Is the future of computing quantum? CIFAR Virtual Talk Brilliant Minds | Bold Questions

Is the future of computing quantum? CIFAR Virtual Talk Brilliant Minds | Bold Questions

Show Video

welcome everyone to brilliant minds bold questions a five-part virtual series exploring questions being addressed by some of the world's top researchers each talk pairs speakers and moderators from across disciplines sectors and the world to explore the questions shaping our future if you are not familiar with us cfar is a canada-based global research organization that convenes extraordinary minds to address the most important questions facing science and humanity my name's mark daley and i'm vice president research at seafar it's my pleasure today to introduce you to cfar associate fellow dave bacon and cfar fellow stephanie simmons stephanie is an assistant professor at simon fraser university and the canada research chair in quantum nanoelectronics dave is vice president software at ion q and an expert on quantum error correction today we will explore the question is the future of computing quantum today's session will last 30 minutes and i'll ask our speakers questions and we'll also be taking questions from all of you joining us today throughout the session if you have a question please submit it through the q a function at the bottom of your screen let's get started first question where is quantum computing at today right at this moment what can a quantum computer do for me that a classical computer can't who should jump in so i think you just volunteered stuff yeah so um there's a lot of discussion about all the things that quantum computers are going to be able to do that classical computers cannot right so there's all these wonderful examples of worked examples of things that quantum computers are going to be able to do such as you know break cryptography but simulate quantum systems and all these there's there's a long list and if you are interested in you're on the technical side you can go to quantum algorithm zoo um search for that it gives you the big one list uh we are not at that full large scale thing in fact it's actually phenomenal that we can see that that's where it's going given where we are we're just kind of passing the hello world stage almost we're seeing we're able to do things by we i mean the community not necessarily like me but like the community has been able to demonstrate a few things that uh a few computational challenges that have not been able to be done are met uh as quickly let's say on on classical systems but i think in that context people typically forget the other cool things that quantum technologies are able to do um not just from a complexity sense like not just in terms of like this is how hard it is to compute a thing but for example quantum computers quantum devices are able to teleport information there's no there's no way that you can do that with a classical system there's all these like little neat things like you can't um copy quantum information so there's there's little things that uh quantum little things there are quant there are things that quantum technologies can do that classical systems you just never can do and so we're at this stage where it's it's going commercial i gotta pause there because i know dave has a lot to say on the matter but but basically yeah it's at this really wonderful tipping point where we're it's starting to really come alight right like this the smoldering and it's soon gonna burst into flames i love that characterization of it's smoldering okay so dave on the topic of going commercial do you want to pick up yeah i mean i didn't talk a little bit about that i think one thing that i always think is funny is we often times we hear that quantum computing is like we're sort of the fact that some people will say we're like at the bad engineer of the computer right or some people say we're just at the invention of the transistor or some people say it's more like the integrated circuit what's kind of happening is that i feel like those are usual it's very useful to look at history and especially the history of classical computers but the chart that quantum's taking will likely be different so and an example of that is you know it is true that we have very small quantum computers you know uh google's function computer is not very large ideas they're very few number of qubits right so they're not super powerful right now but there are other interesting challenges that occur in the field so for example in commercialization one of the things we have to do is we have to get all these to work together at the same time and programming that until we get that to work is a lot like the 80s actually in the very weird parallel computing architectures that people try to build um and so that's about it because we try to understand what it is but it's sort i think the thing we probably could agree on is that it's an exciting time because we're able to do these experiments now for larger systems and for more coherences and we're just in different headings to that point and places where we wouldn't expect it to be useful i think we're just getting more and more control over the common world and in that respect it's extremely exciting because it's just it seems like nothing creepy we don't see sort of our limits there um you know we may stall out at some point who knows but it's we sort of see that i mean when we talk about the actual commercialization then it gets even crazier because then of course money enters into the field and all sorts of crazy disciplines because people are you know want to make money off of this eventually right i will say that in the field i've been in now two quantum computing companies i was at google and then i'm at this company inq that does science computing i would say that well from the external point of view you do a lot of classes on internet and stuff like that internally the people are all just trying to build the damn thing and get it to work and trying to do the best thing they can right and so it's a weird sort of world where there's a public frothing at the mouth whereas internally there's a lot of like this is a really hard thing that we're trying to do and everybody involved wants to figure out how to do it and you know the best people are are looking at each other and trying to learn from each other uh and so it's really in my view i had like these two different things going on and they kind of freaked me out in some way but that's sort of the sort of the way the community works so i love that connection back to the the parallel architectures of the the 1980s you know i'm