1 00:00:00,160 --> 00:00:02,960 Speaker 1: Hey everyone, it's Robert and Joe here. Today we've got 2 00:00:02,960 --> 00:00:05,240 Speaker 1: something a little bit different to share with you. It's 3 00:00:05,280 --> 00:00:09,400 Speaker 1: a new season of the Smart Talks with IBM podcast series. 4 00:00:09,720 --> 00:00:13,520 Speaker 2: This season on Smart Talks with IBM, Malcolm Gladwell is back, 5 00:00:13,520 --> 00:00:15,760 Speaker 2: and this time he's taking the show on the road. 6 00:00:15,880 --> 00:00:19,160 Speaker 2: Malcolm is stepping outside the studio to explore how IBM 7 00:00:19,239 --> 00:00:22,880 Speaker 2: clients are using artificial intelligence to solve real world challenges 8 00:00:23,079 --> 00:00:25,520 Speaker 2: and transform the way they do business. 9 00:00:25,480 --> 00:00:29,560 Speaker 1: From accelerating scientific breakthroughs to reimagining education. It's a fresh 10 00:00:29,600 --> 00:00:33,040 Speaker 1: look at innovation in action, where big ideas meet cutting 11 00:00:33,120 --> 00:00:33,960 Speaker 1: edge solutions. 12 00:00:34,320 --> 00:00:37,160 Speaker 2: You'll hear from industry leaders, creative thinkers, and of course 13 00:00:37,280 --> 00:00:40,920 Speaker 2: Malcolm Gladwell himself as he guides you through each story. 14 00:00:41,440 --> 00:00:44,640 Speaker 1: New episodes of Smart Talks with IBM drop every month 15 00:00:44,680 --> 00:00:48,279 Speaker 1: on the iHeartRadio app, Apple Podcasts, or wherever you get 16 00:00:48,320 --> 00:00:52,240 Speaker 1: your podcasts. Learn more at IBM dot com slash smart Talks. 17 00:00:52,880 --> 00:00:56,000 Speaker 1: This is a paid advertisement from IBM. 18 00:00:56,160 --> 00:00:58,720 Speaker 3: Hello, this is Malcolm Gladwell and you're listening to Smart 19 00:00:58,720 --> 00:01:02,160 Speaker 3: Talks with IBM. Every year, Tech Week brings thousands of 20 00:01:02,240 --> 00:01:05,640 Speaker 3: people together to network and learn about what's emerging across 21 00:01:05,680 --> 00:01:10,039 Speaker 3: the technology ecosystem, and at this year's conference in San Francisco, 22 00:01:10,400 --> 00:01:13,120 Speaker 3: I had an amazing opportunity to sit down in front 23 00:01:13,120 --> 00:01:16,160 Speaker 3: of a live audience with Jay Gambetta. Jay has been 24 00:01:16,200 --> 00:01:19,520 Speaker 3: with IBM for years and was recently promoted to Director 25 00:01:19,560 --> 00:01:23,280 Speaker 3: of Research. In this job, Jay has an important mission 26 00:01:23,720 --> 00:01:27,080 Speaker 3: helping the company build the future of computing. In the 27 00:01:27,160 --> 00:01:29,759 Speaker 3: last episode of smart Talks, I began to learn about 28 00:01:29,800 --> 00:01:34,679 Speaker 3: quantum computing from IBM Chairman and CEO Arvind Krishna. But 29 00:01:34,800 --> 00:01:38,200 Speaker 3: this conversation I had with Jay went even deeper and 30 00:01:38,360 --> 00:01:42,000 Speaker 3: convinced me that the development of quantum isn't just a fun, 31 00:01:42,360 --> 00:01:45,560 Speaker 3: exciting new paradigm of computing, it may be one of 32 00:01:45,560 --> 00:01:55,960 Speaker 3: the most important scientific achievements of my lifetime. Jay, Good morning, morning, 33 00:01:56,280 --> 00:01:59,320 Speaker 3: Welcome to Smart Talks with IBM. Thank you special live 34 00:01:59,360 --> 00:02:02,960 Speaker 3: recording here for tech Week and congratulations. How long have 35 00:02:03,000 --> 00:02:05,160 Speaker 3: you been Head of Research at IBM? 36 00:02:05,680 --> 00:02:09,320 Speaker 4: Since October one? It's October tenth today, since nine days, 37 00:02:09,440 --> 00:02:10,000 Speaker 4: nine days. 38 00:02:10,400 --> 00:02:12,800 Speaker 3: Can you just talk a little about the position. This 39 00:02:13,000 --> 00:02:17,200 Speaker 3: is one of the most important positions in research in 40 00:02:17,240 --> 00:02:17,680 Speaker 3: the world. 