Quantum Supremacy and the Cryptology Time Machine – CF058
This week on Cyber Frontiers Christian and Jim recap the latest major development in quantum research with Google’s “Quantum Supremacy” milestone. We review the headline narratives and compare today’s quantum progress with where we and others expected it to be. We delve into the tangentially related topic of cryptography, quickly leading to a walk through time of where secure computing has been and where it could be heading looking forward. Scattered throughout we bring in what it all means for the average guy, and even get some time to relate it all back to… ham radios?
Cyber Frontiers is all about Exploring Cyber security, Big Data, and the Technologies Shaping the Future! Christian Johnson will bring fresh and relevant topics to the show based on the current work he does.
Support the Average Guy: https://www.patreon.com/theaverageguy
WANT TO SUBSCRIBE? We now have Video Large / Small and Video iTunes options at http://theAverageGuy.tv/subscribe
You can contact us via email at jim@theaverageguy.tv
Full show notes and video at http://theAverageGuy.tv/cf058
Podcast, Cyber Frontiers, crypto, blockchain, ham radio, cryptology, certificate, quantum computing
The world frenzies over the news of quantum supremacy:
https://ai.googleblog.com/2019/10/quantum-supremacy-using-programmable.html
It’s not your grandfather’s ham radio anymore… welcome to SDR:
Jim Collison [0:00]
This is The Average Guy Network and you have found cyber frontiers show number 58 recorded on November 4 2019.
Jim Collison [0:19]
Here on Cyber Frontiers we explore cyber security big data and the technologies that are shaping the future. If you have questions, comments or contributions, you can always send us an email Jim at the average guy.tv you got you can contact that guy over there Christian at the average guy.tv of course find me on Twitter at j Collison and Christians @borgwhisper. Don’t forget the average guy that TV powered by Maple Grove partners get secure, reliable high speed hosting from people that you know and you trust for plans to start as little as $10 a month. Visit Maple Grove partners.com Christian still 10 bucks right
Christian Johnson [0:53]
10 bucks
Jim Collison [0:55]
Plans are still the same. You have changed them
Christian Johnson [0:56]
same plans, same dream, same platform
Jim Collison [1:00]
We are back.
Christian Johnson [1:01]
Great service.
Jim Collison [1:03]
We are really super fast. We are. We’re back for 58 I think August something was our last episode what’s been keeping you busy?
Christian Johnson [1:12]
All things, from engagements to Yeah, congratulations on that, by the way, thank you to work to just every all of the genies seem to win when when it rains it pours.
Jim Collison [1:28]
So 2020 will be the year
Christian Johnson [1:31]
it will be the year. So lots of moving and logistical twisting and keeping keeping boats rowing at work as well.
Jim Collison [1:40]
Yeah, yeah. Ken says what? Congrats, congrats for you. Thank you, considering we were just in the pre show saying how home server show or Home Gadget Geeks whatever I call that thing now. I’ve only done 420 episodes that can’t get the name right. We’ve been doing that nine years and home server show friends before that. So probably known yet a decade. And it’s a long time,
Christian Johnson [2:04]
right? Sure. Yeah, it surely is. Time flies when you’re having fun.
Jim Collison [2:09]
The rest of high school all of college Yeah, getting married. I mean, that’s pretty great. I kind of like it. You’re, you’re kind of like one of my kids, you’re just you’re in the same as, right. So my kids and they’re all doing that too. Yeah. And it’s it’s even weirder from the perspective of having 10 years of YouTube history. If you go back and you watch and it’s, it’s like a window through time it is it is pretty great. We’re going to spend some time talking. Speaking of this, it 10 years ago, we weren’t talking about quantum computing at all. It’s actually kind of become a thing. I hear Christian, different sides of the equation like, yes, in theory, it’s going to work, but it’s not actually possible. And then I say, and then I hear people say, Oh, no, no, this thing’s going to work. So is it going to work? I mean, are we ever really going to be able to get to quantum computing or maybe a working today in some secret back room somewhere some genius.
Christian Johnson [3:03]
Yeah, I mean, the paper definitely shifted the conversation. I don’t remember if the last show, we covered the fact that it had kind of accidentally been leaked on the NASA server early about a month ahead of when it was actually supposed to be released. And then it was finally published properly here at the end October and definitely what I would call a shift. I am hesitant to use the word paradigm shift because I think more testing validation of assumptions and longer term forecasting needs to be done but really well let’s let’s talk a little bit about the experiment first and set it up. So what Google essentially tried to do was take on a classic style computer, the ones you all know and love that do zeros and ones, and specifically one of the IBM supercomputer That would have taken up to 10,000 years I think it is to do the same computation. So what they were trying to test is can they build a quantum processor and framework that makes a classic computing problem that can’t be solved in you know, reasonable time, most would say 10,000 years is not reasonable time to get a answer to a question and a matter of days or hours instead of you know, this huge lead time. And what they found was that using this 54 bit processor they dubbed second more that uses some fancy, quote unquote, processor design and how they build their quantum gates and they’re all of the tangled particles that make up the wacky world of FQ that and and Superstrings, etc, etc. They basically end up proving that they can solve the same problem in 200 seconds, and so 200 seconds and comparison to 10,000 years is obviously pretty substantial. When we talk about what people have been promising quantum computers to be able to do, in all walks of life, it has always had that level of dramatic shift of things that were not possible or tangible for classic computing models would now suddenly become tangible, right? So the terror and Panic of world crypto is going to be irrelevant and etc, etc by quantum. This is really the first actual publishable experiment with results that people can go and look at. That makes the assertion that there is quote, quantum supremacy in the natural solution of this problem.
