1 00:00:00,000 --> 00:00:03,640 Did you ever think you would work in a tense geopolitical 2 00:00:03,640 --> 00:00:07,160 or adjacent to a tense geopolitical space? Because I certainly 3 00:00:07,160 --> 00:00:10,860 didn't. I mean, how far 4 00:00:10,860 --> 00:00:13,200 back do you want to go? If you say a year or 10 years ago, 5 00:00:13,200 --> 00:00:16,880 probably not. Yeah, that's what I mean, like 10 years ago. I mean, I don't 6 00:00:16,880 --> 00:00:20,240 think— because I remember, and living in DC, you get a different 7 00:00:20,240 --> 00:00:23,880 perspective on the world than you do 8 00:00:23,880 --> 00:00:27,260 anywhere else. Candace and I both grew up in New York, and you get a 9 00:00:27,260 --> 00:00:30,020 very different perspective there. But since I've lived here now 20-plus years, 10 00:00:32,244 --> 00:00:35,694 the thinking here is different. So when I talk to colleagues and 11 00:00:35,694 --> 00:00:39,305 even— this would've been 2, 3 12 00:00:39,385 --> 00:00:43,108 years ago, I said, you're going to see a company like Nvidia treated like a 13 00:00:43,140 --> 00:00:46,462 defense contractor. Right. And they're going to have export controls. 14 00:00:47,072 --> 00:00:50,827 And my former colleagues who are based in Raleigh and in 15 00:00:50,923 --> 00:00:53,892 Boston, and if you're clever and you can figure out how to use LinkedIn, you'll 16 00:00:53,892 --> 00:00:57,438 know exactly what company I'm talking about, thought I was insane. They 17 00:00:57,438 --> 00:01:00,780 thought I was tinfoil hat guy, right? 18 00:01:01,541 --> 00:01:04,796 Yet what have we seen over the last, what, 24 months? 19 00:01:05,913 --> 00:01:09,720 Yeah, something pretty close to that. Welcome 20 00:01:09,980 --> 00:01:11,365 to Impact Quantum. 21 00:01:23,447 --> 00:01:26,929 Hello and welcome back to Impact Quantum, the podcast where we explore the 22 00:01:26,929 --> 00:01:30,638 emerging industry that is quantum computing and what it means 23 00:01:30,702 --> 00:01:34,196 for the world at large, because as 24 00:01:34,308 --> 00:01:38,154 it becomes more and more closer to being real, and don't send the hate, it 25 00:01:38,154 --> 00:01:41,888 is very real today, that we're gonna have an interesting conversation today. But 26 00:01:41,920 --> 00:01:45,430 first, I wanna introduce the most quantum curious person I know, 27 00:01:45,510 --> 00:01:48,859 Candace Cahootie. How's it going, Candace? It's great. Thank you for 28 00:01:48,955 --> 00:01:52,754 asking. I'm really excited for today. We've 29 00:01:52,802 --> 00:01:56,151 been talking more about post-quantum cryptography lately, 30 00:01:56,888 --> 00:02:00,238 and today we have the opportunity to talk to an 31 00:02:00,446 --> 00:02:03,928 SME in post-quantum cryptography who is 32 00:02:04,057 --> 00:02:07,829 leading the efforts at KPMG. His name is Christopher 33 00:02:07,974 --> 00:02:11,811 Turk. How are you, Christopher? I'm doing great. Thank you 34 00:02:11,811 --> 00:02:15,343 for joining us today. Also, welcome to the show. Welcome to the show. 35 00:02:15,584 --> 00:02:19,405 Also a fellow resident of the DC metro area. We were swapping 36 00:02:19,453 --> 00:02:22,985 stories in the virtual green room about Northern Virginia and data 37 00:02:22,985 --> 00:02:26,758 centers and things like that. Although you live— you don't live in Data Center Alley, 38 00:02:26,822 --> 00:02:30,440 you do live next to it. But not close enough to hear the, the AC 39 00:02:30,440 --> 00:02:34,146 units going. So it's probably not a coincidence that 40 00:02:34,146 --> 00:02:36,739 you're in the post-quantum cryptography space 41 00:02:37,510 --> 00:02:40,113 and you live in DC and not where, 42 00:02:41,382 --> 00:02:45,061 not any other city in the world. I would imagine that's not a 43 00:02:45,061 --> 00:02:48,823 coincidence. I would say that it's not for sure. 44 00:02:51,110 --> 00:02:54,920 Military, or at least some in doing cybersecurity as well. And 45 00:02:54,953 --> 00:02:58,003 NIST is obviously a big part of, we could talk about that. She doesn't work 46 00:02:58,003 --> 00:03:01,338 in the quantum cryptography stuff or certifications. 47 00:03:01,835 --> 00:03:05,556 Though, funny story, when my oldest was at the daycare 48 00:03:05,556 --> 00:03:09,179 center there, one of the other parents— I didn't know this 49 00:03:09,292 --> 00:03:12,643 until they had for Mother's Day, what does your mom do? And 50 00:03:13,140 --> 00:03:16,587 one of the moms said she, she was a researcher in post-quantum 51 00:03:16,587 --> 00:03:20,211 cryptography. This is 10 years ago. This was like, 52 00:03:20,837 --> 00:03:23,915 now it's, I wouldn't say it, it's not like AI where everybody and their dog 53 00:03:23,915 --> 00:03:27,733 is an expert in it. But, or thinks they're an expert in it. But post-quantum 54 00:03:27,733 --> 00:03:31,247 cryptography used to, was a very niche topic 10 years 55 00:03:31,359 --> 00:03:34,873 ago. And so it's very 56 00:03:34,969 --> 00:03:38,178 interesting. This, I think this is one of the best places to be if you're 57 00:03:38,178 --> 00:03:41,837 in the PQC space, because the government saw this pretty 58 00:03:41,885 --> 00:03:45,398 early. Before I even knew about quantum computing, I was sitting at a public 59 00:03:45,398 --> 00:03:49,105 sector event at Microsoft and the, the, one of the 60 00:03:49,105 --> 00:03:52,570 speakers had said something to the effect of, yeah, this may be 10, 20 years 61 00:03:52,843 --> 00:03:56,519 away, This was 10 years ago. So 10, 20 62 00:03:56,616 --> 00:04:00,373 years as a forecasted date was not unreasonable then. He said, 63 00:04:00,550 --> 00:04:03,825 but think about how slow government systems move, government IT moves. 64 00:04:04,564 --> 00:04:08,401 You wanna start today. So with that in mind, where do we stand now? Because 65 00:04:08,498 --> 00:04:11,356 quantum is as close to being able to 66 00:04:11,966 --> 00:04:15,450 break conventional encryption now than it's ever been. 67 00:04:16,381 --> 00:04:17,618 What's the temperature of the room now? 68 00:04:19,850 --> 00:04:22,933 So when I think about the temperature for 69 00:04:23,351 --> 00:04:26,949 post-quantum cryptography, I think I look at it from 2 different lenses. 70 00:04:27,559 --> 00:04:30,803 So one, there is a legislation lens pushing 71 00:04:30,899 --> 00:04:34,658 mandates. So I would say that's pretty hot. It's pretty 72 00:04:34,690 --> 00:04:38,046 heavy right now. We are seeing a lot of mandates come across out of the 73 00:04:38,046 --> 00:04:41,644 White House. We are seeing a lot of strategy being 74 00:04:41,676 --> 00:04:45,338 developed at the policy level. We know that NIST is already approving 75 00:04:45,916 --> 00:04:49,238 post-quantum safe algorithms for 76 00:04:49,302 --> 00:04:53,122 cryptography. So we're seeing a lot from, let's say, legislation policy. 77 00:04:53,860 --> 00:04:57,407 When I see, when I look at PQC from a 78 00:04:57,407 --> 00:05:01,179 different perspective, meaning industry, and when we 79 00:05:01,179 --> 00:05:04,999 say industry, we know that government, they're very focused on 80 00:05:05,031 --> 00:05:07,551 it. So we can say that we're seeing adoption 81 00:05:08,305 --> 00:05:11,563 discussions happening within government and 82 00:05:12,523 --> 00:05:16,031 Public sector. But then when we talk about the other industries 83 00:05:16,047 --> 00:05:19,861 that we've listed as these are going to be the next in line that really 84 00:05:19,925 --> 00:05:23,590 have to have it, which is like healthcare and finance, we are 85 00:05:23,719 --> 00:05:27,406 seeing the discussions are ramping up, but we're not seeing adoption 86 00:05:28,716 --> 00:05:32,513 come as fast as, let's say, we're seeing it being discussed within 87 00:05:33,197 --> 00:05:34,825 public sector or government. Right. 