Preparing Today For Tomorrow's Quantum Future

would you yeah so I'm just going to try to go through this uh quickly and you know hopefully have time for questions so bear with me uh let's just get started so first of all you should know that there is actually a vision for Quantum across Europe and there is something called a Quantum Flagship which is a billion Euro program that was set up by the European commission uh to further all things that are quantum Technologies and the reason that this is special is because there's only three flagships in the EU the first was the graphine flagship to figure out what the hell to do with this mystery magical material graphine the second was the

human brain Flagship also a billion euros and then we've got the quantum Flagship because Europe foundationally believes that this will lead to us keeping Prosperity this time in Europe when we create you know cool uh advancements in science research and engineering and like the area is divided into four categories and I have lost slides oh my goodness I thought I was going to be left out of the whole slide drama but yes um so uh what happened is we have these four areas so imagine these you know four things the first what you're going to see is a quantum computer uh there and you know that's obviously the thing that we talk about the most but

that is not the only thing when we talk about Quantum so you have a quantum computer but you also have Quantum communication you have Quantum simulation and then you have sensing and mology to do more specific measurements and all kinds of cool stuff and in order for us to actually have this this Vision you know this promise of all this money and intellectual property and retention of patents and all this other cool stuff for Europe there's a lot of underlying work that needs to happen in addition to that as well and that's like all the things that we need in order to have these systems actually work with engineering and control the foundational science that happens underneath that and

also Theory algorithms and protocols so in short this is a huge program involving pretty much every single Europe every single Europe every single University across Europe so there's a lot of people working on this um this is genuinely you know not um a dash on the pen because we've been doing it for a while and the ambition is to get here which is a sort of Quantum Internet which is connecting everything to everything else but at the quantum technological level so I just told you about these other Technologies right imagine a network that is running on Quantum communication that is connecting quantum computers to each other that also enables this sensing and Metrology to work together and do Quantum

simulation as well that's what we're talking about and the kind of Promise of some of these Technologies in sensing Metrology is like to have a quantum gravity meter where you can actually imagine that shell will use it to to find oil or do you need something yes pleas do you want to hug no yeah he um Sor for the in does anyone know Andrew cols in the room all is it cor has C RT yes CTF CTF thank you very much thanks sorry y okay not a hug uh thanks all right thank you I'll hold you to that so um let's talk about what kind of problems can you solve with a quantum computer because this is actually the

thing that everybody wants to know right what's what can you actually do like some people will argue that for a very long time we've been seeing Mo's law in effect right we you guys know what Mo's law is everybody anybody not know what Mo's law is okay good um so that you've been seeing M's law for a very long time but what we're seeing is the emergence of Mall's law which is like that's nice you can keep building your little processors but you're not actually achieving greater computing power in fact it's declining so because of this emergence of andel's law there's this notion that we actually need a fundamentally different architecture in order to build our

future systems which can actually do you know the large data set problems that we're giving them and do the needle in the haast stack problems um and the kinds of applications that are being imagined are things like doing protein mapping and folding stuff that we already have ai systems at like deep mind doing but also earlier detection of cancer and then when you look at beyond that the hype cycle part of this is gigantic and in a lot of ways Quantum resembles things that are happening in AI which is that there's a huge amount of promise and what we had in AI is we had the AI winter where you had this huge promise investors put a shitload of

money in there and then you had this little Decline and the they called it the AI winter and then they're emerging back from that winter you know thanks to llm uh and chat GPT we see a similar type of Phenomenon with Quantum and that means that companies are also pouring again shitloads of money uh collectively into this so Microsoft IBM Google everybody is working on a quantum computer everybody is competing like yours is not a real quantum computer mine is it's that kind of disagreement that we're getting into and it's a complete rap as to who will get where first and is that good enough for the rest so you might remember that Google claimed