Jim Collison [6:01]
So yes, it’s a real thing.
Christian Johnson [6:03]
So it’s a real thing. I want to caveat though that is a very early indicator, right? It shows that a couple things. One, IBM was, of course quick to say, you know, a IBM is also a competitor in quantum space BA obviously own the supercomputer that it was compared against. And so there have been some assertions that they could rewire the classical algorithm to have it be solved in a couple days instead of the 10,000 years. I think the assertions of that are somewhat vigorous at this point. What’s being hailed though, as real progress in this experiment is the fact that it now seems to be becoming less of a problem of if it’s possible from a physics standpoint and more of a matter of is the right engineering and instrumentation in place in order to have these types of achievements. So, I would say up until now, very unclear that there was going to be a driving use case that would allow these systems to become further developed. It seems like Google is well on the track of creating what those other use cases and driving examples are going to be in terms of how useful this particular experiment is not anything, you know, big headline article, so it’s not something where, you know, crypto is broken, or we can now solve some of the hardest, you know, exponential or other types of classic model problems in in nanoseconds. It’s nothing Back to the Future like that. But this lays out a blueprint and a suggestion that for people who have consistently doubted the tangibility of these types of technologies moving in a direction where they will start to solve some of those real world problems. I think this is what we would call a shift. Now the Ethereum co founder who you know all things Bitcoin blockchain, they obviously are very interested in crypto and what quantum can do for their their cryptocurrency. He does call this as a quote unquote, a shift in the computational model and one that’s here to stay and I want to read the quote from Charles. His last name’s interesting, Hoskins and I think that’s the correct pronunciation. Well, that’s enough. Yeah, close. Yeah. Who’s the CO creator of Ethereum? And he says, quote, according to their definition, yes. Unfortunately, there’s a huge delta between what they’ve done and what’s required to actually use quantum computers to become problem that For cryptography, but it’s a major milestone, it’s an example of the endless and relentless progress that companies can make once they’ve set their mind to something and they have the people to execute. The reality is this computational models here to stay, the physics are sound. And it’s more of an engineering problem than a theoretical problem at this point. There’s certainly a lot of science to do and a lot of math to do and a lot of careful programming to do. But today, it feels more inevitable than it did 10 years ago about whether quantum computer will ever exist and be useful to solve real world problems. So yeah, very much. I would say in comparison to last year when we talked about some of the RSA presentations around it and really, you know, debunking every single use case where people thought quantum would be at this point, and it just hasn’t been pretty, pretty big news.
Jim Collison [9:50]
So as you think about it, for the average guy for the next 510 years, does it begin where does it begin to kind of leak in or provide useful, you know, what space or is that still too too cloudy? This one,
Christian Johnson [10:09]
I think so cloudy, where I would expect it to start becoming very useful someday, if we can get to that level of fidelity and a quantum operating system is to be able to run machine learning and other statistical models against it. I mean, imagine if you can train very sophisticated models across huge amounts of data in seconds. And the ability for your AI algorithm or other autonomous systems to respond has now just dramatically taken a whole new level of skill, which could in turn make them seem a lot more quote, artificially intelligent. So I would expect the problems that are most interesting and broadly applicable to society or how quantum will be able to support machine learning and AI data models. I think some of the other ones like crypto and general general purpose Security are interesting. I would expect it to maybe impact blockchain and how you can deliver transactions over blockchain before necessarily, I’d be concerned about it outright breaking cryptology. But certainly, if we’re making the assertion that it has the potential to be relevant and useful to machine learning algorithms and the future, it’s reasonable to also conclude that we would, we would expect cryptology to be challenged in similar ways,
Jim Collison [11:31]
but do okay, so we get faster computing, don’t we just begin to just up the game on then on, on the cryptography itself to make sure it’s even harder and I mean, that just isn’t that just kind of mutually assured destruction on those two sides.
Christian Johnson [11:48]
You know, in theory, right, it becomes a chicken and egg problem. That’s much like when specter meltdown got released, like you could run some quick software patches, but at the end of the day, you needed to buy new hardware in order to run Get out. And it wasn’t clear that there was a single processor available on the market at the time of the disclosure that fully routed out the problem. similar scenario here, right? You’re talking like trillions and 10, how many trillions of devices that now process zeros and ones that would become quote unquote, security irrelevant, that have to be replaced across the board. And then on top of that, you’re talking about, will quantum computers have the same level of portability, stability, and economic scale as a classic computer? So maybe the case that if you buy 100 billion dollar supercomputer, you can crack cryptology someday, but for the average person, they’re not going to go buy 100 million dollar desktop to protect themselves?
Jim Collison [12:49]
Really, really nations right, at this point, we’re talking about a nation state. Yeah, of investment, right?
Christian Johnson [12:55]
certainly could be, you know, along those lines, and I Think the other reality here is that the traditional calculations that have been done for the amount of power that would be required to have a computer like that would be so absurd it, you know, it’s it’s the equivalent of saying, if you tried to build a data center back in the 1960s, that use the size that it did just to get a single megabyte, it would be impossible you need like an entire continent, and even then it’s unclear of what type of power you would need. So there’s going to be some economics of scale that have to be factored in as well, that might transcend like it’s exclusive for some, but not for all.
