Root Causes 83: Quantum Apocalypse - Does COVID-19 Change the Z Date?

Episode Transcript

Lightly edited for flow and brevity.

• 

  Tim Callan

  We are joined today by our frequent and beloved guest, Alan Grau. How are you doing today, Alan?

• 

  Alan Grau

  I’m doing great. Thank you, Tim.

• 

  Tim Callan

  Alan, you know, usually when we talk to you, we are talking about a couple of matters where you have expertise, which is sort of embedded systems and block chain and applications for block chains, but it also turns out that you are reasonably well-schooled in the issues around quantum-safe cryptography and we are going to have a quantum-safe cryptography discussion today, so we invited Alan to join us. How all of this started was the three of us were talking just yesterday and I posed the following question, which is: Does the COVID-19 pandemic and the resulting actions that people are taking change the arithmetic around Mosca’s Inequality? So just as a reminder for the listeners, Mosca’s Inequality, the basic concept is let’s say we have an equation it’s X + Y ? = Z, where X is the amount of time it takes to find quantum-safe algorithms, Y is the amount of time that our secrets to be secure and Z is the amount of time until we have a quantum computer that can break our existing algorithms. And the question is, is X + Y greater than = to or less than Z? Because is X + Y is, you know, if Z is less than X + Y then that’s potentially problematic. And so, I am wondering and I posed the question, in the light of all of these changes that are going on right now in our society, do we feel like the actual Z date and possibly the X and Y dates or the X date change? And that’s how all this got started. So, it was a bit of a head scratcher. I think you guys would agree and so what we decided to do was we were going to hash out it in real-time on a podcast. I’m going to attempt to advocate the viewpoint that it will make a difference. Alan is going to listen to what I have to say and then, you know, poke holes in it and see if it tests and see how well that holds up as a concept and Jason is going to be here just to kind of moderate and be an audience and be convinced.

  So, are we ready to do this, gentlemen?

• 

  Alan Grau

  Definitely.

• 

  Jason Soroko

  Let’s go, Tim. Sounds good.

• 

  Tim Callan

  So, since I started all this trouble, I’m going to start. So, I want to propose that we’ve seen massive disruptions in various forms of work as an effect of lockdowns and self-quarantine and shelter-at-home and all of these things that are going on. It’s very disruptive to the work that people do. Even people who are doing intellectual work who just sit at a desk. Now they are sitting at a desk at a strange place. They might not have their normal computers. They might not have their normal filing cabinets. They might not have their normal methods of communication. They might have other distractions, loved ones they are worried about, kids at home, things like that and all of that takes a toll on our productivity, but in particular, the work that’s really been hurt is the work where you have to physically be someplace. So if you were, let’s use a really easy example. If you were in a manufacturing facility work has gone to zero because people can’t gather to manufacture.

  And one of the interesting things about quantum computers is fundamentally they are physical things. Right? We are working on physical engineering problems. We are trying to make our computers hold qubits for longer and we are building actual machines to do that. We are super cooling them. We are working on various other things to try to make them stable and at least a non-trivial part of that is a physical engineering task and that is going to be hurt from a productivity perspective that is going to be hurt more than just purely intellectual work is gonna be hurt. And so, it seems to me that we should expect the Z date to move out because of the COVID-19 pandemic that we are going through right now. That the people who would have been in facilities making their quantum computers better – yeah, maybe they are having some deep thoughts and maybe there is some amount of modeling or testing they could do offline, but they’ve gotta be less productive. Gotta be.

  And how long are we all going to be sheltering in place? I don’t know but I know the U.S. Congress just passed an unemployment package that is targeted to last for 39 weeks. Right? So 39 weeks, that’s three-quarters of a year. So hypothetically, if you can’t go into your lab and work on your quantum computer for three-quarters of a year how long is that setting back our milestones? Well, three-quarters of a year. Maybe longer.

  So when we are talking about timeframes like ten years it seems like adding a year to that is non-trivial and might be the difference between certain secrets being safe and being not and so I actually suspect – and we are never gonna be able to prove this, but I suspect that one of the odd consequences of this whole thing is that the quantum apocalypse might actually be held off a little as a consequence of what’s going on right now with COVID-19.

  So that’s my premise, Alan.