Quantum Supremacy and then Microsoft was like really is that really Supremacy there are plenty of memes you know that you can make just on that discussion alone uh IBM's development road map is impressive just for one reason this is a project plan that actually meets its Milestones so every Milestone that they have promised they have metant so earlier this year they released Condor with an amount of cubits and what's interesting about this is it's like this expectation of exponential expansion so once you get to a particular threshold with the amount of cubits that you have then uh you will be able to like incrementally increase you might be thinking wait a minute what the [ __ ] is

a cubit it brings me to my next slide so uh the biggest property of a quantum computer is that we stop talking about bids and we start talking about things that are called cubits and a cubit is BAS basically a Quantum bit and you know you might have heard people say that it's basically a zero and one at the same time think of it more abstractly it's basically a place where you can run computation so instead of having a single space to run a computation you can basically hold that in two spaces and they're just being held there virtually you know and uh while you have this like virtual hold you actually have this potential space for doing all kinds

of cool stuff so uh when you take those cubits and you do this funky thing that Quantum lets us do which is entanglement which I think actually is the most romantic property of quantum which basically means that you have two uh particles uh that are coupled together and regardless of the distance this is what every girl wants regardless of the distance right something that happens to one of them is immediately felt by the other right I know you didn't expect this to get mushy but here we are so uh and entanglement is super funky because this is the thing that actually gives a quantum computer scale because when you entangle these cubits right and they

become coupled in this way maximally coupled you're basically creating more States let's do some math who's ready for that ready okay so let's do two to the end states which is the amount of possibility spaces you have to do your computation so let's start easy what's 2 to the second power right two to the thir power is I heard a six shame on you okay and then two to the fourth is thank you so you get the idea we have this doubling okay all right so here's a notion this is why qu computers are so amazingly funky from this perspective so if I needed if I had the biggest supercomputer on earth right where is

the biggest supercomputer on Earth anybody know come on guys China probably so um you know you have the biggest supercomputer on Earth if I like want to double it's Computing capacity what do I need to do no dude stop with the punchline no that's what I get to say no look so if you want the biggest supercomputer on Earth that's currently in China what do you need to do you need to build another supercomputer if you want to double it if you want to double a quantum computer what do you do now you go thank you you had a cub it okay so notion is pretty darn powerful all we need to do is add

one more cubic to exponentially expand our Computing potential space that becomes interesting okay so this is why everyone is like slathering over the potential possibilities of doing this because right now our current computers don't do certain things well for example the properties uh around Quantum that make it so interesting is are these fragility no cloning principles because when you are existing in these Quantum States any perturbation right any observation or noise or someone trying to look at these like Quantum particles actually means that the whole thing collapses it doesn't like it it's finicky it's it's very temperamental and the quantum state will collapse so they don't do well with trying to do a replay attack all right or if you're trying to

observe the traffic in motion they don't do well so this becomes interesting from an information security perspective and now when we want to talk about because so all nice I get it but why are we actually worried about this as a threat to cryptography well it's because all of our modern cryptography especially our publicy cryptography is based on some foundational problems right so we always talk about Alice and Bob our friends right Alice and Bob they basically exchange communication with each other using cryptography which underlies two foundational math problems integer factorization and discret log integer factorization basically means I've got an integer right a number a big number and I want to figure out what its

constituent parts are made out of and let me explain something to you like so uh can I pick on someone who can I pick on just I feel like I need to pick on colleagues Johnny what is I know he's dying now it's so good Johnny what is um I don't know 11 * 10 [Music] six this is going in his performance review so um no but what's 11 * 10 D you can do this 110 thank you so okay so you got that right so 11 * 10 is 10 that was not so difficult except for the six um the idea is our brains work really well when I ask you to multiply something the

the person in over here so if I ask you um what's 9 * 82 okay what are all the factors of 72 what are all the numbers you can multiply together to get 72 all right his six was legitimate because our brains work the same way that our computers do when we need to multiply forward we can do that quickly but so can the computer when we need to reverse that oh damn it when we need to reverse that uh oneway function our brain struggles but so do our modern computers that's integer factorization a quantum computer however does not struggle so they can do this stuff really well and more importantly really fast the thing

is discret log which is basically the clock arithmetic you see on the bottom if I give you the Formula 3 6 for the modulus which is basically a remainder of 17 you get the answer of one but if I only show you the clock right you're not going to be able to derive the formula that was used to create it is that notion clear all right now those two problems are foundational in our cryptography a quantum computer can reverse it thanks to Peter Shore and love Grover not this Grover Grover love Grover and love Grover basically figured out how to do these calculations while trying multiple possibilities at the same time and this optimization of this