88 00:05:38,032 --> 00:05:41,805 That makes a lot of sense because you mentioned the 2 89 00:05:41,805 --> 00:05:45,321 use cases where I would think privacy and things like that would have it. We 90 00:05:45,321 --> 00:05:48,757 had another guest, and hopefully by the time this is released, that show has been— 91 00:05:48,789 --> 00:05:52,240 we do too many podcasts, you kind of forget what's been released. But he 92 00:05:52,240 --> 00:05:55,853 basically said bad actors could destroy the global 93 00:05:55,853 --> 00:05:59,096 financial system, not just the US, and time that with a particular 94 00:05:59,385 --> 00:06:02,933 aggressive military kinetic type option and could 95 00:06:02,997 --> 00:06:06,673 literally destroy wealth overnight by breaking the 96 00:06:06,673 --> 00:06:10,446 financial system and the encryption and the keys, not just to our personal accounts, 97 00:06:10,462 --> 00:06:13,978 but to the higher-level accounts that kind of link all these things together. And 98 00:06:14,010 --> 00:06:17,670 that, that would cause more chaos than any bomb alone could 99 00:06:17,670 --> 00:06:21,379 do, right? And it would basically— it'd be a big problem and a 100 00:06:21,379 --> 00:06:24,429 big mess. So yeah, I would imagine that 101 00:06:25,457 --> 00:06:28,972 financial institutions, healthcare institutions are— And we've 102 00:06:28,989 --> 00:06:32,713 seen kind of just with ransomware stuff that healthcare institutions are 103 00:06:32,761 --> 00:06:36,498 not They can't handle conventional threats. I would 104 00:06:36,498 --> 00:06:40,345 say the banks are probably better prepared for 105 00:06:40,393 --> 00:06:44,161 kind of conventional cyber threats, but hospitals, as we see, and 106 00:06:44,161 --> 00:06:47,799 increasingly also local governments are not. So I 107 00:06:47,847 --> 00:06:51,262 think it's a wake-up call because for years you've heard this kind of 108 00:06:51,454 --> 00:06:55,285 Chicken Little type people, not that they're wrong. I don't mean it that they're 109 00:06:55,285 --> 00:06:59,037 wrong. They're like, there's gonna be a cyber digital Pearl Harbor type thing. But 110 00:06:59,165 --> 00:07:02,194 if you cry wolf, even if the wolf is really there for so long and 111 00:07:02,194 --> 00:07:05,849 everybody ignores you and every normal day is, it seems like you're crying wolf. 112 00:07:06,218 --> 00:07:09,808 eventually you're gonna be right and everyone's gonna be like, how come no one 113 00:07:09,808 --> 00:07:13,479 predicted this? I always like to say, during the pandemic, there was a 114 00:07:13,479 --> 00:07:17,069 historian talking about the 2 most— I think it was right after the invasion of 115 00:07:17,101 --> 00:07:20,692 Ukraine, 'cause we had the pandemic and then we had that. And he goes, the 116 00:07:20,692 --> 00:07:24,458 2 most common things in human history are pandem— are plagues and wars. And 117 00:07:24,538 --> 00:07:28,049 yet every time they happen, we're all shocked that it happened. 118 00:07:28,305 --> 00:07:32,072 So what's interesting about post-quantum 119 00:07:32,072 --> 00:07:35,858 cryptography, or at least getting ready for it, is you, You 120 00:07:35,858 --> 00:07:39,681 can see the iceberg. You can see it coming. We're on the ship and 121 00:07:40,196 --> 00:07:43,055 we have to make sure that we're ready for it. 122 00:07:44,035 --> 00:07:46,798 Is this a time to panic or is this a time to prepare? I think 123 00:07:46,798 --> 00:07:49,561 that's really kind of like the 2 questions you get out of this. And then 124 00:07:49,690 --> 00:07:52,646 I'll stop talking and let Candace ask questions. I 125 00:07:52,646 --> 00:07:56,067 wouldn't say it's time to 126 00:07:56,083 --> 00:07:59,634 panic. I would definitely say it's time to prepare, but, 127 00:07:59,794 --> 00:08:03,216 but from where I sit, And from conversations that I have, 128 00:08:04,195 --> 00:08:07,841 most people are unaware of what that looks like. Right. How do 129 00:08:07,905 --> 00:08:11,583 we actually prepare? As I mentioned, government are releasing mandates, 130 00:08:11,663 --> 00:08:15,036 they're releasing legislation, we're getting those things, but then the 131 00:08:15,036 --> 00:08:18,794 adoption I feel is not moving 132 00:08:18,987 --> 00:08:22,633 as quickly as the legislation that's being released is because there's 133 00:08:22,633 --> 00:08:26,375 just a knowledge gap or skill gap in saying, okay, we 134 00:08:26,375 --> 00:08:30,147 know we need it, but what's the first step? And so I 135 00:08:30,212 --> 00:08:33,839 think preparation definitely needed, but I would say 136 00:08:34,384 --> 00:08:37,787 one precursor to preparation is understanding or 137 00:08:37,851 --> 00:08:41,590 education. I think companies need to focus on 138 00:08:42,024 --> 00:08:45,442 being educated on what this is 139 00:08:46,180 --> 00:08:50,016 and how they can proceed to prepare, because preparation for one 140 00:08:50,706 --> 00:08:54,461 industry versus another, or even one organization 141 00:08:54,461 --> 00:08:58,028 to another, is going to look different. So I think education It's 142 00:08:58,077 --> 00:09:01,035 time for us to get educated, for us to build awareness. 143 00:09:04,119 --> 00:09:07,633 Absolutely. So as quantum is increasingly being treated 144 00:09:07,810 --> 00:09:11,578 as a national security technology, how does 145 00:09:11,610 --> 00:09:14,837 that change the way countries collaborate on 146 00:09:14,935 --> 00:09:16,224 research and innovation? 147 00:09:18,377 --> 00:09:20,544 That's a great question. 148 00:09:22,453 --> 00:09:25,997 So definitely we are seeing it happening. So 149 00:09:26,093 --> 00:09:29,786 we know that out of the White House we have Project Genesis, for 150 00:09:29,786 --> 00:09:32,484 instance, in which we do have 151 00:09:33,062 --> 00:09:36,899 US-based organizations and foreign entities both coming 152 00:09:36,932 --> 00:09:40,512 to look like, and we are seeing alliances 153 00:09:40,608 --> 00:09:44,269 being formed. But then those alliances are 154 00:09:44,302 --> 00:09:47,689 lost in different groups. So maybe you could say there's a lot of 155 00:09:47,689 --> 00:09:51,382 collaboration between, let's say, the US and the UK, But 156 00:09:51,382 --> 00:09:55,116 then when we go to the eastern side of the world, we're not seeing 157 00:09:55,484 --> 00:09:59,314 so much collaboration between, let's say, Russia, China, 158 00:09:59,394 --> 00:10:02,935 and then America. So I think that if we 159 00:10:03,384 --> 00:10:07,166 really look at it, we say, okay, we have to first identify 160 00:10:07,166 --> 00:10:10,931 who our allies are. And then once we identify that, then 161 00:10:10,931 --> 00:10:14,745 we can identify what does the communication path look like? What does 162 00:10:15,018 --> 00:10:18,693 responsibilities, roles, and delegation look like? But I think it 163 00:10:18,709 --> 00:10:21,608 first starts with identifying who our allies are. 164 00:10:23,702 --> 00:10:26,764 Yeah, it's not a surprise to see that it's going to fall along existing kind 165 00:10:26,764 --> 00:10:30,290 of military-political alliance strategies. 166 00:10:30,662 --> 00:10:34,251 It's not— it's interesting. I never thought when I started 167 00:10:36,762 --> 00:10:39,547 playing with computers when I was a kid or 168 00:10:40,786 --> 00:10:43,909 studying it in school or starting this career that this would become a geopolitical 169 00:10:44,643 --> 00:10:48,435 Hot button point, right? Whether it's— this is very much a— it's a space— 170 00:10:48,435 --> 00:10:51,408 it's like the space race. I'm too young to remember the space race, but you 171 00:10:51,552 --> 00:10:55,268 see all these like shows and like history things about it. It almost feels like 172 00:10:55,268 --> 00:10:57,880 it's that same type of vibe. It is. 173 00:11:00,329 --> 00:11:03,751 Did you ever think you would like work in a tense geopolitical 174 00:11:04,042 --> 00:11:07,775 or adjacent to a tense geopolitical space? Because I certainly didn't. 175 00:11:10,092 --> 00:11:12,819 How far back do you want to go? If you say a year ago, say 176 00:11:12,819 --> 00:11:16,652 10 years ago, probably not. Yeah. So I mean like 10 years ago, I mean 177 00:11:16,652 --> 00:11:20,277 like, I don't think, 'cause I remember, and you know, living in DC, you get 178 00:11:20,277 --> 00:11:24,014 a different perspective on the world than you do kind of anywhere else. 179 00:11:24,127 --> 00:11:26,565 Candace and I both grew up in New York and it's, you get a very 180 00:11:26,853 --> 00:11:30,414 different perspective there. But since the thinking here is different, so like when I 181 00:11:30,622 --> 00:11:34,247 talk to colleagues and like, even like, I just would've 182 00:11:34,311 --> 00:11:37,680 been, see a company like Nvidia treated like a defense 183 00:11:37,680 --> 00:11:41,255 contractor. Right. And they're going to have export controls. And 184 00:11:42,329 --> 00:11:45,775 my former colleagues who are based in Raleigh and in Boston, and if you're clever 185 00:11:45,775 --> 00:11:48,420 and you can figure out how to use LinkedIn, you'll know exactly what company