• 

  Alan Grau

  All right. So, when we talked about this really briefly yesterday, my initial gut reaction was that, you know, what we are talking about here is, you know, I mean clearly what’s going on right now is having an impact across the board on productivity. I mean no one is arguing that that is not the case. But I look at this as more of from an impact on the zed date and what companies need to do to prepare for the zed date as more of a rounding error than a meaningful movement of the milestone. So just doing a little bit of quick math, if we are saying that the zed date is ten years from today, well that’s 2,500 workdays from today.

• 

  Tim Callan

  Ok.

• 

  Alan Grau

  And so, in the meantime, you know, if the zed date moves by 10%, you know, that’s a year. Maybe that’s meaningful but the reality is during that time you and I and Jason are still working. We are not working specifically on building quantum computers, but I suspect that most of the people working on quantum computers aren’t going to totally shut down over the next year or the next six months and not continue to make productive strides. So, it may be slowed down a little bit, but it won’t be stopped. So that end date, again, may shift a little bit but I don’t think it’s going to shift in such a meaningful fashion that it’s gonna change what companies need to do to prepare. And as we look at that date, too, the other thing to keep in mind is that’s not a hard fixed date where, you know, we can look ahead and say well the 2020 Olympics got moved out by a year so we know the exact impact of that and if you are a young athlete and you are thinking maybe you are preparing for the 2028 Olympics, you know, so it’s eight years out so comparable to the zed date you’ve got a hard and fast fixed date that you are training for. For quantum computing and the zed date for what we are talking about, it’s a much more fluid or flexible endpoint. Meaning some systems, I mean there is some data that needs to be kept secret for decades and so if that data is already in a hacker’s hand or an opposing nation state’s hand today, they’re still gonna have that data protected by non-quantum-safe crypto algorithms after the advent of quantum computers. So, in some ways, the zed date or for certain datasets, the zed date has already been missed in my opinion.

• 

  Tim Callan

  Yeah, and real quick let me pitch our Episode 43 where we go into detail on exactly this point you are making now, Alan. If you are interested in this idea of how the zed date could really be many dates or multiple dates or unknown dates, get into our Episode 43 cause we go into depth on that concept there but go ahead Alan, please continue.

• 

  Alan Grau

  Yeah, no, absolutely. I mean that’s really an important part of the concept, right. For some systems they are not, you know, keeping the data safe is not nearly as critical. It doesn’t need to be protected for a long period of time so that systems are not as critical or not as time-sensitive in terms of an early zed date if you will. But the other thing that’s happening at the same time is while maybe the zed date, you know, if theoretically we agree with Tim’s point that it’s pushed out by a year in terms of when quantum computers will have reached a state that they can be utilized for this type of activity. The other thing that’s happening right now is there is tremendous amount of IT professionals some of whom might be working on quantum readiness within corporations are now scrambling to address remote workplace issues, scaling up systems to still maintain security while they are dealing with suddenly instead of a small percentage of the workforce, you know, the vast majority of the workforce working from home so the preparations that could have been taking place are also being put on hold. So I think, you know, best case things are gonna kinda slide at the same rate, but in fact, it’s possible that the preparations will be impacted even more severely than the research being done on quantum computing.

• 

  Tim Callan

  Now what’s interesting about that point and I think that’s an interesting point, is though certainly some not all, but some of that preparation is in fact becoming crypto agile and so part of what I’m doing is I’m getting my systems in place so that I can easily spin up identity certs and things along those lines and that crypto agility is also a non-trivial part of everybody getting ready to deploy new quantum-safe crypto once we settled on those algorithms. So, in that regard, you know, I don’t know. It seems like a decent amount of this work might be reusable.

• 

  Alan Grau

  Yeah. There are probably pieces of that that is true. There are other pieces obviously that won’t be, right? I mean we’ve talked about some of the providers that we work with that are having capacity issues and how they are addressing those. So, yeah, I think there is certainly a lot of things that are moving that may affect both when the quantum solutions are available when companies are ready to adopt them and when they are needed.

• 

  Jason Soroko

  Ok guys. So, I got some thoughts.

• 

  Tim Callan

  Go.

• 

  Jason Soroko

  Yeah. So, you know, if I try to break down the crux of Tim’s argument, which is the zed data specifically, which is the people who are working on the side of quantum computing and advancing it forward, we are talking about Mosca’s Inequality and one of the major things that you have to accept if you believe Mosca’s Inequality is that there is a linear function of quantum advancement. In other words, the number of Eureka moments that put together more and more stable qubits is much less than the day-to-day linear experimentation. In other words, experiments done by engineers that are leading to linear steps towards more and more qubits being stable. And we see that.