means that the time you need to reverse these oneway functions where because sh already had an algorithm before there was ever a notion of the Quan computer that was going to be viable for us to build and work on sure already had an algorithm that we could potentially use that we could run if we had such a computer and Grover had this potential to make this optimized thing that would try all these possibilities which would mean that we could calculate things quicker is this notion clear this is why it's a foundational threat to uh our current cryptography so look you guys know if you've been here today you know that even if you would have great

cryptography you still need good Hardware operating systems protocols applications and you guys all know how good we are at that right so assuming that you have this bulletproof layer of cryptography if you just add that you're fine is [ __ ] it's complete [ __ ] you need to fix all the other underlying stuff because no intelligent attacker is going to try to attack your bulletproof crypto if all this other stuff is broken and even if you do all this stuff well magically and you're a really hard target uh then you still need to understand what it is exactly you're trying to protect and from whom because even then the potential that they have even greater resources to attack you

that is still there and when we look at when it really depends it depends on what that attacker has at their disposal so what you're looking at is um two axes on the x- axis you're looking at the time to factor an Elevate number so like let's just that's the length of a number that you're trying to factor and then this bottom uh axis is the amount of bits key bits you have have at your disposal so if you take a look at the N recommended key length for RSA which is right here this dot of line in the middle what you're seeing is that operations that would have taken potentially a billion years complete

will be reduced uh to a few seconds or a second or under a second the more cubits you would have at your disposal yeah so when is it dependent on what you saw me show you with the IBM road map when do we get to this point where we have hundreds of thousands or even millions of cubits and we have this viable quantum computer that's actually capable of breaking RSA because of the amount of computations it's able to perform is this notion clear all right so there's another problem it's not just that we need to wait around for a quantum computer the other reason is and I will tell you the honest truth we're already [ __ ] and that is because we

have a foundational issue in this capture now decrypt later thing this is a photo of the NASA data center facility in Utah I think it's actually really pretty but this facility is you know known for ages to make snapshots of the internet specifically of encrypted Communications and look if you're just going to capture encrypted Communications and store it until you do have a viable means to decrypted I think that is already an interesting attack Vector just an in of itself and you take a look at okay so there's a quantum computer you showed me RSA what about everything else well that's the problem because even with symmetric Keys we worry about things like okay even with

AES you know you'll still need to take the largest key length you have so you'll still need 256bit keys and we're still worried about key exchange mechanisms the Cs and every single uh public key algorithm that we have is all Al at risk so bottom line there is a hell of a lot of impact which means we need to have a plan to do something else than do what we currently do and when you want to ask yourself what does that mean to my Enterprise or company or country well I'd like to start with asking three questions first of all how long is it that you need to keep that stuff secure that you're using your

encryption for today does anybody he worth for healthcare 1 years 13 years in healthcare wow hats off to you um and like obviously like healthc care records for me would be the time that you would need to have the longest amount of preservation let's say that you would theoretically need to keep Healthcare records safe for a human lifetime I have to tell you we don't know how to do that yet if we would have genetic information if you were 23 in me and you had sequence everybody's DNA it's not just about one lifetime it's about potentially like having information that would be interesting for their progyny so multiple lifetimes we don't know how to do that yet so that's the first

question is how long do you need to keep your data secure the second question is how long before there is a viable quantum computer that breaks our secrets so I used to be the member of this Quantum Flagship I was the vice chair of the EU Flagship the the billion thingy and I have to tell you that speaking to people who are actually building these things that the uh Outlook and this is already ready back in uh 2020 was 10 years so by 20 30 we will have a viable quantum computer that was the Outlook you know and the question is will it be in the public domain is a different question that's a national security

question but we will have one and then the third question is how long do we this group not only in this room but in every room like this all across the planet need to work on having a transition plan from our current systems to one that is quantum safe how long do we need let's take a look at how we did with ipv4 to IPv6 transition that's an indication so like what we really should do is think incredibly pragmatically about our Network architectures and think about okay let's be smart you know we have the advice we already had it from the NSA sweet B stuff you know back in 201 what was it 15 2014 where they