41 00:02:18,200 --> 00:02:21,280 Speaker 4: IBM research has been around for eighty years and it's 42 00:02:21,320 --> 00:02:26,320 Speaker 4: done some tremendous technology, a lot of inventions and fundamentals 43 00:02:26,400 --> 00:02:31,519 Speaker 4: for semiconductors, algorithms, AI. Yeah, I think if we look 44 00:02:31,560 --> 00:02:33,520 Speaker 4: back to where a lot of the innovation and the 45 00:02:33,560 --> 00:02:36,120 Speaker 4: technology of the world comes from, I think you can 46 00:02:36,160 --> 00:02:39,680 Speaker 4: find Ibram's footprints on it, and you can find IBM research. 47 00:02:39,800 --> 00:02:43,040 Speaker 4: So yeah, I'm very excited for the opportunity, but I'm 48 00:02:43,080 --> 00:02:45,640 Speaker 4: also aware that there's big shoes to fill, and I'm 49 00:02:45,720 --> 00:02:50,400 Speaker 4: looking forward to how we take IBM research forward. Obviously, 50 00:02:50,520 --> 00:02:53,440 Speaker 4: I'm going to be bringing a lot of the quantum side, 51 00:02:53,440 --> 00:02:57,000 Speaker 4: which we're going to talk about later. Beyond quantum, there's 52 00:02:57,080 --> 00:03:00,520 Speaker 4: important work that needs to happen in AI hybrid cloud, 53 00:03:01,040 --> 00:03:03,400 Speaker 4: and I think we're going to also enter into this 54 00:03:03,440 --> 00:03:06,880 Speaker 4: new period of mathematics where we get to use quantum 55 00:03:06,919 --> 00:03:11,000 Speaker 4: machines and also AI machines. And there's some really good 56 00:03:11,200 --> 00:03:13,400 Speaker 4: hard mathematical questions to answer. 57 00:03:13,520 --> 00:03:15,120 Speaker 3: How many people do you have working for you? 58 00:03:15,480 --> 00:03:19,240 Speaker 4: I've been researchers in the three thousand researchers across many 59 00:03:19,240 --> 00:03:22,520 Speaker 4: different labs around the world. Our main lab is in Yorktown, 60 00:03:22,600 --> 00:03:24,680 Speaker 4: but then we have the lab actually out on the 61 00:03:24,760 --> 00:03:28,160 Speaker 4: West coast in Armadan or SBL Now and then we 62 00:03:28,240 --> 00:03:31,560 Speaker 4: have one in Zurich, Japan, and a few others around 63 00:03:31,560 --> 00:03:31,919 Speaker 4: the world. 64 00:03:32,600 --> 00:03:35,080 Speaker 3: Tell me a little bit before we get into quantum. 65 00:03:35,240 --> 00:03:38,520 Speaker 3: I'm just curious about your path. So you're Australian. Yep, 66 00:03:38,920 --> 00:03:42,320 Speaker 3: we were talking about earlier backstage. Your accent has become muted. 67 00:03:42,880 --> 00:03:45,600 Speaker 3: You should crank it up because it's. 68 00:03:45,560 --> 00:03:48,360 Speaker 4: Yeah, I'm slowly losing my Australian accent. I've been in 69 00:03:48,400 --> 00:03:51,400 Speaker 4: the US since two thousand and four, so accent, you know, 70 00:03:51,480 --> 00:03:54,200 Speaker 4: to sound very Australian. Yeah, but how do you practice it? 71 00:03:54,280 --> 00:03:57,320 Speaker 4: Maybe I got to go back to Australia. Here a 72 00:03:57,480 --> 00:04:00,080 Speaker 4: more Australians say gooday, how's it going? 