Jim Collison [13:40]
Yeah, and I think, you know, there’s always this ease in period for these kinds of things. Right. I think your 60s example is pretty good. 50s and 60s, even right before we were trying to put a guy on the moon, you know, the technology that they had there was super expensive and we look at it now and kind of make fun of you know, we Hey, we landed a guy on the moon with this kind of technology, right? But it took years it probably took you know, so let’s 1965 to 1985. So 20 years to really kind of get to a point when we first really begin to see a PC on the desktop. And so Okay, let’s just say quantum is today and even though we think about Moore’s law and some of those other things, we really just, there are really some challenges to overcome. Will we overcome them? Why if we can hold society together long enough to have those kinds of things happen, right? Yeah, probably but it’s I think it’s going to take some time i do i get i grow tired. A little bit of the doom and gloom, you know, we’re gonna crack you know, all of a sudden, you know, cryptography is irrelevant. Well, is it and you know, it like you said the, the costs are pretty prohibitive. We’ve always been in a state where the nation states have had a lot of the technology available. To them before everybody else did and then there’s some checks and balances in that space as well. Because as soon as that stuff comes out, then everybody knows right? I mean we need to kind of the secret of some of that needs to be kept just for them so they can do certain things with it. So I to Christian, as you think about this, you know, just on the crypto side and not cryptocurrency, but crypto, the cryptography, the securing of things right. I think it’s gonna it will grow as we get more, you know, this is the way it’s been happening, right. I mean, think about some of the early days of encryption. Yeah, 15 years ago, maybe. Right? And it’s like, now we’d laugh at that, right. Some of the cryptography that are thumb some of the things we expected to keep things secure. It’s had to kind of change, right, change and morph. How have you seen as you’ve watched that, and, you know, we’re just talking, you know, 10 years of podcasting together, how have you seen cryptography change in the 10 years, you’ve been Following.
Christian Johnson [16:01]
Yeah, I mean, certainly a lot of bugs, which has been kind of interesting because we have claimed they are such a fundamental, quote unquote, plumbing of the internet to see bugs like heart bleed. As a prime example the so predominant and implementations kind of shows how much effort it continues to take to turn over these code bases and to fully enumerate them. We’ve also seen a lot of encryption standards break over time as the technology has improved, right, like Duffy Hellman, not exactly what you’re seeing in production anymore, right. So, you know, a 512 bit RSA certificate, which by no means is used in any deployments today can be cracked with enough time and patience on a conventional non supercomputer application using general saving techniques which try to anticipate what’s polynomials are going to be before computing them. So you’ve seen bits move up, right? Like your average SSL certificate today, it’s 2048 bits 1024, generally untrusted. You’ve seen algorithms fall by the wayside, like how many of you still have shot one certs? No one, and those were only killed a year or two ago. And so I would say what we are talking about with quantum we’ve seen on a micro scale over the last 10 years as the infrastructure and computation and plumbing of the Internet has gotten fatter. The crypto standards have needed to get more robust. And those are both from a security and correctness of the theoretical implementation as well as higher bits, better padding mechanisms, etc, etc. So even with AS you’re seeing, you know, as CBC was acquired gold standard for the 10 years that you and I are talking about and now only in the last six months it was funny is actually a a average guys side tangent was I think it was Joe we were talking to in the discord channel a couple days ago, and we were running cipher suite tests against the web servers, which is interesting because a typical web server will offer multiple cipher implementations so that different browsers and different devices can use quote, unquote, their preferred mode of cipher exchange. And so part of good operational security of any web server is that you don’t offer algorithms that are inherently insecure because the idea is that if a client connects to you, and they use that dedicated or insecure algorithm, they can then you know, make attempts at compromising your your keys or your customers data etc. etc. So back to the good upset the idea of any production web server deployment is that you want to limit it down to the smallest set of ciphers possible and you want those ciphers to be new, modern, etc. So as CBC 256 has been recommended, approved, great standard for, you know, last few years and now you go and review run a quality service or any type of other SL SSL test. It won’t be new, but it will say, hey, here are the specific interested. The specific tests that were run and specific instances where CBC proved not to be as exciting as people thought. And it was specific to an implementation with Oracle databases I believe of I need to go back and do some fact finding on that.
Christian Johnson [19:52]
But that aside, we really would not have expected that algorithm to go away so quickly. And so GCM and some of the other ciphers for AS are now taking over. So while AS is still the base encryption, yeah, there’s different implementations kind of coming and trickling their way through the pipeline. So I would say implementations have changed, bugs have gotten caught. Cypher sizes and bit sizes of certificates has gone up. I think the core concepts surrounding confidentiality, integrity, availability have not changed. I don’t expect those to change anytime soon. And those are just what customer needs are. But yeah, we’ve definitely seen a shift in how robust these things are and how robust they will be and that, you know, do I have confidence that there’s not going to be another Cypher problem in a year to not really, I think 10 years of precedent shows that it’s a shifting target. Much like Do we have confidence that TCP IP or other based OSI transport mechanisms are inherently flawless. No, but does the entire internet use them? Absolutely. So there’s definitely a it’s, it’s much like how Intel CPUs are built off of hardware that turned out to be fundamentally flawed at its core with speculative execution. And now, we can make somewhat of the same comparison to, we built this entire thing called the internet over the last 40 years. And the reality is that the OSI model is an imperfect world. And on top of that cryptology is an imperfect world and oh, on top of that, certificate authorities and how we really delegate trust is an imperfect world because trust is a human construct, right? So it’s a it’s, it’s almost a, a implies be that if it if it’s a human construct, and it therefore can be foiled in some way, and I think certificate authorities are under example where it’s much more on the human side than necessarily on the machine side where foiling can take place. So it’s moving, it’s improving, but I think we’re seeing on a micro scale, what we would expect to see if quantum started really challenging this space.
