will be solved this way. but most of the most promising pages your thank you. thank you very much. it's been a very interesting conversation and that's the final show this season we're back very soon. but until then when we book gusting, some of your favorite episodes of the season, you still keep in touch of i will as social media is not sense in your country. and i do i channel going on the runs, you've got a normal don't come to what's new and all the episodes going on the ground, continued condolences to those very by you can us, you, i'm boning and palestine, lebanon, him and syria,

and iraq. we'll see you versus the i look forward to talking to you all that technology should work for people. a robot must obey the orders given it by human beings, except we're so shorter that conflict with the 1st law show alignment of the patient. we should be very careful about our personal intelligence at the point, obviously is to trace a trust rather than fit the barriers. i mean with the artificial intelligence, we have somebody in the, in the

a robot must protect this phone. existence was alexis the 3rd generals are indicated in tara tax against the russians. civilians are both says, 6808 says so these are already on the ground in a few things. so this is andrew looking bracelet. mary king, so so stationed in any gera as the applicant nation, 6 of big washington on size of the u. s. i'm, you've been a tardies. i'm pilots from kind of seeing them the spaces universities in the middle east or for a safe haven to students expelled from west some colleges. the speaking out against is