73 00:04:00,200 --> 00:04:00,560 Speaker 2: Like that? 74 00:04:01,520 --> 00:04:04,040 Speaker 3: And you you didn't grow up thinking you're going to 75 00:04:04,080 --> 00:04:05,560 Speaker 3: be a scientist one day, now. 76 00:04:05,600 --> 00:04:09,640 Speaker 4: I grew up in a pretty normal life. My dreams 77 00:04:09,880 --> 00:04:12,560 Speaker 4: as a kid was building things, so I was either 78 00:04:12,600 --> 00:04:15,640 Speaker 4: going to be a carpenter or a mechanic. But I 79 00:04:15,680 --> 00:04:18,600 Speaker 4: had some great teachers that inspired me to go to university. 80 00:04:18,680 --> 00:04:21,120 Speaker 4: And I didn't even know, honestly what a scientist was. 81 00:04:21,440 --> 00:04:25,960 Speaker 4: And then I found myself at university doing science, particular physics, 82 00:04:26,080 --> 00:04:27,360 Speaker 4: and I ended up loving it. 83 00:04:27,680 --> 00:04:30,520 Speaker 3: So you go from there to what do you do 84 00:04:30,560 --> 00:04:31,120 Speaker 3: your PhD. 85 00:04:31,279 --> 00:04:33,800 Speaker 4: So I did my undergrad in Australia. I did it 86 00:04:33,839 --> 00:04:40,279 Speaker 4: actually in laser science, so I think I watched some 87 00:04:40,440 --> 00:04:43,000 Speaker 4: TV show in lasers seemed interesting, so I wanted to 88 00:04:43,080 --> 00:04:46,279 Speaker 4: learn about lasers. And then I realized in trying to 89 00:04:46,360 --> 00:04:50,360 Speaker 4: understand lasers there was this quantum mechanics, and so I 90 00:04:50,400 --> 00:04:53,279 Speaker 4: was like, all right, I want to actually understand this 91 00:04:53,360 --> 00:04:55,520 Speaker 4: quantum mechanics. So I did my equivalent of what you 92 00:04:55,600 --> 00:04:58,680 Speaker 4: and the US school masters. We call it honors in Australia, 93 00:04:58,720 --> 00:05:01,320 Speaker 4: but we do a research project. I said, I wanted 94 00:05:01,320 --> 00:05:05,760 Speaker 4: to shoot lasers into atoms and measure cross sections and 95 00:05:05,800 --> 00:05:08,440 Speaker 4: I got really into quantum physics. So then I decided, 96 00:05:08,440 --> 00:05:11,599 Speaker 4: all right, I don't understand this quantum physics. I want 97 00:05:11,640 --> 00:05:16,040 Speaker 4: to do my PhD in interpretations of quantum mechanics. So 98 00:05:16,080 --> 00:05:17,839 Speaker 4: I jumped in and said, all right, what is this 99 00:05:17,920 --> 00:05:21,760 Speaker 4: quantum mechanics? Why is everyone arguing on these different interpretations. 100 00:05:22,240 --> 00:05:25,360 Speaker 4: Then I finished my PhD in Australia doing that. Then 101 00:05:25,400 --> 00:05:28,720 Speaker 4: I moved over at the end of my PhD interpretations, 102 00:05:29,720 --> 00:05:34,479 Speaker 4: it's more people arguing about the equations whilst I think 103 00:05:34,480 --> 00:05:37,279 Speaker 4: it's really important. I decided if it's going to be 104 00:05:37,400 --> 00:05:40,760 Speaker 4: like a collapse equation versus many worlds, or a hidden 105 00:05:40,800 --> 00:05:46,520 Speaker 4: variable model, or that just quantum mechanics decoheres because we 106 00:05:46,560 --> 00:05:50,200 Speaker 4: don't see supersitions in the everyday world, because it interacts 107 00:05:50,240 --> 00:05:53,400 Speaker 