Jim Collison [22:16]
About three or four weeks ago, I was logging in to the average guy.tv. And I got a security warning. That was like, Hey, you know, the site and I was like, okay, we this certificate must have expired, right? You know, you get that warning from Google. So I pinged you, and you immediately got back to me, and you’re like, Oh, yeah, it’s been five years. We need to we need to update it, when you updated the site. And when you think about the future of this, it’s one did things change in the five years since you had put this certificate in, in effect, and then will there be Do you see over the next five years anything changing just a certificate on a website?
Christian Johnson [22:57]
Yeah, you know, that’s a that’s a really good point. I would say that five years ago, you had to, you know, it was a premium paid thing to get an SSL certificate shell over $30 for whatever the time frame was, etc. And now we’ve moved to, you know, I go back to the certificate authority. Now we’ve moved to these temporary certificates where every three months I renew and extend the certificate, it’s free. I use things like Let’s Encrypt, which has become wildly popular. And so yeah, you can still go by and pay for the premium, huge certs. But ultimately, the movement for everyone being on SSL has definitely been a paradigm shift. I mean, browsers and coded showing you big red banners if any website didn’t have SSL starting about two years ago. And these are things that people kind of forget now looking into looking forward into 2020 but you know, we go back five years, browsers to tell you if your if your password is being sent in the clear, you got I had no idea what the certificate issue or was of the certificate authority, and they weren’t as easy or as accessible to the average user. So
Unknown Speaker [24:09]
definitely.
Christian Johnson [24:13]
I don’t know if I’ve answered your original question, the Delta there and just how one goes about accessing crypto has become much more consumable for the average guy. Another example is with VPN, right? Like everyone left and right wants a Private Internet access or VPN package for 10 bucks a month for some type of mobile phone endpoint security. And so these are things that people just a weren’t asking for as much five years ago, but BR now, like rapidly available for the consumers at a cheap price.
Jim Collison [24:48]
Yeah, I think I’m paying 30 a year to use bit defenders VPN and it’s available on my phone I can any software that’s running on my computer, it’s available. So not just Not a device. It’s just whatever I want to install it on. I can use their VPN services to kind of make it work for 30 bucks a year. Unlimited no data, like it’s it’s it’s pretty amazing. On the flip side of that Christian is we think about the next five years. Do you see in that area of certs, do you see something or anything coming that will fundamentally change or even just tweak, right? Nothing ever just I shouldn’t say that. Because we feel we have seen some massive overnight changes. But you see anything coming that may that may change that process or make a difference?
Christian Johnson [25:35]
Yeah, that’s interesting.
Christian Johnson [25:38]
Nothing in the one to two year horizon. I think by and large, you can continue to expect some of the trends that we’ve talked about with certificates and algorithms and implementations in terms of what the consumer has access to paradigm shift. No, I think it a paradigm shift in privacy and overall ability of leveraging crypto and day to day life will come in two areas. One will be a more decentralized internet. So you’re kind of seeing two things happen. One is like whole parts of the internet are testing their own private firewalls, which is kind of interesting. So it’s like Internet except when we don’t feel like it. And then it’s an intranet. Can you give an example of that? Yeah. So there’s an article that had come out recently that the Russian government tested basically blacking out their country and going to, quote unquote, servicing all their web traffic internally in the country. So all their DNS resolves had to fail over and all these other kinds of backdrops needed to take place for them to quote unquote, go on and off the internet, which is kind of interesting. There’s a ton of other interesting research being done right now in general censorship around the world. China being one of them, because they, you know, inject all sorts of funky things into packets. And there’s plenty of academic papers about what they do. We call it the Great Firewall of China, right? But like, what do they try to keep out by default? And so based on what part of the internet you’re on, you might have a very different experience. And so you could start to see some of that behavior. But where crypto comes in is that if the internet starts becoming more decentralized instead of the gatekeeper model, I would expect that crypto would continue to play a bigger role in providing Ethereum like services to the broader Internet, and so I would expect crypto have a big impact there. I also expected to kind of do the natural plumbing things like continue to make financial transactions easier. So I think about the apple card is a good example of this right? Was there some breakthrough in cryptology that enabled Apple to come up with a sleek, real design a credit card? No. But does the average consumer still see it as some great sign of progress? insanity? Absolutely.
Christian Johnson [28:09]
So there’s a little bit of fluff factor there as well.
Jim Collison [28:12]
Oh, some really good marketing does, you know so as it is, of course, as equipment continues to get better, we see this all the time, right. And lately, I think hardware is kind of flown under the radar, even among tech enthusiasts. It doesn’t have the same feeling today as it had three to five years ago. Like in our circles, we used to talk specs and all those kinds of things all the time. And that’s largely dropped off in a lot of ways. Maybe there’s other groups doing that, but hardware is definitely continuing to improve. Does this change when we think about crypto now I’m thinking about cryptocurrency with the blockchain as we continue to see hardware improvements and those blockchains being able to take advantage of faster processing or faster My name so to speak. Does that in your opinion? Does that change anything behind the scenes? I mean, we’ve moved from GPUs to a sick miners to specialized equipment now just for Bitcoin mining, Bitcoin is not falling off. Like I think everybody thought in fact, as we’re recording this right now, about nine, the 404 block for Bitcoin right now, which I think by now everybody thought maybe well, you know, is it 3000? We were kind of laughing at it. And that’s still pretty good. I mean, 3000 is still pretty good. Sorry, did I invoke? I think I invoked your, your digital assistant there. But what do you think is we think about the hardware and the how does that play on the blockchain and some of those kinds of things?