• 

  Tim Callan

  Which seems to be what’s really happening, right?

• 

  Jason Soroko

  And in fact, that’s exactly what’s happening. We’ve been told that that’s what would happen and it looks like that is exactly what’s happening because every couple weeks, we get one of the big three or four players who all are playing with their quantum computers all saying, look, I now have quantum supremacy. I mean how many times has that been announced in the past while? And that’s great. It’s actually fantastic and it’s usually the announcement is because one exceeded the other by a few bits and the other cries foul because it was done, you know, because of a different math routine or the problem they were solving wasn’t the same.

• 

  Tim Callan

  Or it was on a Tuesday.

• 

  Jason Soroko

  In other words, scientifically they weren’t comparing apples to apples. But it doesn’t matter, we don’t do science anymore, we do news releases. Right? That’s what we do now. It’s the - - it’s public opinion that now rules the world. So in the backrooms of the people doing this the question is are these people being disrupted and will that lead to anything significant and I think – I’m going to very roughly paraphrase what I think that we are discussing here together with Alan is that that might end up being rounding error and we weren’t really sure when the zed date was anyway. So in other words, how could we measure what the disruption would be. We are just theorizing here whether or not there will be disruption. Well, here is what I can tell both of you from my own very recent experience is, I actually know people on both sides of the equation. I know people who are working on the quantum computers and I also know people, probably better so, people who are working on quantum resistance, which ends up the two together being the zed date, or Z date to my American friends.

• 

  Tim Callan

  Yes.

• 

  Jason Soroko

  What I can tell you is they are working full steam. They are not being disrupted because the people who work on the physical computers – I think there is two aspects here. One is a lot of them are sometimes working on computers that require liquid cooling which requires them to be in a clean room, in other words, all the different health and safety measures are such that they are probably safer from COVID doing their work than they are being at home. That’s the ironic thing about this. And the other half of it is there are also people working on non-liquid cooled quantum computers who are able to do their work more or less theoretically. In other words, they don’t spend most of their day building a computer – building a physical computer. These are engineers who spend more of their time probably sitting in a chair with their head in their hands thinking and so therefore they can maintain their productivity in that way.

  The other side of the equation, which of course is the people working on quantum resistance and the math behind it, a lot of those people are university researchers. I happen to know one, so it might not be a statistically significant sample but that one person I happen to know very well is steaming ahead at full, full charge. In other words, I think, Tim, the ironic thing here - -

• 

  Tim Callan

  And I would expect that, right. If the work you do ultimately means all you need to have is a computer and perhaps a whiteboard there’s no reason why you are not steaming ahead no matter where you are.

• 

  Jason Soroko

  So here is the humorous part of this podcast, right? I always gotta inject a laugh. People who are really, really good at thinking through maintaining stable qubits in a computer and the people who are thinking hard about the mathematics behind quantum resistance, those people are already very, very good at social isolation.

• 

  Tim Callan

  Very good point. That is a valid point and, yeah, if you are an academic and fundamentally you are a mathematician at your core, certainly working by yourself is one critical part of being able to be successful at your profession. Not to say that collaboration isn’t incredibly important but also just sitting and doing the hard, intellectual work is incredibly important as well.

• 

  Jason Soroko

  So, Tim, here is where I’m gonna kind of lay out where I think you are both right. In a sense. And that is this. If we take a look at all of this as a probability problem, right. So there is going to be moments of, you know, it’s just like think of the analogy of tectonic plates. They are always shifting. They are always shifting in a more or less linear pattern, but every once in a while, you get a little earthquake and a tectonic plate will shift under another one by a few more feet unexpectedly. And we cannot predict when those moments will happen. If you think about the whole zed date and the components that make it up, which include Eureka moments that are unexpected in mathematical understanding and you take into account progress and engineering that leads to maybe a few more qubits than last week, there’s gonna be a bunch of these unpredictable moments in time where we are either gonna move a little faster or things like COVID-19 which absolutely may have the effect that you are saying, Tim, which is maybe there will absolutely be disruption of X number of months, up to a year. Who knows what we are talking about here. Again, it’s an unknown amount of time and the probability of it is probably greater than 0.

• 

  Tim Callan

  Yeah probably.