imagining the old connection machine and there was that sort of pre-cambrian explosion of different architectures and ideas and we're sort of going through that in the in the quantum realm right now yeah that's that's great okay no go ahead i mean that's why i love that analogy is because i wanted to growing up i wanted to participate in that when i was a little kid so you're getting your wish this is a human habit right like this is how disruptive technologies tend to go it goes through this this you know incubation period within academia and then there's this mass proliferation of approaches and what's going to come next right like if you can use that as a way of predicting the future what comes next is a giant consolidation event and a massive talent shortage so if you want a really nice nicely paying job get really good at quantum computing because there's a there's going to be a big talent scramble where everybody kind of when the there arrives a dominant design where people forms consensus around what the actual winning horse is gonna look like and it's it's gonna be in the next couple years i think so all the young people listening take stephanie's advice learn quantum computing every day i'm bombarded by papers press releases and honestly i can't always tell the difference between them uh about breakthroughs in in quantum computing how should i as a non-quantum specialist be evaluating these claims this is hard like the usual sort of things always occur which is you can you can try to listen to the experts who are saying things and you can hear what they're saying but even there it's always challenging because we don't really you know for example in industry we don't really know what's going on you know i know what's going on right now and so it's really hard to get an idea about what's going on the breakthrough thing is always a little bit wonky right like it it's just it's hard to say and tell you know it's nearly impossible to judge breakthroughs in this green bay weird way right and so i think almost all of them when they say breakthrough you're kind of skeptical but it could be that somebody had a great idea that's going to change things i mean we have seen events in quantum computing that are in somewhat breakthroughs right so this is there's an invention of an architecture called architecture which is predicting cubic and you can see at the breaker because like all the other types of people are doing sort of like dissipated and that's not totally true there's still lots of interesting things being pursued but it was it was such a performance improvement that like immediately kind of being used by tons of people who could get it you know the reason that that was the breaker right and that may not have been obvious and i don't think it was obvious when the paper came out and i'm not sure what to press release so it's it's fortunate that we haven't gone fully over to um commercial metrics so you can still use um publications and the uh who's invited to like the big conferences it's kind of a proxy it's not a great proxy but if i were to actually try and give advice to somebody who isn't like i didn't know the space they'll be like well is this person well cited do they publish in like high high-profile journals are they invited to high-profile things it's not great it's better than nothing in terms of a proxy because there is and will continue to be a lot of proliferation in the space where people claim to be able to do things that they can't and and and trying to read that out is is going to be really hard so at the moment there's still this academic proxy right where like quality is kind of gauged by these publications that's not going to last forever right like through this evolving um commercialization phase we're gonna start having performance metrics that become standardized and people start being able to report upon them commercially these those are gonna be broadened beyond the ones that exist today because there's still gonna be a lot of stuff that we need to figure out on how to properly characterize uh and and put things around these systems but yeah i would say if anything that would be the bit of advice i would give to people trying to learn the space which is not great but it's better than nothing the other thing i like about this question though which is sort of the hidden thing was is that there isn't necessarily consistent consistency across the field the field is still i mean the thing that i hope people are trying to do is talk to each other and continue to debate things this is hard stuff and you know we don't because we don't know what the future things are going to look like there will be a diversity of opinions uh one of my favorite quotes is is by terry rudolph i will win everybody but from the extent of like the look that all of us are working on a particular approach to quantum computing because that's what we did absolutely and there's some truth to that right like there's goodness about that like being an expert in something and pushing that as possible is great but i don't you know there is a lot of diversity in the way that i can see the future unfolding and i think that sometimes we we forget that there are competing ideas and people who are trying things and people may believe that this is the way they do it it turns out it's wrong right it goes the wrong way right and that that's actually fine it's just hard for us to do that because we want to be able to see what the future is and what where things are going and i think we just we struggle with that because that's nearly impossible right so i think we sort of sometimes beat ourselves up when thinking about this too much um but it is important to try to make sure like i think the biggest thing is trying to pull back to reality make sure that people are validating what they're doing like this is clearly something that everybody in the field needs to do and you know that people are still publishing papers that's probably the healthiest thing that can continue to occur dave that's a it's a really good point you know as as researchers you know we've sort of learned by experience that you're going to fail at just about everything you try because it's research and every once in a while you'll have a success and so in an early stage like this there's a lot of people who very passionately believe things that are potentially wrong but uh i really like stephanie's intervention of you know there's no single right or way right or wrong way to to tell what you should look at is community consensus if if a large group of people all with passionate ideas think oh that's probably