Christian Johnson [29:44]
Yeah, pretty interesting. I would definitely say that we would need to see some type of fundamentally stable quantum processor in order to jump into the next realm of whatever Bitcoin mining looks like in a non traditional platform immediately. affordable rate. I mean it, but just in theory. Yeah, I mean, some of the interesting things that Google calls out and they’re looking forward is supporting generative machine learning as future use cases, which is consistent with the narrative that I illustrated earlier in the show. The second is they described very specifically building quote, fault tolerant quantum computer as quickly as possible. So they’re looking for general purpose devices here at the end of the day the same way you would with a classic computer I think that’s the ultimate kind of holy grail of where people want quantum to be. And then of course, they paint you know, use cases like lightweight batteries for cars, new materials, airplanes catalyst preferred, I mean, the whole thing opens up right but at the end of the day is because people are able to crunch different types of data or move different types of things quicker with the development of blockchain cryptocurrency I don’t think Expect quantum is going to impact it dramatically in the next one to two years, I would expect at a minimum, we’re talking five years here. And it would have also begun some assumptions that we have some type of watershed moment, like what Google did here about once a year. Moving, you know, dramatically and linearly in that direction, I think short of that, you’re going to see things that try to emulate hybrid solutions, right. So maybe a GPU is most efficient for this type of computation. But then a basic CPUs most efficient for this part of the computation. So rather than a only situation, it’s not like a and or, but it’s an it becomes an sorry, becomes an indoor situation, right? So you’re not exclusively locking yourself into one type of computational platform for your overall workload. You’re basically pipeline in your workloads to be symmetrically aligned to the best processor that can support that particular computation. And so potentially, maybe you start to see things in cryptocurrency, where those things get partitioned out in order to iterate without having the next big thing. Typically, that’s what you see, right? Like new capabilities come out, we then refine those new capabilities. And then we implement them in multiple tiers of buckets so that we can best align what the computation is to what the capabilities are that are available for it.
Jim Collison [32:27]
Well, in maybe an application in the machine learning space, any I would throw those two together sometimes even though they’re different, but to think through what scenarios run best in run and what environments and then be able to dole out those processing jobs, not from a, I’m going to program that to happen, it’s going to it’s going to begin to learn in this sense of like, Hey, I can I can test these things out and I can figure out where what’s best and when and then be able to make some Decisions that’s just across to, but there could be multiple ways of, you know, we think about a chipset, where it begins to distribute that load out kind of based on its knowledge of what runs best in what situations down to the calculation level. And that all of a sudden, that makes machine learning, I think, kind of really interesting in the sense that now we have kind of distributed computing, even in a single data center. And, and so that brings some I think that just brings some interesting thoughts to I still think we struggle a little bit with AI and what everybody wants to call it that but and we’ve talked about that on previous shows, but
Christian Johnson [33:37]
well, and density of computation is going to become hugely important, right? I mean, hard drives illustrate this quite brilliantly. If you built a data center five years ago, you were going to be populating it with all two terabyte spinners. Now, if you build a data center, or you read provision all of your hardware in place for what the same size that that data center is in terms of square feet, you’re populating it with 10 terabyte drive So, now you have five times the capacity of 500% increase with zero percent change in physical space. You know, Moore’s law has continued to be artificially extended for the last decade. To the point where the AMD CPUs that are out on the market today are seven nanometers, I mean, who would have thought seven nanometers for a general purpose consumer CPU that you can buy off the shelf for a couple hundred bucks. So I expect that level of computational density to continue to increase and it might not be purely in hardware. It might also continue to be software opposite optimizations or other general purpose computing techniques get better. But certainly, you’re you’re looking at multiple moving horizons here of denser storage, more compact superconductors and computational ability. Larger pipes to move the different data silos or different computational locations that might end up specializing in different parts of your overall needs. And so the result is that it really comes down to the three basic utilities compute storage network. And that I think, is going to drive us in the interim of, you know, quantum coming with a giant hammer and somehow, you know, redefining the game for us. And I think we’re still well on our way to seeing dramatic improvements and artificial intelligence and machine learning purely based on the fact that utility cloud computing is still moving in that direction of just hyper growth, hyper performance, and always kind of outdoing itself, you know, iteration to iteration.
Jim Collison [35:47]
Well, we saw these gigantic in response to crypto cryptocurrency. We saw these gigantic data centers being built with GPUs. Yeah. And that was not like we that had not been in traditional data set. are set up in and I remember the you know, the tech industry can make it a big deal about it about Hey, now you can do you know, now you can come and do these GPU, whatever computations are fast on that it’s super fast and math. And so we’re beginning to see now hybrid versions of that is Microsoft in, you know, again driven by gaming but I think there’s lots of applications around this beginning to move their gaming platforms to the cloud. And I think that whole idea then now because the fundamentally changed when we think about processing as a service, and what is that or computation as a service? What does that mean when I have different you know, we just mentioned three here in the last two minutes of CPU calculations, what’s happening there? We thought about a sec this specialized, how do we build a specialized chip? In a way an iPhone is a little bit like that, right? I mean, Apple has kind of customize that chip to be the way they want it to be. But let’s say a six when we think about what’s being used in crypto and the way it’s doing cryptocurrency and then they buy GPU space. And yeah, those are all forms of kind of similar ideas, but they are all specialized now and highly specialized and some of the way they’re doing calculations. Imagine if we now on top of that, we got a little bit of quantum But imagine if we have a control that is beginning to, on the fly, begin be able to send those computations based on what they are and what they’re trying to do to get the maximum speed. If that’s what’s important. In some case, maybe it’s not, I mean, does all the world’s computation need to be done at the fastest possible speed? does all the world’s storage need to be done on the fastest possible storage? Right? I mean, we, we kind of think this way, but economies of scale means I think in the future. Things can be decisions the machine will make the decision for us kind of based on our usage of where does it get stored the most expensive fastest or can it go to cold storage? We humans make that decision today, right? When we do backup, we choose to cheaper, slower, more stable, maybe storage, the humans do that. Maybe in the future that is a machine learning or an AI service for us that kind of begins to learn what we do and then directs that into the right places for us to make sure it’s available in whatever form we need kind of based on our own internal SLA of like, when am I going to need that data? And can it sit in other places? So I don’t, I really don’t feel like it always has to be the fastest. There’s always sometimes when that’s appropriate, but certainly we’ve known from just storage but I mean, think of think of some of the SSD and some of the future of storage as well. Quantum Computing has taken the lead I think on on the thought leadership of in the tech world, but man storage is making some amazing advancements that don’t don’t get a lot of necessarily get a lot of talk with. You mentioned a 10 terabyte spinner. Well, okay, that’s just a variation. The two terabyte of the six terabyte of the 12 terabyte, right? And yet there’s some greats kind of some cool storage things coming when we think about what they’re doing in that space to be able to jam a bunch of those things in there and get them back out fast. So anyways, I say all that to say, I think there’s some really cool things. When we think about those things. Can you add anything else to that?