• 

  Jason Soroko

  So this is trying to get to the real truthfulness part of how we think about this, which is, you know, Tim, I think the things that might unexpectedly speed it up, which are things like, hey, it turns out that quantum annealing is all that’s necessary. We don’t really need full-blown Shor’s algorithm with a fully-gated quantum computer. All we need is a quantum annealing computer and some really fancy math to do the same breakage that leads to a zed date. For example, and I’m not saying that that will happen, but we have podcasted about that. Those are the kinds of probability moments, right? Moments of wow, something changed and we didn’t expect that combined with, heck, pandemics. Right? Which will happen from now to ten years to now when we expect this thing to come to its conclusion. There will be both and the probability of both happening are greater than zero. Do they cancel each other out? I don’t know. I don’t know. I’m not thoroughly convinced yet except to say that I think, Tim, your point is utterly valid. It’s just I have no way of measuring the effect of it.

• 

  Tim Callan

  Yeah. We are not going to be able to measure any of this of course.

• 

  Alan Grau

  Yeah. There is one other consideration that nobody has mentioned and, again, is still unknown, but there is predictions that this is going to have a significant – clearly, everyone agrees this is going to have a significant impact on the economy globally and, you know, one possible outcome from that is some of these firms may choose to cut back on their funding for research. And so, if that happens, that could have an impact on either side of the equation. So, again, it is an unknown but that could be one that could significantly impact what the end result is. But my view on this is this is still a critical problem. It’s still something people need to pay attention to. Obviously, today a number of folks are getting redirected onto other short term issues, you know, whether it’s taking care of personal issues and family and relatives that are needing assistance in this time or working on IT issues to allow people to work from home or to support some of the changes in how people are working or whatever it may be, you know, and we have to address those first, but this is an issue that if a company is trying to figure out how they are going to be quantum safe as they move forward it can’t be an issue that gets simply dropped. It may be something that gets deprioritized for a short time but it’s still an issue that people need to pay attention to and given that the zed date was unknown before this happened and remains unknown but is looming, we can’t just ignore this.

• 

  Tim Callan

  So you mean we are not gonna be able to come back in 12 years and see which one of us right and go nah, nah. That’s so disappointing.

• 

  Alan Grau

  Well, I suggest Jason should declare winner but I don’t think it’s a - -

• 

  Tim Callan

  But he is far too polite for that. We all had our good points. I think that was very polite. I think that was very polite the way you did that Jason.

• 

  Jason Soroko

  Yeah, absolutely. So what I will tell you on just the point that Alan just made about funding. That absolutely will make differences. I think the funding though for a lot of projects is already in place. I used to live in the world of university funding and a lot of times these projects are funded at least a year or two, so they will continue to a certain point and I don’t think NIST is gonna pull this because of a pandemic. That’s just my thought. The other thought I had is if you look at the evidence, one thing Tim that you and I theorized was that the public clouds would start to democratize early forms of quantum computing. In other words, I don’t have to have a quantum computer all I need to do is hook up to a public cloud and I can spend fractions on a dollar to buy quantum computing time. And that’s already happening.

• 

  Tim Callan

  Yeah.

• 

  Jason Soroko

  So that was something we theorized not that long ago and it is actually happening. So what will look like a series of traumatic events is actually exactly what Mosca’s Inequality says, which is in the fullness of time it’s a linear function with a bunch of quick movements. A bunch of small, quick movements. And I think that between the news releases we see, the announcements we are getting out of the public clouds to democratize this and also the work that I know is being done on the pure academic side for the math, all of this together is typically working out. I have not seen anything yet, Tim, for us to have a podcast and say, whoops, Mosca was wrong. We have to rethink this.

• 

  Tim Callan

  Right. Completely agreed and, you know, all valid points including ultimately the fact that at the end of the day we are never gonna be able to put a number on it and say, ah, it moved by this amount. And I’ll accept your point Alan that even if it did move some, it’s not moving by dramatic amounts. Right? If it’s plus 10%, minus 10%, that’s within the realm of we didn’t really know for sure anyway.

• 

  Alan Grau

  Yeah. And again, it’s still something we need to pay attention to. We need to assume that it’s going to be here sooner than we think and work hard to be prepared for it.

• 

  Tim Callan

  All right. Well that’s probably a good place to leave it. Thank you for indulging my wild idea gentlemen and it’s always fun to have you on Alan. So thank you, Alan.

• 

  Alan Grau

  Thanks, Tim.

• 

  Tim Callan

  Thank you, Jason.

• 

  Jason Soroko

  Always great, Tim. Thank you.

• 

  Tim Callan

  Thanks everybody. This has been Root Causes.