I'm 186 00:11:48,420 --> 00:11:51,883 talking about, thought I was insane. They thought I was like 187 00:11:52,107 --> 00:11:55,714 tinfoil hat guy. Yet what have we seen over the 188 00:11:55,714 --> 00:11:59,529 last, what, 24 months? Yeah. Something pretty 189 00:11:59,562 --> 00:12:03,393 close to that. I would say. So like 190 00:12:03,393 --> 00:12:06,666 today's paranoia, is tomorrow's policy, 191 00:12:06,698 --> 00:12:10,198 right? It is. And DC, I think, is— 192 00:12:10,455 --> 00:12:14,131 people always bag on DC and politicians as being behind the curve, and there is 193 00:12:14,244 --> 00:12:17,920 truth to that. When it comes to stuff like this, they're actually ahead of the 194 00:12:17,920 --> 00:12:21,517 curve, right? The whole notion of Jensen Huang traveling 195 00:12:21,549 --> 00:12:24,904 places and it becomes a political event, right? Like, he was in 196 00:12:24,904 --> 00:12:28,388 Taiwan and like he had to be very careful about what he said, right? Like, 197 00:12:29,239 --> 00:12:31,792 I don't know, it's just interesting to see that. And if you had said that 198 00:12:31,888 --> 00:12:35,386 3, 4 years ago, you would've been like, really? Is, is anyone gonna care 199 00:12:35,995 --> 00:12:39,137 that he visited any point on earth? I know there was some 200 00:12:39,185 --> 00:12:42,471 controversy. I guess when you become like the most valued company in the world, like 201 00:12:42,471 --> 00:12:45,742 the, the lens on you becomes a little bit brighter, right? It does. 202 00:12:46,158 --> 00:12:49,733 Yeah, absolutely. What did he do? He paid people off to skip a 203 00:12:49,845 --> 00:12:53,612 line in some streetside, like, it's not, it wasn't like a taco truck. It 204 00:12:53,612 --> 00:12:57,444 was in Korea. So whatever the equivalent of a taco truck in 205 00:12:57,636 --> 00:13:01,355 Korea would be, like he cut the line. He goes, would you mind if I 206 00:13:01,403 --> 00:13:04,865 each give you pay for your food? I can go first. And everybody's like, yeah, 207 00:13:04,930 --> 00:13:08,411 free food. But then all these people were like, oh no, that's terrible that he 208 00:13:08,411 --> 00:13:12,118 did that. Absolutely. Yeah. Perspective, right? It's very important. 209 00:13:13,040 --> 00:13:16,744 Yeah. To your point, the paranoia, it definitely 210 00:13:16,856 --> 00:13:20,239 drives action and it's going to determine our actions for sure. 211 00:13:22,369 --> 00:13:26,149 So let me ask you, what do you think the quantum industry is 212 00:13:26,246 --> 00:13:30,010 getting right today and what is it still 213 00:13:30,203 --> 00:13:32,029 talking about in the wrong way? 214 00:13:35,509 --> 00:13:35,915 Ooh. So, 215 00:13:39,938 --> 00:13:43,317 and to my quantum experts, I want to talk to just regular people for now. 216 00:13:43,413 --> 00:13:46,842 So if you're listening, you're a quantum expert, I want to talk to normal people 217 00:13:46,842 --> 00:13:49,917 for now. So it's going to be simple. So 218 00:13:50,400 --> 00:13:54,104 yeah, just keep that in mind. So I think what we're 219 00:13:54,104 --> 00:13:57,580 getting right is I think most being 220 00:13:57,596 --> 00:14:01,409 developed I think it's great. I think we're 221 00:14:01,409 --> 00:14:05,008 starting— I think we have great research on error correction. 222 00:14:05,361 --> 00:14:08,985 I think we have great research on— we have great research on— it's 223 00:14:08,985 --> 00:14:10,700 great. I think 224 00:14:13,176 --> 00:14:16,800 one thing that we do need to improve is 225 00:14:16,800 --> 00:14:20,538 collaboration. I do see that there's quite a few silos across 226 00:14:20,538 --> 00:14:23,857 the industry. Now, I can talk to physicists, 227 00:14:24,416 --> 00:14:28,155 and they have their own focus for research. But then I 228 00:14:28,203 --> 00:14:31,782 can talk to quantum engineers. Like I know software 229 00:14:31,782 --> 00:14:34,863 engineers that work for quantum hardware companies like PsiQuantum, 230 00:14:35,489 --> 00:14:38,923 and their research is focused on a totally 231 00:14:39,035 --> 00:14:42,871 different paradigm of quantum. And they're more focused on, can we 232 00:14:42,871 --> 00:14:46,305 develop quantum software that improves 233 00:14:46,417 --> 00:14:50,253 hardware characterization? But then you, it doesn't play a role here. We just need 234 00:14:50,253 --> 00:14:53,863 to focus on making the hardware better, use better materials, use better 235 00:14:53,863 --> 00:14:57,617 equipment. So I think even though we are seeing 236 00:14:57,970 --> 00:15:01,756 collaboration, as we mentioned, at the geopolitical level where 237 00:15:02,221 --> 00:15:06,039 policymakers and congressmen are coming together talking, I 238 00:15:06,056 --> 00:15:09,296 would say down on the ground, when you talk about scientists and 239 00:15:09,296 --> 00:15:12,039 engineers, we need to see more 240 00:15:12,248 --> 00:15:15,616 collaboration between scientists and engineers for sure. 241 00:15:16,739 --> 00:15:20,557 The other thing that I would say, something that I feel we're getting 242 00:15:20,670 --> 00:15:24,165 wrong in addition to working in silos, is 243 00:15:24,919 --> 00:15:28,222 the approach. Now, if I go back to 244 00:15:28,847 --> 00:15:32,551 when cloud first became a thing and AWS was on the scene and they were 245 00:15:32,696 --> 00:15:35,951 building cloud, there was just a lot of 246 00:15:35,983 --> 00:15:39,831 communication with business leaders and then architecture. Mm-hmm. 247 00:15:40,328 --> 00:15:43,791 So they had architects and they had business 248 00:15:43,904 --> 00:15:47,752 owners really driving, this is how we should design 249 00:15:48,088 --> 00:15:51,440 the AWS console. So there was a lot of business and 250 00:15:51,440 --> 00:15:55,192 architecture in the beginning, but I feel with quantum, we're not necessarily 251 00:15:55,385 --> 00:15:59,073 leading with business and architecture. We're 252 00:15:59,073 --> 00:16:02,890 just saying, I just wanna build something, but overall strategy is 253 00:16:02,922 --> 00:16:06,690 not present from what I can see as an industry. Now you do 254 00:16:06,690 --> 00:16:10,443 have Quantinum and IonQ, which I believe do a pretty good job, but the 255 00:16:10,523 --> 00:16:14,196 industry as a whole, there's not really a unified strategy 256 00:16:14,532 --> 00:16:18,246 that we're marching towards. So I would like to see 257 00:16:18,536 --> 00:16:22,164 quantum architecture take some legs and 258 00:16:22,164 --> 00:16:25,970 start walking, and I would like to see maybe quantum business or 259 00:16:26,100 --> 00:16:29,751 business leaders now weighing in on these are the use cases that are 260 00:16:29,751 --> 00:16:33,369 important to now drive engineering and scientific research. 261 00:16:36,688 --> 00:16:40,395 So if there are several competing quantum architectures going on right 262 00:16:40,509 --> 00:16:44,340 now, What do you look at when deciding which 263 00:16:44,340 --> 00:16:47,512 approach has the greatest long-term potential? 264 00:16:49,412 --> 00:16:52,856 I say it's based on the use case. Because 265 00:16:54,037 --> 00:16:57,628 if I'm an organization, if I'm a healthcare organization, and I say I want to 266 00:16:57,661 --> 00:17:01,050 use quantum for drug discovery, I want to figure 267 00:17:01,131 --> 00:17:04,954 out what is going to be the best sequence 268 00:17:05,179 --> 00:17:09,019 of treatments to help this person with Parkinson's disease. I 269 00:17:09,164 --> 00:17:12,725 want to know what that is. And so now your quantum approach is, I want 270 00:17:12,725 --> 00:17:15,128 to use quantum to improve GPS navigation. 271 00:17:17,270 --> 00:17:20,989 It's a totally different paradigm to say I want to use 272 00:17:21,102 --> 00:17:23,340 one for medical treatments, I want to use— 273 00:17:24,966 --> 00:17:27,453 so I think the use case is going to be the most critical. 274 00:17:32,129 --> 00:17:35,066 Okay, interesting. So where do you think 275 00:17:36,126 --> 00:17:39,155 What about regular non-regulated industries, right? Which is 276 00:17:39,829 --> 00:17:43,323 fancy DC speak for everybody but 277 00:17:43,339 --> 00:17:46,625 healthcare, government, and finance. What I own, 278 00:17:46,802 --> 00:17:50,312 let's pick on Tim Hortons, right? 'Cause they own, the only Tim Hortons in 279 00:17:50,312 --> 00:17:53,615 Maryland is actually near my house. Who I have to worry about, 280 00:17:54,320 --> 00:17:57,991 or Starbucks, whatever, don't really deal with a lot of 281 00:17:59,129 --> 00:18:02,768 regulations to my stance on post-quantum cryptography. 