said already NSA said to the US government everybody yoyo increase your key length for all of your current encryption algorithms they already said that we can already do that now so if you're using crypto in your network make sure you're using the maximum key length afforded by that algorithm the second thing is look for options there might be very specific places where you can do something called Quantum key distribution this is not ubiquitous currently but there might be some specific opportunities and finally what I do think is the best long-term bet is investigate postquantum algorithms even though there haven't been officially announced by nist there are options available to all of us for this so um

let's explain qkd in a minute we you know go back to our friends Alice and Bob Alice and Bob are in a relationship unfortunately Eve the jealous ex-girlfriend wants to know what Bob's doing so as a result of which Eve wants to spy on Alice and Bob's communication and what she tries to do is you know hook up to their Quantum Channel and what did we say about no cloning and fragility and replay taex right if Eve is trying to spy again on Bob Alice and Bob no it's kind of sad for Eve right so um the basic way that it works is really simple um we basically have you guys know what polarizers are right we have

polarized sunglasses so it basically lets light in one way and doesn't let it out another way it's usually when you rotate at a specific angle roughly 90° so you rot it at rotate at a 90° angle light uh does get let in or does not let get in so what you do is you have a configuration of polarizers diagonal and horizontal you set them up in a certain way and that configuration is relayed from Alice to Bob over a secondary Channel there are some flaws already here this sounds like typical key exchange issues there are typical key exchange issues here um and then there's a single Photon emitter on one end from Alice it's emitting a single Photon it

goes through Alice's uh polarizers if Eve is in the middle of this that perturbation occurs and then Bob gets gibberish that's how they know they can detect for the presence of Eve is this notion clear all right so that's qkd in a nutshell and there's all these different flavors of qkd that you can Implement you know on production telone networks uh the favorite one is MDI qkd which is measurement device independent but there's some really cool stuff happening because how many of you have dedicated fiber between you and everyone you communicate with no one and that's the point of free space qkd so the idea is because we don't all have dedicated fiber between Alice and Bob which you might have by

the way for a primary and secondary data center or some backup facility freespace qkd says no I can actually transmit these signals with a very powerful light emitter just over free space so this was one of the first experiments that happened between larot and tenf in uh Europe where the pioneers of everything free space um and this showed that over like a a length of 144 km this also worked problem is the Chinese did it bigger and better and in space so in 2016 they launched the first uh Quantum communication satellite uh this professor panan we he was uh a student in Austria uh and he then managed to make a secure video conference call between him and the

University of inbrook um in China I told you about my billion Euro Flagship which we're super proud of in Europe right that's a billion EUR over 10 years China every year on Quantum Technologies spends 10 billion EUR per year we're not going to win on the basis of money ball right it's if we if there was a race based on who had the most money we're not going to win the race right 1 billion over 10 years versus 10 billion over every year and all these commercial companies because there's always a sort of government stake in it they have all invested into this and now this network you know initially it was only a 2,000

km Network now it's a 10,000 km Network it covers like most of China uh you know a qkd network so really fiber cable between all of the major critical infrastructure and satellite connectivity on qkd so misus is the name of the Chinese satellite and what you see is that it actually like you know how Co did more for Innovation uh and digitalization than like every CIO ever that's kind of like what um missus did for the spur uh to do things in terms of space based qkd and so as a result of which you had a bunch of projects all over the world that were actually trying to figure out how we could do more with space based

qkd and free space Quantum key distribution and there are some real options here like there you could use this legitimately if just as a bridge between terrestrial repeaters you could have additional gains for confidentiality you know the usual you could actually use this to mitigate some risks that you would have in terms of uh space based issues and you know still it's about finding those use cases one of the things that I did when I was he at kpn which is a Telco is we found critical infrastructure projects for banks that could actually make use of this space based qkd to actually further their regular goals but you could also use it for like sharing infrastructure

so vone I don't know if you guys know this totally esoteric subject but vone is putting a 5G Network on the moon because you can um and one of the real options to do this is if you're going to share infrastructure like the early days of the internet or the early days of telecommunications one of the most important things you need it for is to do interconnections and billing and that kind of stuff and so it be a really good use case for that too but it's sometimes feels like a solution looking for problems rather than the other way around uh and there is a a lot of moving Parts in terms of resilience it's hard