worthwhile it's probably a good good proxy which is not to say the winning horse has been well identified right like this is the whole point is that there will be things that come from left field and there will be breakthroughs and they won't all look the same right they'll probably look very different so it is a bit of a wild ride and you can keep your eyes open but yeah it's it's to try and distinguish what's worth paying attention to versus what's not worth paying attention to that's the proxy i would use and you need that proxy once the number of papers on archive starts becoming literally inhuman i used to be uh so i used to be able to read every single quantum paper that came out or at least stim it um when i first got going and there's there's just no way that's gonna happen there's and it's even hard to train to stay on top of all the news articles from all the companies that's not even talking about yeah it's fun it's a wild ride it's a wild ride that's where you want to be okay so let's extend the wild ride out it's 2050 and cfar has reconvened this panel for a reunion do i still have to ask you the first question or are quantum computers now ubiquitous and a familiar part of our world and if so what would those technologies look like and what societal changes might they have ushered in i think they're going to be boring i think that's what success looks like and i'm an optimist well no like think about it electromagnetism used to convene things like this you know there'd be people that have workshops and they would go and sit in in big halls in the royal society in england and people would look at how mad magnets work and that would be an evening experience right and that's just now it's electromagnetism is as boring as you plug something into the wall and it just works and it will it'll just work and it's going to be a long trajectory and it's going to work under the hood in a way i think may many of the the big ones are um the kind of cloud services that are supporting uh so much really important research today it's under the hood nobody's you know you're not walking around downtown and you see a big giant you know cloud services thing that like pops into your consciousness it's just supporting the broader infrastructure of humanity and it's going to be like that and and if it's successful even the kind of quantum communication things in your cell phone which may eventually come like that is still going to be to the point where you're not going to have to like know about it the like if we're lucky all it will be is like a little local logo like powered by quantum and that's like the extent of it right so that's that's my optimistic streak i love it success is boring well i mean that's what commercializing a branch of physics is yeah right it's taking it from being something that is synonymous with magic to being something that's useful and routine and a tool yeah i mean there's there's the commercial future they're also i mean there are alternative realities that i think are fun to think about right so there are things where like maybe quantum mechanics breaks in some weird way along this journey right and this is one that the total possibility and people will bring up uh you know people there are people who are very skeptical who bring it up and then you have to sort of try to engage with them and understand what they're trying to say but this is one possibility right and i think one of the things that's weird is to me is that that like that that one weird probability so maybe it would all fall that could happen right like we have to acknowledge it it's not a impossibility right now um so that's one weird question like it could be like oh man it's time when you were super optimistic and then all of a sudden you discovered mechanics is broken in network that's something different in this weird way and then likely the story is oh and then we use that in some way right or something else right so it's unclear that that like is the end of the story so there could be some future like that um there could be some future in which like um you know it could be that you know the real question is do we get to the point where we stall out you know usually what people always bring up even though you know right now to me it appears that fusion is undergoing a huge explosion of interesting work sort of a revival even though you know if you started long ago the writer's gonna come back and people are doing that so it could be that it stalls out in some weird way for reasons that we don't quite see right now uh and then we're kind of limiting that right and trying to figure that out i i just i put that in low just because i just don't see them right now i see a new fourth approach which goes pretty fast right you would think maybe there are we will likely laugh at the brute force approaches that are being taken right now like being through what i mean by blue forces you know individual control of single qubits with ion traps or you know bigger again circuit that they need to put together in this air correcting code to build a larger quantum computer like that very brute force and technology the darkening technologies have sort of really done that particularly well that's hard right like you know it may be that we look back at this era to be like oh this is before they discovered you know specifically really good quantum i'll call it that you know quantum system that does this and it's now obvious that that was the right solution and we kind of laughed that there as like the you know these people who are trying this this amazing thing like we sort of do this when we look at early computing too right but there are these crazy things that people trying to build right some of our storage devices back in the early days were just nuts right but but it turned out that like they were on the they were all doing great work but they just didn't scale in the right way i mean like we as human creatures we take what we know and we adapt it much more than we invent right and so like it's going to be true that the first cars look like horse-drawn carriages and that's fine we shouldn't we shouldn't like laugh at it really like we should be like you know that was what was known at the time and and honestly i think it's actually a tour to force some of the stuff that's coming out that's demonstrating what it is with with all of those caveats in mind like with all of those constraints and the rest of it is heroic it's heroic stuff and it's definitely paving the way for more venture funding for other things you know like there's there's it does service the overall dream of this rising tide