'Cause I think in this industry, 282 00:18:02,992 --> 00:18:06,583 you live in DC, when you talk quantum cryptography, I would imagine with most of 283 00:18:06,583 --> 00:18:08,893 your customers, they get it. To people 284 00:18:10,114 --> 00:18:13,649 outside of that DC bubble, they probably think you've 285 00:18:13,665 --> 00:18:17,119 lost your mind. They probably see you as a tinfoil hat. 286 00:18:17,585 --> 00:18:20,082 You know that meme with the guy from It's Sunny in Philadelphia with all the 287 00:18:20,082 --> 00:18:22,806 stuff together? What do you say to people like that? 288 00:18:25,882 --> 00:18:29,465 So PQC, we really only talk about it from one lens, 289 00:18:30,044 --> 00:18:33,354 and that is some form of protection. Harvest now, 290 00:18:33,434 --> 00:18:36,422 decrypt later. A hacker is now gonna come in. 291 00:18:36,972 --> 00:18:40,707 steal my stuff. I need post-quantum cryptography from a protection lens. 292 00:18:41,300 --> 00:18:44,908 But I also say there's another 293 00:18:44,908 --> 00:18:48,242 lens in which PQC is useful, and that is 294 00:18:48,595 --> 00:18:52,362 operationalization of processes. And I'll explain what I 295 00:18:52,362 --> 00:18:55,745 mean. Now, if you have a banking app 296 00:18:56,690 --> 00:19:00,249 that you log into and you want to transfer money, now 297 00:19:00,762 --> 00:19:04,224 most people will say, I want it to be— I want to maybe use my 298 00:19:04,224 --> 00:19:07,957 thumbprint, get into my app, process something, 299 00:19:08,214 --> 00:19:11,899 have it go through immediately. The money transfers and 300 00:19:11,979 --> 00:19:15,552 everything is a matter of seconds. So we call that performance, operational 301 00:19:15,600 --> 00:19:18,981 performance. It's easy, it's fast, it's accurate. 302 00:19:19,846 --> 00:19:23,692 Go back to, let's say, 10 years ago, we didn't 303 00:19:23,692 --> 00:19:27,521 really have really good mobile banking applications. You pretty 304 00:19:27,521 --> 00:19:30,469 much have to go into the branch to do everything. And that is just because 305 00:19:30,966 --> 00:19:34,607 If I want to transfer money, I need this system to talk to this 306 00:19:34,607 --> 00:19:38,441 system. Now, this system talking to this system, you want to make 307 00:19:38,441 --> 00:19:41,810 sure it's secure, but you also want to make sure that it's fast and it's 308 00:19:41,810 --> 00:19:45,259 reliable, which is a performance metric. So I 309 00:19:45,484 --> 00:19:48,805 think if you're outside regulated industries, I own a 310 00:19:48,869 --> 00:19:52,446 restaurant, then now I'm saying I want to use 311 00:19:52,799 --> 00:19:56,216 post-quantum cryptography to make people ordering 312 00:19:56,232 --> 00:19:59,457 food easier. Because I have PQC 313 00:20:00,484 --> 00:20:04,256 embedded in my systems, someone can set up a profile for my 314 00:20:04,256 --> 00:20:07,770 restaurant. They can save previous orders that they've done. 315 00:20:08,444 --> 00:20:12,200 They can say, I like this type of food, and they can store 316 00:20:12,200 --> 00:20:15,907 their card information on the app. And it's secure, but it 317 00:20:15,907 --> 00:20:19,085 also helps with user experience 318 00:20:19,518 --> 00:20:22,166 because now I can just go in and say, hey, this is the order I 319 00:20:22,166 --> 00:20:25,839 got last time. I want to click on that. I know it's secure because PQC 320 00:20:25,839 --> 00:20:28,627 is there. I know it's going to be fast because I can easily 321 00:20:28,659 --> 00:20:32,249 just iterate through the different systems. If I'm the 322 00:20:32,249 --> 00:20:35,743 restaurant owner, I also want that because if I now 323 00:20:35,872 --> 00:20:38,917 need to track my sales, if I now need to update my 324 00:20:38,917 --> 00:20:42,571 inventory, update my sales, then I can 325 00:20:42,603 --> 00:20:46,225 do all that and I can have the integrations between my systems. 326 00:20:46,594 --> 00:20:49,703 And I know that it's secure and I know it's going to be fast because 327 00:20:49,735 --> 00:20:53,485 now I don't have to manually perform that step to update my general ledger. 328 00:20:53,886 --> 00:20:57,351 I can just have it connected directly to my accounting software and know that it's 329 00:20:57,351 --> 00:21:00,640 secure because PTC is present. So I think if you're 330 00:21:00,865 --> 00:21:04,619 outside of regulated industries, you're not worried that I need to meet this 331 00:21:04,715 --> 00:21:08,550 compliance standard. You're thinking, I want more automation and 332 00:21:08,566 --> 00:21:12,143 I want more integration so that users can get in and out of my apps 333 00:21:12,192 --> 00:21:15,721 really fast and it's going to be secure. And I want my 334 00:21:15,882 --> 00:21:19,732 backends to also be integrated and automated really well. but I 335 00:21:19,732 --> 00:21:23,257 also wanna make sure it's secure as well. So I can have post-quantum cryptography in 336 00:21:23,257 --> 00:21:26,669 between connecting everything for automation, but securing 337 00:21:26,861 --> 00:21:30,514 everything for peace of mind. So that's my stance on 338 00:21:30,931 --> 00:21:34,360 outside regulatory agencies or regulatory 339 00:21:34,392 --> 00:21:37,981 industry clients. So this is a 340 00:21:37,981 --> 00:21:41,826 question that I've often thought of, and I've kind of asked it to other 341 00:21:41,842 --> 00:21:45,046 experts. I never got really a great answer. So 342 00:21:45,591 --> 00:21:47,562 I'll pose it to you. No pressure, right? 343 00:21:49,714 --> 00:21:53,530 Given that there's no— how do we know these algorithms are 344 00:21:53,530 --> 00:21:57,216 secure against quantum computers since the quantum 345 00:21:57,216 --> 00:22:00,669 computer, they can— so 2 346 00:22:00,702 --> 00:22:04,139 questions. How do we know they're secure? And 2, is there any 347 00:22:04,172 --> 00:22:07,754 guarantee that with more qubits and more processing power, they won't be secure? 348 00:22:07,883 --> 00:22:11,529 So we'll have to continually upgrade our encryption algorithms, right? 349 00:22:11,997 --> 00:22:15,788 Because I know when they had the— NIST had the contest 350 00:22:15,788 --> 00:22:19,526 and there were a couple of candidates. One or more of the candidates failed, 351 00:22:19,558 --> 00:22:23,314 right? Because cryptography's a tricky business. And if you ever 352 00:22:23,330 --> 00:22:26,892 want to, I recommend the Bruce Schneier book. It used to be red and now 353 00:22:26,956 --> 00:22:30,695 it's blue, or I don't know, but it came out. What? Because cryptography is an 354 00:22:30,936 --> 00:22:34,579 ancient science. People don't realize that. And it really got heavy into 355 00:22:34,627 --> 00:22:38,350 math sometime in the 1800s and 1900s, obviously World War II. But 356 00:22:39,232 --> 00:22:42,313 so if you really want to go deep into that, read that book. There is 357 00:22:42,346 --> 00:22:46,036 no substitute. Will you get a migraine? Most probably, because it's not for the 358 00:22:46,036 --> 00:22:49,387 timid. But You know what I'm talking about, right? So, but 359 00:22:49,807 --> 00:22:53,123 how do we know it's secure? Like, how do we know? And is this gonna 360 00:22:53,123 --> 00:22:56,099 be a, a new kind of cat and mouse game? We built it for it 361 00:22:56,131 --> 00:22:59,954 to survive this kind of attack, which is really the history of cryptography if you 362 00:22:59,954 --> 00:23:02,284 think about it, going back to the Romans. 363 00:23:05,069 --> 00:23:08,743 So I feel like there's— I know it's a whopper of a, 364 00:23:08,824 --> 00:23:12,514 it's a whopper of a question. I know. And I apologize. So 365 00:23:13,111 --> 00:23:16,208 I, Quantum is so 366 00:23:17,235 --> 00:23:20,875 new in terms of talent pool and skillset. 