to fix stuff in space crypto agility is still the name of the game what if your algorithm is broken you know you need to swap that out and then finally supply chain security for all of this is really difficult and this above image is of the Kesler effect which if I don't know if you remember but uh there was an issue where Russia had destroyed a satellite and then subsequently little pieces of that satellite then exponentially attacked other pieces of infrastructure which then attacked other pieces So eventually threat in the lives of the astronauts on board the ISS so there's a lot of moving Parts literally uh in trying to get this stuff right um and one of the things that we

try to do terrestrially we built in the Netherlands a Quantum internet backbone and this is already a couple years old when we initially started but the reason that it's the case is because we started initially doing this in 2016 with two locations dedicated fiber uh commercial equipment and then what it really looks like under the hood when you try to do this at scale when you want ubiquity it's way more complex this is the other reason that I prefer a postquantum approach so as you all know nist started with a project in 2016 it'll be uh finalized by next year in 2024 um you know and the most of it will be finalized it's going to be a

continuous iterative effort and then eventually like you know we know that we're going to need 5 to 15 years to get everybody on track and migrated to the new solution so you might have seen the round four finalists so we had a bunch for um key change mechanisms and encryption so it's mostly uh and I'm really happy to say this because it was a professor that we worked with Peter schwaba who uh worked on crystals and kyber uh and dilithium so those are accepted by nist and we still have a bunch of Alternatives you might have seen the news about psych uh anyone see the news about psych Okay so you don't always need a quantum

computer to break a Quantum algorithm and that was what was proven by psych so someone took pretty much an old computer and found a mathematical weakness in pych and this is actually the kind of thing this is why we need more eyeballs on this problem because the what can go wrong here is still significant you know um we still have companies that think that they will find their own Quantum algorithm for this this is the worst thing we can do there is no security and obscurity we can keep saying it but do not roll your own crypto there are still attacks on our current crypto which involve Crypt analysis and trophic analysis and you know looking for uh

faults in entropy and these same problems also plague any postquantum algorithms so really there's always stuff that can go wrong the best thing we can do is kick the [ __ ] around a lot more and watch it break before we're all deploying it globally um and even when we do that we're still going to have to balance equities between you know governments and their arguments to either weaken or ban this cryptography altogether so that requires a security Community to kind of stay on par and figure out how to make sure that we're still doing the right thing uh when necessary and not saying oh yeah sure we'll enable the Ghost Protocol and WhatsApp and you can have

everything you need we need to make sure that doesn't happen in the quantum ra as well and we're not deliberately weaking these algorithms like we did when we introduced algorithms for mobile communication um and in terms of cryp postquantum crypto the biggest thing is and this is a question that I genuinely have to you someone earlier today asked who knows all of the things on their Network right that's a good question who knows all of the crypto you currently use in your Enterprise how who knows any of it no seriously how much even if we have asset management and inventory management how many of us have any concept of the current cryptography that we use across

our Enterprise any of it like we don't know we assume that it will work somehow magically because it needs to talk to someone else and that's not true we need to actually start with an inventory of crypto assets we need to figure out how we like get it ready for this implementation we need to look for opportunities for crypto agility so that if stuff breaks if one algorithm winds up breaking through this process that we can swap it out with something else else because we can't swap these algorithms arbitrarily for what we have now we to have policies and hardware and software vendor stuff and all the usual blah blah compliancy stuff so the biggest thing is

we need to do this start failing early because the road map in order to get us to Readiness and I I will give you my slides so you know please don't feel think but um with starts with awareness we need to figure out what Discovery start implementing and test test again and then finally make sure we have a backup plan and monitoring and to look at these algorithms and obviously I'm not asking you all to run out and go start your Quantum Journey if you're not doing the basics right so if you don't know your own network if you don't know your vendor elements you're not doing any third party managing if you don't do

anything around vulnerability management programs or understand your data security needs start there but what I hate about our industry is we wait for stuff to get urgent that is important that we knew ages ago we know stuff is going to be end of life how many of you have zero end of life in your networks we we all do so I I hate the fact that we don't fix things that we could have fixed early before they turn into you know a triple fire alarm and then everyone's running because we enjoy the crisis uh when it comes to vendors I think it starts about asking the right questions so engage those things again you can uh get these but I think it