Christian Johnson [39:20]
Yeah, definitely. I mean, it’s it’s kind of interesting that it’s not only just thinking about what the infrastructure is computing for you, but also turning the computational and the algorithmic side back on the infrastructure to optimize the efficiency of the infrastructure. So it’s somewhat of a self self fulfilling prophecy.
Jim Collison [39:39]
It’s the beginning of the machines taking over when they can optimize themselves, right, that’s really in that in that regard. It’s really, you know, could they and then, you know, using some robotic mechanisms, could they add to themselves kind of based on their needed Yeah, that’s the beginning of Skynet, I think is as we joke about that does this Christian does this add anything? You know, we get crypto updates from you from time to time you haven’t that the show is notoriously crypto week sometimes cryptocurrency that is but any thoughts on anything? You know the blockchain now is being tested in a lot of new ways. There’s you know, any additional thoughts
Christian Johnson [40:26]
I was amused to see that the Bitcoin numbers jumped from like seven K to 10 k because China has now indoors Bitcoin is here to stay so it’ll manipulation. So apparently we continue to have weigh ins on whether it will whether it won’t whether it will, whether it won’t so I suppose with your Ace speculative investor.
Christian Johnson [40:53]
This is an interesting roller coasters for you to ride on and why the direction that you’re on your own successful apparel.
Christian Johnson [41:01]
I don’t have a great sense for how long it’s going to take for us to quote unquote, production allies the thing like it is still very much a geeks paradise and not very much an average guy. platform, opportunity point of entry, etc. If anything, I expect that we’re still in that kind of wonky timeframe of we just have to wait and see. Because I think you’re going to continue to have to wait for governments to pick sides on this and then derive to some kind of we need to go paperless and we can’t have any paper currency anymore. And well, how are we going to do that? And then someone’s going to make some argument for why credit cards are not the correct way to scale that and then blockchains going to be everywhere. I also think that as smart contracts start to become more legally enforceable and verifiable with artificial intelligence, there might actually be better use cases for putting blockchain in the middle of things because now you can have the kind of digital validation aspect of a contract and take the human element out of it. And that’s very interesting to me. I think that’s where there’s a lot of promise for the technology still is that it has the potential on the upper bound to scale really well for what I call high quality validation. Right? And so, can you trust the human to always properly be able to legally enforce in 1000 page contract across maybe hundreds or hundreds of thousands of customers? Probably not. But a machine can certainly do, it can certainly do it very well. And the thresholds that would be interesting is automated action versus action that requires human review. That’s an area we haven’t even traversed yet, but I would expect To start coming into focus, as financial institutions really grapple with, this is pretty much a technology that’s here now. It’s not even that it’s coming. It’s here, but what are we doing with it?
Jim Collison [43:14]
In a Forbes article, we will put in the show notes. what’s kind of interesting Detroit revs its blockchain engine is the title of it. For BMW, kind of behind the mobility open blockchain initiative mo VMOBI. took to the stage talk about the blockchains potential to transform the auto industry. What’s really really interesting in this is again, no though cryptocurrency side with abusing the blockchain to check, validate and track every bit on a car from the second it’s manufactured to it’s being put together. Then the life of the car itself in a sense, I could Christian I could know every part where it came from how it was put together, who Put it together. My car could be a blockchain in a sense, right or track on blockchain, right to think. Well, you know, interesting enough, I was out doing some work on my hand and I have no 600. So it’s an older one. It was new at one point in time, but it’s showing some age. And I was working and trying to get my headlights replaced. And then they have Honda’s buried those deep in the corners. And it was not like the old days, you take a few screws off and put the lights in and you were done. Right? This was pretty hard to get to. But it’s missing some is missing some screws in there and I’m, and I’m like, Okay, well, I have to go online and order them and whatever. But how interesting it would be. If I was thinking along the lines of my maintenance records and some of those pieces and where they were manufactured, where I would get them from, if that was individualized to my vehicle, and I could kind of keep track of it. I do think the blockchain is a really good way to do that. It’s you know, when we when we think about the Y or the what’s behind the blockchain That’s another one of those, you know, like property records. Those are another one of those. When you think about my housing record, there’s no reason in the United States property records are kept at the county level. Like it’s a horrific Lee awful system. That is very inefficient. It doesn’t work very well. It could be nationalized and streamlined. It using the blockchain, is that the right thing? You mentioned this a little bit earlier. We need some precedent. We need some legal precedent of some things to happen before we can make some of these decisions, right. The courts need to start working through some of these things to make sure because you just can’t go wholesale in Okay, let’s do this thing. We’re going to need some legal precedent. But I do I still believe there’s some great implementations and I think, in this sense of on the other vehicle, being able to track my vehicle would have a that’s the way I would track it, including every single part that was put on it. I don’t think it’s a bad example.