367 00:23:21,260 --> 00:23:24,436 So what I mean by, we don't have a lot of 368 00:23:24,612 --> 00:23:28,157 experts in this space. We have cryptographic 369 00:23:28,157 --> 00:23:31,782 experts for sure that really understand cryptography. But when we talk about 370 00:23:31,926 --> 00:23:35,759 these new quantum-safe algorithms, we 371 00:23:35,759 --> 00:23:39,384 know there's experts at NIST. We know there's a handful of experts at your 372 00:23:39,384 --> 00:23:43,218 different hardware vendors. So we know they're there, right? When we talk about— there's 373 00:23:43,218 --> 00:23:46,574 probably experts at places we don't know about, but yes. But 374 00:23:46,927 --> 00:23:50,669 largely, hold on, say application. So the reason why I'm saying that 375 00:23:50,781 --> 00:23:54,618 is because I see people kind of diverge in 2 different directions so 376 00:23:54,634 --> 00:23:58,151 far with PQC. Either they're going to be focused on the compliance 377 00:23:58,151 --> 00:24:01,603 aspect of PQC. Can we do assessments or 378 00:24:01,603 --> 00:24:05,296 analysis that says this is all the things you need to have to 379 00:24:05,328 --> 00:24:09,069 implement PQC? So compliance standards. But then you have people like 380 00:24:09,069 --> 00:24:12,569 me that are more technical in nature. I'm more curious about how 381 00:24:12,762 --> 00:24:16,245 can we address this technically? What type of microservices can we 382 00:24:16,646 --> 00:24:20,306 build? So the reason I say that is my response is going to be 383 00:24:20,306 --> 00:24:24,014 more technically driven. And so for me to 384 00:24:24,014 --> 00:24:26,952 say, how do we know it's secure? If you ask someone that's focused on compliance, 385 00:24:26,984 --> 00:24:30,740 they will say, oh, because all the controls are implemented. So if 386 00:24:30,740 --> 00:24:34,207 all the controls are implemented, then we can say it's 387 00:24:34,207 --> 00:24:38,044 secure. Right. And that's a compliance answer. 388 00:24:38,076 --> 00:24:40,721 But me, if I'm going to give a technical answer, then I'm gonna say that 389 00:24:41,490 --> 00:24:44,807 the only way we can really know it's secure is through 390 00:24:45,336 --> 00:24:49,102 testing. We have to test it, right? And 391 00:24:49,102 --> 00:24:52,771 so me personally, I have developed 392 00:24:52,819 --> 00:24:56,377 something that we call a purple teaming exercise. 393 00:24:57,771 --> 00:25:00,078 So I deployed like a quantum 394 00:25:01,216 --> 00:25:04,613 environment at my house. So I've built it. So I have, I have my, 395 00:25:04,982 --> 00:25:08,741 Quantum libraries there, I have Qiskit deployed, I 396 00:25:08,741 --> 00:25:11,023 build like a quantum server, 397 00:25:11,874 --> 00:25:15,650 hypervisor. You've taken home labbing to the next 398 00:25:15,650 --> 00:25:19,248 level. That's awesome. I'll have you back and geek out on that one, man. Because 399 00:25:19,280 --> 00:25:23,056 the question you just asked, if I'm leading PQC and I'm 400 00:25:23,056 --> 00:25:26,687 considered a PQC expert, then how do I know this 401 00:25:26,847 --> 00:25:30,655 algorithm and this configuration will stand up to a quantum attack? The only 402 00:25:30,687 --> 00:25:33,786 way I can know that is if I launch a quantum 403 00:25:33,818 --> 00:25:37,590 attack and I see, will it hold? And we just call that purple teaming. Right. 404 00:25:37,847 --> 00:25:41,683 And so in my purple teaming, this is what I've identified. Now, 405 00:25:42,470 --> 00:25:46,306 again, based on the maturity of the industry, we really focus 406 00:25:46,402 --> 00:25:50,158 heavily on cryptographic identities. Do 407 00:25:50,254 --> 00:25:53,449 I have RSA? I do. Do I have 408 00:25:53,721 --> 00:25:56,530 EDS? What do I have? Do I have PKI? I've 409 00:25:56,530 --> 00:25:59,933 identified a cryptographic identity. So, but we're not 410 00:25:59,981 --> 00:26:03,738 focusing on the configuration of cryptography. And 411 00:26:03,738 --> 00:26:07,208 I think that really, if we want to know it's secure, we have to get 412 00:26:07,208 --> 00:26:10,807 into the detail. I have RSA. Great. I know I have 413 00:26:10,807 --> 00:26:14,599 RSA 2048. It has 248 bits, 2048 bits. 414 00:26:14,920 --> 00:26:18,455 I know that. But now, how is it actually configured? What is it 415 00:26:18,455 --> 00:26:21,845 actually doing in this system? Am I using it 416 00:26:21,990 --> 00:26:25,653 just to protect my web pages so that when people put their credit 417 00:26:25,669 --> 00:26:29,252 card information in, it's not exposed? How am I using RSA? 418 00:26:29,927 --> 00:26:33,636 And I think that when we can go down into the weeds and 419 00:26:33,636 --> 00:26:36,413 say, I'm using RSA really just to 420 00:26:36,991 --> 00:26:40,748 provide some type of encapsulation of people's socials 421 00:26:40,780 --> 00:26:44,521 when they entered on my website, then now we have to evaluate 422 00:26:44,939 --> 00:26:48,728 things like, okay, if someone does a 423 00:26:48,872 --> 00:26:52,356 bypass attack, will it work? And that's purple teaming. 424 00:26:52,758 --> 00:26:56,337 So to your answer, how do we know it's secure? We have to 425 00:26:56,353 --> 00:26:58,437 understand how is the 426 00:26:58,437 --> 00:27:02,044 cryptographic identity configured? 427 00:27:03,086 --> 00:27:06,821 What is its job? Why is it there? What is it doing? 428 00:27:07,398 --> 00:27:09,898 And then once we know that this is what it's supposed to do, this is 429 00:27:09,962 --> 00:27:13,810 how it's configured, then we can use purple teaming to test it. We 430 00:27:13,810 --> 00:27:17,529 can use a quantum system like Qiskit, like IBM. 431 00:27:17,753 --> 00:27:21,584 GCP also has it. We can use that to now attack it once 432 00:27:21,584 --> 00:27:25,063 we deploy it. If we deploy that key, we can attack it 433 00:27:25,752 --> 00:27:29,503 with an actual, some type of quantum simulator, and we see how well it does. 434 00:27:30,176 --> 00:27:33,863 But without that, without understanding the configuration of it, without 435 00:27:33,863 --> 00:27:37,262 understanding what it's being used for, and without testing it, 436 00:27:37,951 --> 00:27:41,798 we have no real way of knowing if it's secure or not. Just 437 00:27:41,798 --> 00:27:45,036 have to take it on faith, right? If you hit all the compliance things. So 438 00:27:45,085 --> 00:27:48,836 here's a real-world example for folks that are not used to the compliance 439 00:27:48,836 --> 00:27:51,865 world in the DC kind of bubble. I have 5 dogs 440 00:27:52,122 --> 00:27:55,828 and 2 of them are escape artists. And there's been 441 00:27:55,828 --> 00:27:59,087 some recent issues I'm not gonna get into for what could be a potentially 442 00:27:59,601 --> 00:28:03,359 legal problem. 5 dogs. We have a wireless fence in the 443 00:28:03,359 --> 00:28:07,004 ground. We have wireless fence that is truly wireless. It broadcasts a radio 444 00:28:07,004 --> 00:28:10,713 signal. We also have physical barriers and we have a 445 00:28:10,858 --> 00:28:14,631 zipline tie-out for some of them. But earlier, so 446 00:28:14,696 --> 00:28:18,036 we, this is all because they've been escape artists, right? Mm-hmm. So all of these 447 00:28:18,052 --> 00:28:21,440 things I've been doing, they are, I, I like the way you're putting this, right? 448 00:28:21,745 --> 00:28:25,578 They are red teaming my ability to keep them fenced in. Right? 449 00:28:26,127 --> 00:28:29,415 And 2 nights ago, the 3-year-old had 450 00:28:29,496 --> 00:28:33,284 basically opened the door for them and had— 451 00:28:33,284 --> 00:28:36,705 so then that was something I had not anticipated. So now there's zip ties on 452 00:28:36,721 --> 00:28:40,037 that door. So, so basically you're right. The only way to know something is secure, 453 00:28:40,585 --> 00:28:44,256 if somebody from, say, animal control came by and 454 00:28:44,288 --> 00:28:47,285 said, how are you securing them? Mm-hmm. And let's just say this might have already 455 00:28:47,285 --> 00:28:50,984 happened. I showed them every possible barrier that I have 456 00:28:51,000 --> 00:28:54,836 created, right? And But I can't account for 457 00:28:54,868 --> 00:28:58,627 the 3-year-old just running out and opening the door, right? Now I 458 00:28:58,643 --> 00:29:02,080 have, but, and that, so that's kind of like the difference between compliance and virtual, 459 00:29:02,160 --> 00:29:05,710 like what you're saying, you actually know or not, right? Red teaming against it and 460 00:29:05,710 --> 00:29:09,019 things like that. I had, my assumption was that the dogs would be 461 00:29:09,870 --> 00:29:12,536 on their own to figure a way out, but little did I know there'd be 462 00:29:12,536 --> 00:29:15,893 an outside actor, an outside threat compounding that too. 463 00:29:16,375 --> 00:29:19,763 So now I have to think about it very differently. So I think it's very 464 00:29:19,763 --> 00:29:23,469 much a similar type of problem, right? Absolutely. Compliance, I can 465 00:29:23,517 --> 00:29:26,372 show all day. I have this control, I have this control. 