starts by saying what are you doing about this what are your plans for this transition you know how are you going to uh work with your own vendors and there are already products out there that you can play with now first of all you can already get an account on the IBM quantum computer kkit which is the programming language is all based on python uh there's a lot of other demonstrator uh quantum computers because they do want to have cloud computing available with Quantum that's one of the major goals is accessibility so there's a lot of places where you can already start playing but there's also a lot of places where you can also look at

other products so op SSH is working on Quantum stuff M that VPN is using that there's code Crypt there's all these postquantum companies and then there's of course the open Quantum safe uh Alliance which is a whole resource uh for all of you to just kind of get started but I think this is the most important thing I can say is that you just need to get started and also when we look at issues that will inevitably occur like they did to psych that was broken uh that we actually need to be able to evaluate that information correctly because some of it is also equally overhyped there were a reports of other algorithms being broken they

are not correct but it's only because when there are um uh Power uh analysis attacks they can break multiple things not just the one algorithm that was reported and in conclusion and I appreciate you guys listening to all of this on Quantum after doing math after five it's a huge accomplishment I want to remind you that it's a journey and you just need to get started these are uh two reports that uh I worked on uh as part of the world economic Forum work and there for your like you know yourself to take a look at uh where is the current state of quantum Computing and what kinds of things can we do to transition they're also for your CIO CTO

whatever C layer to just you know lay on their desk and let them see so that you actually have the funding to go ahead and do this make sure that you organ and find your friends so your ISAC Community is a potential source of support so is your shirt community and please if I could ask you one thing is to get your hands dirty and then talk about it we need way more cooperation and communication here because we are not collaborating because of all the National Security interest in this space so that's something we really need thank you so much for your attention thank you

[Applause] questions or just beer yes yeah um with the kqd thing does that mean that if it comes out as a noise if there the middle is that new kind ofing yeah actually one of the initial um attacks that was done was um you know so it's a fiber optic cable with specific color lights so one of the initial attacks was Eve was just being horrible and just shining a bright light into the thing so everything was gone so that was just a yeah you can absolutely deu that but then you have to be you know when that dark fiber is being in the middle you have to F either sort of dark fiber vampire attack or actually

physically have an Intercept in the middle which will mean that from day zero you'll notice it need there yeah oh by the way I should say why the light it wasn't just e being like in a horrible mood that's so you could read uh the configuration of the detectors so remember there's polarizers on the other side and the detectors on the other side so in order to understand which configuration was communicated by Shining Light you can try to get some partial information from how those detectors are figure yeah oh yeah fig we you know [Music] we don't know that that's the whole point so if you remember there was um oh gosh uh so look

I my favorite example is a historical one so you remember the Enigma machine right everybody remembers Alan ching and okay so who can tell me how long like so we we used Enigma we you know had the War uh then subsequently what happened to those machines what do we do as as like use this crypto it's really good and then meanwhile you could read all the communications for roughly 30 years after the war okay what do you think we're going to do if somebody suddenly finds a way to have a computer that works to read everybody else's secret they're not going to run out and be like hey I can see it stop communicating that they're not going to

tell you this is my point about like go out get your hands dirty and tell everyone yeah it sounds like they always say know the quity in five years and then again in five years and then and then we don't even have the algor than scale it's really hard to justify the investment yeah yeah it's going to be there but then has yeah don't forget the algorithms by the way they're all already there and I want to I want to warn you when the algorithms are there it is not a drop in rep place there are bandwidth issues there's optimization issues there's uh CPU utilization you can't pop it on an iot device just like that so what nist

is trying to do in fact n is need okay so we had we are round four with uh the encryption algorithms but they have laun yet another call for new digital signatures because they want more stuff that is not lattice based and also has better optimization uh for like some of these considerations so we're not there there so I don't want us to wait till there's like this perfect moment the N then says yes you know here sh Thou shalt have the Holy things because even then stuff can get broken and replaced we need like a sort of baking time we need the toothpick to keep checking the middle of the BR so uh this is a terrible analogy but you

get where I'm going with this we need to kind of continuously be able to check and verify that these things work and are properly suited for the problem we're using them for this is why I would encourage you we already have a set I gave you that Set n has already released those go ahead and play yeah anyone else well thank you again it's been a pleasure thank you