lifting all boats because it will be a computational resource that will be used across across the board right like just across the board and so having that and anything that unlocks that even like culturally um conceptually should be celebrated i think stephanie i love your optimism for human ingenuity we're we're pretty persistent we are so i have trust that people have at least some motivation fueled by one mechanism or another to make it happen right especially when it's known when you can kind of see like the the road map a little bit not the path story but like the the mountain that needs scaling yeah or people are pretty uh persistent all right i've got some good questions coming in from our audience here so i i'm going to turn over to them steven asks there are many applications for quantum computing and one of these is cryptography will quantum computing mean the end of security by cracking security algorithms or will quantum offer new approaches to security that cannot be cracked by simply increasing the number of qubits there's a type of cryptography that people call post-sponsored cryptography which is the idea to do things that are resistant to a large quantum computer now what's fascinating is cryptography a lot of our crypto systems are based on the hardness of the homes but the classical example is short either factoring a number is far uh and though the hardness method gets the security guarantee now why didn't we believe it was hard but we believed it hard because a lot of smart people were trying to come up with things that failed there's some other things going on there but it's roughly that uh and then peter short came along and said hey this weird computer can break it pokemon cryptography is sort of now in a similar situation where they they have some cryptosystems that they're trying to develop it should be secure and the idea right being that like smart people are thinking about this and they're trying to choose problems that are resistant to a quantum computer but again we're now back in this processing that there's no way to break those cryptosystems and i think that the jury on this is a little bit out and we're at the beginning of this the scary thing that if you want to put it in terms of the fear right the thing that's kind of scary is that a lot of the photography that's used right now not just like the internet one's easy i used to write google software today right like going into changing google's infrastructure in terms of internal stuff to use different systems it's a lot of work but it's like not a major scary thing but if you think about all the cryptography that's used in the wilds on the like the internet of things and that you can't just upgrade right like it's often uncreateable and it's going to be with us for say 10 20 years then you have to stop worrying about those types of systems being broken because they don't already have this postponed cryptography the other part of this story is just but there are things like there are other quantum cryptographic principles like on key distribution that can be used and i think that's i'm less of an expert in it but i just you know there are there is some way in which there are other benefits they're not necessarily uh they're not it's not sort of this like quantum cryptography stalls all the things the shores algorithm breaks that's not quite correct for my understanding but there are these other photographic principles there's actually probably a lot of them working out there that we're still trying to understand it's kind of this there's a just there are people who work in sort of these weird distributed systems and trying to think about uh you know how do you think they're in interesting ways and cryptographically secure ways maybe their way the quantum community could benefit there um and i think that's probably like if i were young i would be like oh that's a fun place to work in because i don't see it i don't see a ton of people thinking about that but it's definitely a club field that exists where people are really trying to think of what interesting cryptographic print it is yeah so so qkd and these other tools are are not reliant on the principle that something is hard to compute it's relying upon physical principles so it's not as if it's resist they're not as if as if a quantum computer can necessarily break it right so it is it is very much it offers it offers one and takes you know it's if you have to um think about both sides of it but teleportation like the information does not travel through intervening space it's a very different way of thinking about the movement of information and cryptography more broadly so lots of opportunities there interestingly the technology that's going to be used to unleash that is kind of similar to what a small-scale quantum computer is going to look like so a repeater isn't going to be so different you know it's going to have you know a lot of similarities with small-scale quantum computers so it's going to kind of come in intent but there is going to be um yeah a bit of a onboarding to a new protocol that can uh withstand or withstand shore's algorithm so stephanie on that point james asks you mentioned the teleportation aspect of quantum computing so information teleportation does it look like that aspect will be useful in addressing latency and and can you imagine a world where our communications infrastructure is entirely quantum so those are two separate questions latency is an easy one in that there exists um kind of bounds on what uh that means in terms of information propagation so you can uh you when you when you teleport information quantum information the information is actually teleported but it needs like a tidy up um signal you need to be able to send a signal to be like okay well but it's it's correct but like up to a sign and you need to like fix that sign basically and so that still involves a signal that needs to be sent so it doesn't change the latency so much um which is kind of this wonderful boundary condition uh in terms of how the laws of physics all work because you can have this this beautiful effect that um that seems to go faster than but not you can't signal faster than light and that like nuance is enough to um yeah keep a whole bunch of other uh uh things from from going wrong from a physics perspective so what what do you see in terms of modern communications infrastructure so we're still we're still bounded by the speed of light so do we just use classical fiber optics and we we never have quantum communication oh no we will um we absolutely will there's a there's giant satellite programs underway