466 00:29:26,981 --> 00:29:30,574 But until I have my 3-year-old do some red teaming on me, I really don't 467 00:29:30,590 --> 00:29:34,103 know. Absolutely. Absolutely. And I'll stop 468 00:29:34,103 --> 00:29:37,616 talking and let Candace talk. I'm just wondering if 469 00:29:37,648 --> 00:29:41,225 PQC primarily is a tech— a technology problem, 470 00:29:41,963 --> 00:29:45,476 or is the bigger challenge identifying where vulnerable 471 00:29:45,508 --> 00:29:46,967 cryptography exists 472 00:29:51,312 --> 00:29:54,506 Ooh, they're both really big problems. 473 00:29:55,732 --> 00:29:59,362 Okay. My experience tells me, so I think 474 00:29:59,362 --> 00:30:02,835 everyone is saying, I see a lot on LinkedIn. I see a lot of posts 475 00:30:02,835 --> 00:30:06,408 on LinkedIn all the time, and people inbox me quite a bit. And so 476 00:30:06,877 --> 00:30:10,602 I see a lot of talk of which algorithm should we use. We don't know 477 00:30:10,683 --> 00:30:14,201 that yet. Most organizations don't know that yet. And then they're 478 00:30:14,201 --> 00:30:17,685 saying, oh, can we do an assessment and 479 00:30:17,861 --> 00:30:21,473 find where our cryptographic identities are? I feel 480 00:30:21,553 --> 00:30:25,085 like organizations aren't ready for that either. I think 481 00:30:25,085 --> 00:30:28,681 that kind of step one for us is, 482 00:30:29,275 --> 00:30:32,903 hey, where do I even want to look for cryptographic 483 00:30:32,903 --> 00:30:36,114 identities? Should I look in my 484 00:30:36,114 --> 00:30:39,630 mainframe? Are we thinking to look there? Are we thinking to 485 00:30:39,630 --> 00:30:43,280 look in James's laptop where he's been 486 00:30:43,280 --> 00:30:47,060 there? We're probably not thinking to look there because he has 487 00:30:47,060 --> 00:30:50,740 so much code on his laptop. He has so many cryptographic identities on 488 00:30:50,740 --> 00:30:54,580 his laptop that's not in a production server. Are we thinking to look in 489 00:30:54,580 --> 00:30:58,120 a sandbox environment that we got from Salesforce? Because Salesforce said 490 00:30:58,120 --> 00:31:01,440 they're going to give us a sandbox we can use for 3 months and learn 491 00:31:01,440 --> 00:31:05,240 how to deploy apps, but we have already started integrating with our system. 492 00:31:05,240 --> 00:31:08,284 So now there's cryptographic identities there. Are we thinking to look there. 493 00:31:09,166 --> 00:31:12,726 And so trying to identify the target systems, I 494 00:31:12,918 --> 00:31:16,606 think we need to do that. And even having a 495 00:31:16,606 --> 00:31:20,038 complete list of what are our targets as an 496 00:31:20,134 --> 00:31:23,838 industry, we haven't really— we don't really have a structure for that 497 00:31:23,902 --> 00:31:27,446 yet. So to your point, Candice, if I'm looking at, 498 00:31:28,488 --> 00:31:32,240 is it a technical problem? Is it a compliance problem? Should 499 00:31:32,240 --> 00:31:35,830 we be looking for our identities? I think From a 500 00:31:35,830 --> 00:31:38,609 technical problem perspective, we don't even know what we have. 501 00:31:39,420 --> 00:31:42,130 And then to say we don't know what we have, we don't even know where 502 00:31:42,130 --> 00:31:45,810 to look. So I think, where 503 00:31:45,810 --> 00:31:49,550 are we actually looking for quantum or 504 00:31:49,550 --> 00:31:52,830 for cryptographic identities? Can we first get a list of 505 00:31:52,830 --> 00:31:56,650 assets that we need to actually look at? And 506 00:31:56,650 --> 00:32:00,270 then how can we actually look in someone's laptop to look for cryptographic 507 00:32:00,270 --> 00:32:03,230 identities? How are we actually going to do that? Now it comes back to the 508 00:32:03,230 --> 00:32:06,953 technical aspect of it. But it's like we need a target. So then 509 00:32:07,113 --> 00:32:10,610 when we find RSA, what do we do about it? Like how now, 510 00:32:10,979 --> 00:32:14,076 where does this technology part come in? Do I now try to do a hybrid? 511 00:32:14,349 --> 00:32:18,183 Do I deploy a quantum-safe algorithm next to it? Do I rip it out and 512 00:32:18,183 --> 00:32:21,857 replace it? So now it, it's kind of a loop in my, in my opinion. 513 00:32:22,210 --> 00:32:25,868 You find a target, you find a cryptographic identity, you develop some type of 514 00:32:26,044 --> 00:32:29,574 technical remediation for that. But the, the 515 00:32:29,670 --> 00:32:32,927 first part is identifying our targets, identifying our 516 00:32:33,103 --> 00:32:36,671 assets. That we need to even look, we 517 00:32:36,671 --> 00:32:40,228 haven't done. We haven't done. So I would say— Yeah, 518 00:32:40,406 --> 00:32:43,462 I'm sorry, I didn't mean to cut you off, but I think you, you bring 519 00:32:43,462 --> 00:32:46,956 up something that is true in at least conventional cryptography, right? Which 520 00:32:47,068 --> 00:32:49,194 is, it's never— 521 00:32:50,788 --> 00:32:53,753 people always like to give you the big numbers of a brute force attack will 522 00:32:53,753 --> 00:32:57,598 take this many years. How often has a brute force attack actually been the thing 523 00:32:57,630 --> 00:33:00,894 that actually gets the thing? It's always the key management is always the issue. And 524 00:33:00,894 --> 00:33:03,569 you mentioned The developer laptops and things like that. That, 525 00:33:04,643 --> 00:33:08,348 and I'm by far from a cryptic, you know, cryptographic or cryptanalysis 526 00:33:08,348 --> 00:33:11,764 type expert, but I can appreciate the fact that 527 00:33:12,357 --> 00:33:16,094 it's that last mile is the issue, right? Mm-hmm. An 528 00:33:16,207 --> 00:33:19,607 unwatched key, an unlocked laptop, a left behind thumb 529 00:33:19,719 --> 00:33:23,440 drive will, has the potential to blow a 530 00:33:23,568 --> 00:33:27,305 massive hole through a multi-billion dollar defensive thing, 531 00:33:27,433 --> 00:33:30,945 right? It becomes like the Achilles heel of all of this, right? 532 00:33:31,507 --> 00:33:34,218 And I can't imagine 533 00:33:34,474 --> 00:33:38,163 quantum cryptography would be much different, right? Because you still 534 00:33:38,163 --> 00:33:41,821 have the— there are probably more niceties that it has in terms of being able 535 00:33:41,821 --> 00:33:45,606 to detect eavesdropping, blah, blah, blah. But the whole notion of key sharing, 536 00:33:46,055 --> 00:33:48,942 all right, key sharing, you can say you can mitigate it because I have a 537 00:33:49,488 --> 00:33:53,161 direct quantum thing that is anti-eavesdropping thing. But I can't imagine that there's 538 00:33:53,161 --> 00:33:56,465 probably got to be some weak point that can be 539 00:33:56,465 --> 00:34:00,236 exploited and blown wide open, or— Probably the most likely 540 00:34:00,316 --> 00:34:04,090 thing is there's, since these are systems that are by and large not really 541 00:34:04,171 --> 00:34:07,993 built yet, or if they are built, they're not heavily tested through years 542 00:34:08,105 --> 00:34:11,623 of rigor. There's probably weaknesses we don't know about yet. 543 00:34:12,923 --> 00:34:16,697 I can tell you in my own testing, my own red teaming and purple 544 00:34:16,697 --> 00:34:20,199 teaming, I've identified 2 threats that are proving 545 00:34:20,295 --> 00:34:22,945 pretty difficult to stop. And one's called 546 00:34:23,443 --> 00:34:27,189 cryptographic envelope bypass. That's one. Ooh, interesting. 547 00:34:28,091 --> 00:34:31,860 And then there's another one that's called operational context 548 00:34:31,893 --> 00:34:35,294 bypass. So those 2 549 00:34:35,747 --> 00:34:39,516 are already proven tough to stop. So even if you 550 00:34:39,645 --> 00:34:43,483 deploy a quantum safe, a lot of attackers have a 551 00:34:43,483 --> 00:34:47,224 way to bypass it. And so that's what I'm seeing. So whether it's 552 00:34:47,272 --> 00:34:50,837 operational context that they're using to bypass or whether it's, hey, 553 00:34:51,380 --> 00:34:55,160 I know you have this envelope that sits here on the network. 