Christian Johnson [45:56]
Yeah, now well, it kind of reminds me of The analog case of this, which is I think of ham radios that are they work with the what’s the name of the system?
Christian Johnson [46:12]
I should be I should be disciplined for not remembering this Wieger needs to be here to
Christian Johnson [46:16]
Yeah.
Jim Collison [46:18]
I don’t know either
Christian Johnson [46:19]
the name of the
Christian Johnson [46:21]
Gosh.
Christian Johnson [46:24]
It’s essentially a system that automatically reports out the location of a particular quote unquote, ham radio site or a cell site. And so people install these transceivers in their cars that do nothing but emit a analog to them converted to a digital frequency that plops them on a map so you can see little cars moving around and stations and all this kind of ridiculous stuff. I am very depressed that I’m ARPs. That’s what the protocol is not
Jim Collison [47:00]
Not the method, not a RP, the retirement group
Christian Johnson [47:04]
but a prs. So ARP ARP is like not like your art packet, but a PRS, which is a packet protocol tracking system that’s specific to these ham radio relay relay. So if anyone’s ever curious about this, you can go to your browser and type aprs.fi is the website and it’ll kind of drop you to where your nearest geographic areas and you’ll get kind of this map of all the pinpoints plotted on Google Maps, you’ll see all of the different radio towers, transmitter stations, things marked as houses, cars driving around, and these are basically analog ham radio devices emitting their callsign to the nearest relay station and it gets converted into a digital signature and signal and then stored on a map like this. So, I only bring this up because it’s kind of really funny when you read a callsign like KN for YPM dash nine, which is someone’s car somewhere in Virginia, driving at 54 miles per hour and 67 with a with an altitude of 79 feet and transmitting on a frequency of 146.74 or five megahertz and it just this reminds me very much of an analog way of like, oh, how would I track cars? Well, apparently we’ve had this system around for decades that can do it and no one talks about it. But now we want to use the newfangled you know, sexy thing called blockchain to do it for all cars. Yeah, or Tesla and cell phones and all these other different ways to piggyback signals, right. Sure. One of the takeaways though, is that it becomes a very decentralized, fast, potentially cheap medium to do. at scale. And I think right now, you look at even the cell towers and other types of implementations, there’s when we talk about fidelity of the metric, there still is not kind of a, what I call a digital grid, right? Like we have a United States grid of electricity. We have the United States grid of like, internet, plumbing and all the other things, but we don’t really have what I call a data grid. And that data grid would be something that, again, traverses the same global surface area, but allows for with very little bandwidth and very little compute power to emit tons of statistics. And I guess most people are going to be like, Hey, isn’t that what IoT supposed to do? Well, kind of, like IoT is a device. It would be a great candidate for maybe emitting some of those metrics. But when we talk about everyday devices that can emit with very little fidelity, like today, you have to have a cell phone plan to do that. You have Some type of carrier signal signal coming in, there’s a lot of like prerequisites for just getting a device that can be have service, right? Whereas, you know, you can go anywhere you want in the United States and turn on an FM radio and get the quote unquote data plane of FM radio, no matter where you are, what altitude you’re at, etc, etc. All you need is a very little amount of power. And you can listen to a lot of data. And it’s analog, right? What does the equivalent look like for that in a digital currency digital blockchain world is a fascinating concept. And I don’t think we’re anywhere close to it, but I don’t think that the technology is missing to do it. And it would have huge implications on the type of data models that could be built and collected.
Jim Collison [50:50]
I just want to know how you remembered this website. EBRSFI. That’s super random Christian.
Christian Johnson [50:57]
Yeah. I’m weird. Like that.
Christian Johnson [51:01]
actually spent part of my time at DEF CON, getting my ham radio license, certificate and all things weird. It’s mostly become a point of fascination for me to like, instead of, you know, charging into cyber frontiers of the future I like kind of want to charge into the past and see if we’re reinventing the same things.
Jim Collison [51:21]
Do you have a hand? Did you buy one?
Christian Johnson [51:25]
Well, that’s interesting. So have I bought one? No, but my interest in ham radio is predominantly because of my interest in software defined radio, which is the kind of cyber frontiers newfangled thing of what people did 40 years ago with ham radio, and you’ll kind of hear the quote from all the old timers that ham radio isn’t what your grandfather used 40 years ago. It’s cool and a tip. Well, when you ask them why it’s cool, and it’s epic, because oh, software defined radio. So when you look like that, a lot of the things that are on prs as a website, a lot of these things are being supported with software defined radio to the point where like, you can build a cheap little Raspberry Pi that connects to an RTL dongle and be emitting these things while driving down the highway. And that’s an example of something that’s low fidelity, cheap economically and it has high data impact. So I, I don’t really do much in the way of transmit, I play a lot with different RTL dongles and and related chipsets and trying different listening patterns and really honing the radio art, which is what you’ll read in any ham operator model that the reason the FCC allows for amateur bands is to quote unquote, perfect and mature the radio technical expertise in the country. And so there’s some of that with SDR as well. There’s also a lot of interesting stuff you can do and receive satellite wise Like CubeSat type stuff and other low Earth orbit stuff that you can get with real cheap dongles. So in all it’s it’s a whole new space of just wacky out there and it’s kind of funny because for some reason analog excites me as much as digital does so I’m I’m regressing I’m going back into analog and some of my my free time
Jim Collison [53:24]
learn some things of the past if you if you go to that site and you click on stations currently moving just kind of interesting right? And then you can you can find some really interesting like I’m finding some that are moving at hundreds of miles per hour like so and moving at 269 miles per hour.