their telecom is going to be able to support quantum communications the repeater stations are going to look different um there will be i imagine you know that's so there are these quantum safe approaches but i will say that you know the only thing that's truly truly quantum safe is going to be some amount of quantum key distribution or similar like something where you teleport information uh to some level right like some people will really insist upon that level of security and there's because some of the algorithms that people have put forward as you know something that ought to be secure against quantum attack have fallen right like people have found uh ways through so it's one of these things that you know i imagine they'll have something but not just from a communications perspective but from a distributed computing perspective like quantum computers when you combine them the reason why they're so powerful is because when you combine them it's much more than some other parts they combine and the space that they encode grows you know exponentially in the qubit number and so like two ten qubit systems joined together is so much more than like it's it's so much more than just uh the sum of the parts so you're going to need to have distributed quantum information if we wanted to really leverage large-scale quantum computing resources across large distances you know if you needed to rally that together for whatever large-scale computation you're going to need to be able to send quantum signals and teleport stuff around just for the computing sake so that's going to be part of our future you know like it sounds amazing to say but it's absolutely going to be part of our lifetime these are these are going to be happening on some small scale and then growing bigger and bigger and yeah distributed quantum computing is going to be happening on the cloud and again it's going to be boring i love it dave i have a question that looks tailor-made for you from dawn when quantum hardware architecture becomes stable are we likely to see a huge change in how software is written what software architectures will we likely see as a response to quantum hardware and it's in the top question we're very we're in the very early days of writing quantum software it's very you know it's very much like the stuff you do with the high levels very much like assembly language they're like hidden layered there's sort of two layers that people talk about they're sort of like programming the quantum computer up here maybe through a cloud interface and then there's like the software you need to write to get the quantum computer to run all operating systems to me there's a ton of back maybe work going on there it's a little bit more of it's hidden because a lot of it's being done against industry groups but a lot of really fun challenges to get that system to even work right that's part of the one of the hidden things i can remember the day of google when we were running this beyond classical experiment where one of the stoppers came in and he was like it looks like it's going to be able to scale up to be you know perform what it does it was just like he was like you could see the relief sort of knocked off the shoulders so i i suspect right now we're just at the beginning stages of understanding how to run these very simple things we are admitting a level of solutions that can help us build algorithms or do things correctly i think i think we're i think we've always been looking for these in quantum computing people working in quantum algorithms to understand them there are some of these alternative models that people trying to come up with and they could give a new way to think about that i sort of believe i'm hopeful that we're at the beginning of people starting to think you know trying to develop these weird things to help us and guide us in and building new algorithms that's really challenging right but again you have to think back to history of algorithms and just notice that like that that didn't occur in places i thought there was a conference recently that i didn't know existed but i just learned that it's on simplicity and algorithms and i love that this idea that sometimes simple algorithms and explains it algorithm that come out into input very important and i feel like that's one of the things that i'd love to see more of in quantum computing ways that we can conceptualize the model that help us do new things um and i think popular would be a big part of it i do i do think that sort of this this mipping hole and you know what i think i'm excited about our tools to help us do that by building tools for people to try to explore space and understanding and this is this is challenging but if you think it's going to be part of our story um the good thing is it's the beginning right the stuff that we've written in quantum software is very new that's a lot of fun too because you'd be like oh i'm doing this new thing that nobody's thought of before to do in quantum computing so that makes that feel very fun thanks dave stephanie i'll give you the last word before i have to wrap up on on the note okay so on the note of hardware and software development there's a giant giant opportunity for co-design i don't think we're only just kind of getting into it but um yeah there will be as you're saying this dominant design emerge and or a few dominant designs maybe that emerge it'll coalesce around a few and there will be lots of opportunities for co-design where you start thinking about you know through the software stack compiling down to actual hardware and optimizing on that and so that's why there's a lot of really interesting developments happening in the corporate side because within the academic side usually things are quite uh agnostic hardware agnostic and they're kind of done in general to be as broadly applicable as possible but there's a lot of opportunities for optimization and really you know unleashing things when enough people across the entire stack are all working and and having the higher end of the stack informing the hardware too right like that it's not just a one-way street it's up and down and so there's there's oh my gosh it's fun it's so much fun i'm so sorry to say that is all the time we have for today i would like to thank all of you for joining us and special thanks to stephanie and dave for sharing your insights and optimism for a boring world with us please visit our website cfar.ca to find out more about the series and to view recordings of previous talks a recording of this talk will be posted by early next week [Music] you

2021-07-29 19:30

Show Video

Other news