554 00:34:55,160 --> 00:34:58,800 Guess what? I can take a backdoor and go through a load 555 00:34:58,800 --> 00:35:02,620 balancer and just bypass that and authenticate on the other side. And 556 00:35:02,620 --> 00:35:06,420 I don't have to go through that particular handshake. I don't have to go 557 00:35:06,420 --> 00:35:10,140 through that, that little security garden that you have built on 558 00:35:10,140 --> 00:35:13,800 the edge. I can just go around. And to your point, I think if we 559 00:35:13,800 --> 00:35:17,380 go back through history and we look at most of the attacks that have happened 560 00:35:17,380 --> 00:35:21,140 and most of the data breaches, it's an end-user device. It's somebody's phone, 561 00:35:21,140 --> 00:35:24,960 it's somebody's laptop. And that's my bypass. And I can just go 562 00:35:24,960 --> 00:35:28,400 around and say, I'm going to go around, get on this guy's laptop and 563 00:35:28,400 --> 00:35:32,060 authenticate through this jump box, which is far 564 00:35:32,060 --> 00:35:35,760 away from the edge where all my defenses are critical. And to me, that's an 565 00:35:35,760 --> 00:35:39,480 architecture strategy gap. It's not necessarily saying the 566 00:35:39,480 --> 00:35:42,480 quantum doesn't hold, but it's just saying I'm just going to go around it. I'll 567 00:35:42,480 --> 00:35:46,220 find another way in. So those old school style stuff 568 00:35:46,220 --> 00:35:49,760 still sound like they have a— they're not going away. They're just going to 569 00:35:49,760 --> 00:35:53,549 evolve and adapt. Yeah, if folks want to— if folks 570 00:35:53,549 --> 00:35:57,353 want some hair-raising stories, you should just look up books or audiobooks 571 00:35:57,402 --> 00:36:01,174 or YouTube videos on key management. The vulnerabilities are 572 00:36:01,174 --> 00:36:04,866 always at the edge, like you said, and it's always stupid stuff. Even this, like, 573 00:36:05,203 --> 00:36:08,607 if you can narrow down the wear patterns on keyboards, 574 00:36:09,361 --> 00:36:13,053 there's— yeah, I think Black Hat just happened in Vegas. Yeah. 575 00:36:13,519 --> 00:36:17,356 I've never been, but I've always— I've heard stories about kind of like just— 576 00:36:17,452 --> 00:36:20,680 it's just the dumbest things. It's, you know, somebody who's in a cleaning crew, they 577 00:36:20,680 --> 00:36:23,440 take a picture of your keyboard, they send that picture to somebody and it's like, 578 00:36:23,440 --> 00:36:27,070 oh, this key is worn more often. And does that get them close? 579 00:36:27,100 --> 00:36:30,360 No. But instead of tens of billions of options, now you're only down to 1 580 00:36:30,360 --> 00:36:34,020 or 2 billions of options. And you could do other things to narrow that down 581 00:36:34,020 --> 00:36:37,860 to, 'cause no, no one tries. I'm not aware of 582 00:36:37,860 --> 00:36:41,660 anyone having a successful brute force attack. I don't think that the— Doesn't 583 00:36:41,660 --> 00:36:44,600 really happen. The cost of it. I don't think it does. But it's funny 'cause 584 00:36:44,600 --> 00:36:47,640 people always say it'll take 10 million years with a brute force attack. I'm like, 585 00:36:47,640 --> 00:36:50,987 does anyone do that? It's like you have this 586 00:36:51,165 --> 00:36:54,723 fancy security system, but it doesn't have backup power. 587 00:36:54,772 --> 00:36:58,313 So cut the power, then you can just walk right in, right? Exactly. 588 00:36:58,362 --> 00:37:01,780 Security, I think, is a holistic— requires 589 00:37:01,796 --> 00:37:05,392 holistic thinking, whether it's conventional or 590 00:37:05,570 --> 00:37:09,369 post-quantum, right? Absolutely. And to your 591 00:37:09,401 --> 00:37:13,039 point, I don't think the experts— I don't think there's enough expertise or 592 00:37:13,039 --> 00:37:16,866 just time to build out the battle scars of how do you make 593 00:37:16,866 --> 00:37:19,797 a real post-quantum encrypted secure thing. There are things we think will work, 594 00:37:21,923 --> 00:37:24,376 But I think there's gonna be a whole lot of learning on the way in 595 00:37:24,408 --> 00:37:28,127 the next 10, 20 years. Oh yeah, for sure. So as we're 596 00:37:28,175 --> 00:37:31,782 talking about cryptographic identity, I think this is actually the first 597 00:37:31,782 --> 00:37:35,277 show that we, that, that phrase has been brought up. So I'm just 598 00:37:35,453 --> 00:37:39,092 absolutely captivated by it. So could 599 00:37:39,140 --> 00:37:42,908 cryptographic identity become even more important as AI 600 00:37:43,068 --> 00:37:46,530 agents and autonomous systems increasingly act on our 601 00:37:46,530 --> 00:37:50,312 behalf? Oh yeah. We're already 602 00:37:50,312 --> 00:37:52,894 seeing pretty sophisticated AI attacks, right? 603 00:37:53,039 --> 00:37:56,057 We're already seeing it. And so 604 00:37:56,832 --> 00:37:59,322 to your point, cryptographic identities, 605 00:38:00,957 --> 00:38:04,420 and when you look around, people really only talk about 606 00:38:04,759 --> 00:38:08,587 encryption in kind of 2 things. I want to encrypt 607 00:38:08,652 --> 00:38:12,363 my users. So I have some type of what we 608 00:38:12,363 --> 00:38:16,101 call PKI, public key infrastructure. I have an SSH key that I use as a 609 00:38:16,260 --> 00:38:19,652 user to log in. We talk about kind of encryption of 610 00:38:19,828 --> 00:38:23,658 data. So now I have my SSL cert that kind 611 00:38:23,674 --> 00:38:27,455 of encrypts my data on a webpage, my encryption at rest, 612 00:38:27,520 --> 00:38:30,773 my encryption in transit. I want to encrypt my database. So that is now saying 613 00:38:30,981 --> 00:38:34,827 I'm using encryption to kind of help protect my data. And 614 00:38:34,827 --> 00:38:38,400 then I'm also using encryption to help protect my user identities. 615 00:38:38,737 --> 00:38:41,990 And so when we talk about identities, then now 616 00:38:42,823 --> 00:38:46,205 it— For the most part, If you want to use AI 617 00:38:46,686 --> 00:38:50,343 to now hack in, then they know, okay, I have these 2 things I'm looking 618 00:38:50,455 --> 00:38:53,968 for. I know that this user has some type of 619 00:38:53,968 --> 00:38:57,592 identity that's encrypted and coded. I know they have some type of 620 00:38:57,737 --> 00:39:01,473 access key that they're using to access. So I know there's an access key somewhere, 621 00:39:01,762 --> 00:39:05,290 whether it be at the network level or the server level, or even the app 622 00:39:05,290 --> 00:39:08,578 key, app-level key. I know it's there. But then I also know 623 00:39:09,444 --> 00:39:11,802 I want to get to the data. I know that there's also going to be 624 00:39:11,802 --> 00:39:15,202 some sort of encryption on that data. protecting the edge, 625 00:39:15,314 --> 00:39:19,099 trans— in transit and at rest. So AI knows to look for 626 00:39:19,099 --> 00:39:22,852 those, right? And so if we don't know to look for 627 00:39:22,852 --> 00:39:26,187 them and understand that I have these types of identities, this one is 628 00:39:26,364 --> 00:39:30,117 legacy, it's really old, I need to swap it out, then AI is 629 00:39:30,133 --> 00:39:32,634 just gonna have a field day or a hacker's gonna have a field day with 630 00:39:32,634 --> 00:39:36,195 AI because they're like, we already know what you have, we already know what's there, 631 00:39:36,788 --> 00:39:39,899 then now we just have to figure out where are the holes associated with it. 632 00:39:42,420 --> 00:39:46,224 So— Go ahead. I'm sorry. So I would say that's the entry point for AI. 633 00:39:46,609 --> 00:39:50,366 It's looking for those identities. And so our job is to make 634 00:39:50,366 --> 00:39:53,978 sure that those identities are secure or up to date, I would say, is 635 00:39:54,122 --> 00:39:57,814 more, more of the challenge. Because if you, 636 00:39:58,071 --> 00:40:01,844 if a credential gets revoked, you have to know about that, right? Like, but again, 637 00:40:01,860 --> 00:40:05,456 security is a holistic field, right? Like, it's 638 00:40:05,536 --> 00:40:08,667 temp— time is a factor in it. Right? If I hire you, 639 00:40:09,133 --> 00:40:12,120 Christopher, for 6 months, 6 months and 1 day, 640 00:40:13,003 --> 00:40:16,343 or midnight is your last day, that credential should 641 00:40:16,777 --> 00:40:20,551 expire, right? And I know the 642 00:40:20,551 --> 00:40:24,373 Target hack, one of the issues was the credentials 643 00:40:24,405 --> 00:40:28,018 were for, I think it was an HVAC vendor allowed 644 00:40:28,500 --> 00:40:32,258 bad activity, something like, and it didn't expire, right? And then 645 00:40:32,258 --> 00:40:35,863 the attackers were able to leverage that. And I mean, 646 00:40:37,445 --> 00:40:40,479 it's easy to bust on them 'cause they're a big company and like you think 647 00:40:40,479 --> 00:40:43,931 they should know better, but like I wouldn't be so arrogant about it. One, 648 00:40:44,269 --> 