Christian Johnson [53:44]
Yeah, someone like like some calibration about that too big to have a rocket on earth that we
Jim Collison [53:50]
don’t know but it’s like a station. This is the funny part. It’s like a you know, it’s marked as a like, like a physical station, but it was moving at 235 An hour or whatever, and you’re like, well, but no Christian. I mean, I do think it’s interesting. Like, you know what’s old is new? Yeah. And it’s interesting. You’re finding some, you’re going back to it to kind of learn some things. What did they learn about this? And how can we apply that to the new
Christian Johnson [54:16]
right? Yeah. And what’s interesting is that like, a lot of the repeaters stay active now through a service called echo link, which is analog to digital conversions, right? So if I want to open up a repeater flap out in like Seattle, Washington, all I need to do is use an echo link service to digitally Connect myself to the nearest repeater out there and go from digital to analog and vice versa. So it’s very interesting how everyone is so buried in their iPhone and whatever their digital devices they kind of forget how many things still operate and function at the analog scale of things like Oh, did you stop and think that yes, your digital receive signal to your cell phone is going over a quote unquote analog surface medium called a wave. So really understanding those principles can help you kind of understand, you know, have a better appreciation for number one where we are with technology today and realize some of those cases where what it was 40 years ago is now being implemented in, you know, with with new and different fidelity of technologies.
Christian Johnson [55:24]
So
Jim Collison [55:25]
there’s a call sign off the southern coast of Australia right now going at 199 miles an hour.
Christian Johnson [55:32]
Nice. Nice. Wonder what that is, since the rocket, super speeder
Jim Collison [55:37]
boat. No, it doesn’t. Yeah, well, that’s an odd speed like that. Oh, you know, ocean travel, even land travel at 200 miles an hour unless you’re on a train. I suspect
Christian Johnson [55:48]
some of them must not be calibrated correctly. Do you think
Jim Collison [55:50]
you think that’s what it is?
Christian Johnson [55:52]
We can think of many things that would be reading accurately.
Jim Collison [55:56]
Yeah. Well, you know, if you had one on a plane, maybe it says Maybe it’s a maybe it’s a small
Christian Johnson [56:02]
one. If you have one transmitting on a plane, I think you’re going to have other problems called the FAA.
Jim Collison [56:08]
Yeah, but you never know. I mean, southern coast of Australia, you know, they’re down there. But what travels at 200 miles an hour? I mean, it would have to be a kind of small small engine aircraft or something like that if that’s indeed a real you know, a real here’s this one traveling ces 269, which is interesting. It is up in or is this out here? Western Canada. I’m now now I’m hooked. I’m gonna I’m going to be checking out this site.
Christian Johnson [56:38]
This is a cool site. If you’ve ever wondered why you should get back into ham radio or analog things, this is a gateway drug. Sounds good. Christian. Anything else? No, I think that’s a wrap. I dangerously broke back into analog on a digital show.
Jim Collison [56:53]
No, I like it. I like it on the homepage on the Home Gadget Geeks side you know, Wieger has gotten into ham radio. Yeah. So This is this has been a topic of conversation over on that side, we probably need to had a guy kind of lined up to talk about some in depth ham radio. He was busy. So we stopped to get back to that. But
Christian Johnson [57:12]
yeah, we might we might need to have a bridging of shows at some point. Yes. There’s a ton of data dump that we could we could talk about here from random random sponsoring here in the last six months. Yeah. I will say as a looking forward show announcement, have a couple guests for a couple different shows that we’re going to try and line up for the end of the year. So
Christian Johnson [57:33]
good look out for that.
Christian Johnson [57:34]
And hopefully we’ll get them done before the end of the year.
Jim Collison [57:39]
Joe says you could you could cross look up on the flight tracker to see you know if you wanted to see although those Shouldn’t you wouldn’t necessarily want to see a ham signal. I would not know
Christian Johnson [57:52]
that most I would want to see your a cars, airplane box pinging at my at my receiver what it’s called. sign is but
Jim Collison [58:01]
right you want to seem in vehicles and things like that. So we remind everybody, of course, cyber frontiers powered in the average guy.tv powered by Maple Grove partners get secure, reliable high speed hosting really high speed from people that you know and you trust and plans start at 10 bucks Maple Grove partners. com If you’ve got ideas, show ideas, guest ideas, whatever Christian loves to hear about those things. So send them an email Christian at the average guy.tv you can track him down on twitter at Borg whisperer and and he’s already kind of previewed what’s coming up I’m excited about it. We will have some new things on cyber frontiers. We are alive whenever we feel like it it’s just basically tonight we lie literally paying Christian on discord today. Tonight be a good night. Let’s do one of these and he said yes. So you want to follow me Follow me on twitter at j Collison is what an all announced and it’s always last minute so just make sure you subscribe to the podcast player and then don’t unsubscribe and we put use one of these every couple weeks. Just want to thank you for joining us tonight Joe.
Jim Collison [59:07]
And with that, we’ll say goodnight everybody.
Christian Johnson [59:09]
Goodnight
Transcribed by https://otter.ai
Contact Christian: christian@theaverageguy.tv
Contact the show at jim@theaverageguy.tv
Music courtesy of Ryan King. Check out the Die Hard Cafe band and other original works at:
http://diehardcafe.bandcamp.com/ / http://cokehabitgo.tumblr.com/tagged/my-music