00:40:47,933 I've had intermittent experiments with my own home labbing experience. Fun 649 00:40:48,046 --> 00:40:51,758 fact, in actually about, I 650 00:40:51,758 --> 00:40:54,804 was still living in New Jersey, so this would've had about 20 years ago, I 651 00:40:54,804 --> 00:40:58,306 decided to host my own web server. I'm like, this is great. I'm learning and 652 00:40:58,306 --> 00:41:02,130 I'm doing this. And I set up an FTP server and then 653 00:41:02,163 --> 00:41:05,498 for some reason I turned the password off. Oh. I did 654 00:41:05,546 --> 00:41:08,388 something stupid to the password. So then, 655 00:41:09,384 --> 00:41:13,061 and I forgot about it cuz it was just a temporary measure. Why I, as 656 00:41:13,173 --> 00:41:15,726 much as I'm tempted to make fun of people when they make these gaffes, I'm 657 00:41:15,726 --> 00:41:19,163 like, eh, there but for the grace of God go I, right? So 658 00:41:19,741 --> 00:41:22,888 over the course of the next few weeks, my internet connection started really 659 00:41:23,530 --> 00:41:27,095 dropping to like, I had no bandwidth. So I'm like, ah, 660 00:41:27,368 --> 00:41:30,178 I called cable, I called the cable company. They were like, we don't know what's 661 00:41:30,178 --> 00:41:33,853 going on. And so then Now I'm older and 662 00:41:33,885 --> 00:41:37,670 wiser. There's other things I would do, test network traffic, Wireshark, all that. 663 00:41:38,119 --> 00:41:41,551 But I was young. I was 20-some-odd years younger and 664 00:41:41,984 --> 00:41:45,720 not as wise. But, and then I was looking through the hard drive because suddenly 665 00:41:45,736 --> 00:41:48,334 I was, the next thing I noticed, my hard drive was being filled up. Oh, 666 00:41:48,398 --> 00:41:51,974 that's weird. And I'm like, that's weird. And then I noticed that 667 00:41:52,007 --> 00:41:55,519 there are all these, somebody had found my website, my 668 00:41:55,535 --> 00:41:58,213 FTP, and loaded it with PlayStation 2 ROMs. 669 00:41:59,368 --> 00:42:02,995 And basically was a massive pirating operation on my server. 670 00:42:04,424 --> 00:42:08,115 And I was like, how could that happen? And then I found out how 671 00:42:08,131 --> 00:42:11,438 they got in. They got in because I temporarily did it and I forgot to 672 00:42:11,486 --> 00:42:15,033 close the window. So it's very easy to make fun of companies when they make 673 00:42:15,033 --> 00:42:18,644 these huge mistakes like this, but it can happen. And if you think you're 674 00:42:18,644 --> 00:42:22,079 secure, you're not. If you have that level of arrogance and 675 00:42:22,079 --> 00:42:25,626 confidence, you don't. Because even people in industry, as you have to 676 00:42:25,626 --> 00:42:29,093 test it, you have to hold your own things. So fast forward to what, a 677 00:42:29,093 --> 00:42:31,967 few months ago, Candace. So we're, Candace and I are working on a product called 678 00:42:31,967 --> 00:42:35,725 the Dingo, thedingo.ai, which maybe will be launched by the time you're 679 00:42:35,725 --> 00:42:39,241 listening. So when Vercel had their breach, 680 00:42:40,638 --> 00:42:44,460 right? She asked me, are we vulnerable? And I was like, no, no, 681 00:42:44,476 --> 00:42:47,896 because I had all the passwords and user data encrypted from the get-go 682 00:42:48,666 --> 00:42:52,504 because I vibe coded 90% of it and it 683 00:42:52,504 --> 00:42:56,064 was just Doing it myself would've taken X number of hours. Vibe coding it 684 00:42:56,128 --> 00:42:59,703 would've been, would've taken an extra sentence in a prompt and 685 00:43:00,056 --> 00:43:03,744 a few hundred thousand more tokens. So I did it. Plus I, my real 686 00:43:03,744 --> 00:43:07,463 inspiration was, God, if I'm storing plaintext information 687 00:43:07,976 --> 00:43:11,616 on a passwords on a server, my wife will never 688 00:43:11,680 --> 00:43:14,742 let me forget, live that down. 689 00:43:15,752 --> 00:43:19,472 So I actually did it and, which seemed like overkill when I 690 00:43:19,472 --> 00:43:23,225 made that decision. But then again, 6 months later. when the, when the hack 691 00:43:23,225 --> 00:43:26,129 happened, or a few months later when the hack happened, I can be 692 00:43:26,915 --> 00:43:30,365 not so worried about it because I took that extra step. That's why I like 693 00:43:30,365 --> 00:43:33,750 to say today's paranoia is tomorrow's best practice. 694 00:43:35,162 --> 00:43:38,772 It is. It goes back to the paranoia, always back to the paranoia. 695 00:43:39,510 --> 00:43:42,687 I always think to that, that FTP server that became this massive 696 00:43:42,960 --> 00:43:46,682 PlayStation pirating hub, which, you know, thankfully the, the 697 00:43:47,035 --> 00:43:50,677 FBI never knocked down my door. I wasn't running it. If you're looking at the 698 00:43:51,062 --> 00:43:53,564 The IP traffic, it would've not been a good look for me. 699 00:43:54,206 --> 00:43:58,024 Absolutely. Oh, that was probably helped that 700 00:43:58,024 --> 00:44:00,751 I didn't have a PlayStation at the time. If I had a PlayStation at the 701 00:44:00,751 --> 00:44:04,152 time, I would've been, I would've probably taken— I discovered that I immediately pulled the 702 00:44:04,152 --> 00:44:07,328 network cable outta the wall cuz I'm like, and then I had friends tell me 703 00:44:07,408 --> 00:44:10,344 and make fun of me and then run this when you plug it in. And 704 00:44:10,376 --> 00:44:13,215 sure enough, when I plugged everything back in, there was a lot of these failed 705 00:44:13,215 --> 00:44:16,199 requests that kind of came out and then, and then I dropped it from the— 706 00:44:16,408 --> 00:44:20,065 so you live and you learn. Right? And if you're smart, you can learn from 707 00:44:20,113 --> 00:44:23,813 other people's mistakes. Oh yeah. We've had a 708 00:44:23,813 --> 00:44:27,046 great conversation with you. Your Quantum Home Lab, man, we got to— we have to 709 00:44:27,046 --> 00:44:30,345 have a geeky deep dive on that one, man. I'd love to talk to you 710 00:44:30,345 --> 00:44:32,773 about that and invite you back on the show. But we are coming close to 711 00:44:32,773 --> 00:44:35,897 time. Where can folks find out more about you and what you're doing? 712 00:44:37,829 --> 00:44:41,631 I try to be pretty active on LinkedIn, so definitely people can reach out 713 00:44:41,631 --> 00:44:45,141 to me there. I do speaking engagements. I have a few 714 00:44:45,173 --> 00:44:48,748 conferences coming up, so Yeah, if it's a major quantum 715 00:44:49,189 --> 00:44:52,759 conference, I'm probably there. You can catch me there. I 716 00:44:52,759 --> 00:44:56,089 think my LinkedIn, I have some 717 00:44:56,607 --> 00:45:00,261 contact information on there so that they can reach out to me. But 718 00:45:00,261 --> 00:45:04,049 yeah, so I work obviously, but then I kind of build my own stuff 719 00:45:04,049 --> 00:45:07,551 in my home lab. And then I also work with this biology 720 00:45:07,583 --> 00:45:11,214 group and we're working on some quantum biology stuff. So I do that as 721 00:45:11,327 --> 00:45:14,884 well. Oh, interesting. Yeah. They're called the Burden Center for Spatial 722 00:45:14,884 --> 00:45:18,120 Biology. It's a nonprofit. But we, we're doing some pretty cool 723 00:45:18,120 --> 00:45:21,840 research on quantum storage. And so trying to 724 00:45:21,840 --> 00:45:25,680 work on quantum storage. Yeah. You can run into me through KPMG. 725 00:45:25,680 --> 00:45:29,240 You can run into me on LinkedIn. Those will be the main 726 00:45:29,240 --> 00:45:32,860 avenues. And then the major conferences, I'll, you can expect me to be there. 727 00:45:33,460 --> 00:45:36,920 Awesome. Any parting thoughts, Candace? Honestly, I love 728 00:45:36,920 --> 00:45:40,440 this. I love cryptographic identities. I can't wait to talk 729 00:45:40,440 --> 00:45:44,200 more about this with you. So you have an open invitation to 730 00:45:44,200 --> 00:45:46,974 come back for sure. This was wonderful. We really enjoyed it. 731 00:45:47,890 --> 00:45:51,331 Awesome. I enjoyed it too. You guys asked me some tough questions. You really made 732 00:45:51,331 --> 00:45:53,662 me scratch my head a little bit. It's all right. 733 00:45:54,080 --> 00:45:57,713 You'll be ready. You'll be stronger for next time, right? Yeah, 734 00:45:57,841 --> 00:46:01,057 absolutely. And with that, we'll let the outro music play. 735 00:46:01,941 --> 00:46:05,638 Oh, that was great, man. That was fantastic. That was— really enjoyed it.