4: with environment. The only way to answer that question was 108 00:05:53,440 --> 00:05:56,520 Speaker 4: to build a quantum computer. And so then I decided 109 00:05:56,520 --> 00:05:58,960 Speaker 4: at the end of my PhD, I wanted to work 110 00:05:59,000 --> 00:06:01,599 Speaker 4: out how to build a quant computer. And then I 111 00:06:01,680 --> 00:06:05,080 Speaker 4: left there and I went to Yale, and then at Yale, 112 00:06:05,160 --> 00:06:08,479 Speaker 4: that's where I got into superconducting cubits, which just a 113 00:06:08,480 --> 00:06:10,839 Speaker 4: few days ago one of the professors there just won 114 00:06:11,120 --> 00:06:12,599 Speaker 4: the Nobel Prize this year. 115 00:06:12,960 --> 00:06:16,000 Speaker 3: Oh wow, Yeah, I'm very interested in tracing because your 116 00:06:16,080 --> 00:06:20,880 Speaker 3: career follows the arc of quantum computing in a certain way. 117 00:06:20,960 --> 00:06:24,000 Speaker 3: Right at the time when you asked the question, what 118 00:06:24,040 --> 00:06:25,680 Speaker 3: I really want to do is to figure out how 119 00:06:25,680 --> 00:06:28,600 Speaker 3: to build a quantum computer. Where are we in quantum 120 00:06:28,680 --> 00:06:29,880 Speaker 3: computing at that point? 121 00:06:30,040 --> 00:06:32,839 Speaker 4: Yeah, So that would have been nineteen ninety So there 122 00:06:32,960 --> 00:06:36,800 Speaker 4: was Shaw's algorithm came out, let's say ninety five. There 123 00:06:36,839 --> 00:06:39,839 Speaker 4: was a lot of theory. And then the reason I 124 00:06:39,880 --> 00:06:44,080 Speaker 4: went to Yale is because people had started to show 125 00:06:44,160 --> 00:06:48,839 Speaker 4: that they could see quantum effects in electrical circuits. So 126 00:06:48,920 --> 00:06:52,240 Speaker 4: these macroscopic objects they were starting to behave quantum mechanical 127 00:06:52,800 --> 00:06:55,760 Speaker 4: There was a really significant breakthrough in nineteen ninety nine 128 00:06:55,880 --> 00:06:59,960 Speaker 4: where Yazoo Nakamura in Japan showed that a qubit could 129 00:07:00,160 --> 00:07:03,640 Speaker 4: exist in these electrical circuits. And then I found out 130 00:07:03,680 --> 00:07:05,760 Speaker 4: the group at Yale were really trying to take these 131 00:07:05,800 --> 00:07:10,560 Speaker 4: electrical circuits and couple them together. And so it was like, 132 00:07:10,800 --> 00:07:14,880 Speaker 4: if I can build something using electrical circuits and they're big, 133 00:07:15,440 --> 00:07:18,160 Speaker 4: that that's the best way that you cancide to test 134 00:07:18,240 --> 00:07:21,920 Speaker 4: and understand whether quantum mechanics breaks down at a macroscopic 135 00:07:21,960 --> 00:07:25,040 Speaker 4: scale or not. Can we actually make them behave as cubits? 136 00:07:25,440 --> 00:07:27,800 Speaker 4: And I agree. When I came to Yale, the cubits 137 00:07:28,560 --> 00:07:32,840 Speaker 4: were not very good. They were actually a couple of nanoseconds. 138 00:07:33,080 --> 00:07:37,160 Speaker 4: They were unstable. Electron would jump onto the chip and 139 00:07:37,200 --> 00:07:39,760 Speaker 4: then they would change all their configurations, so you have 140 00:07:39,840 --> 00:07:43,160 Speaker 4: to restart your experiment. And so for the first time 141 00:07:43,200 --> 00:07:45,720 Speaker 4: at Yale, it's kind of what the challenge there was, 142 00:07:45,880 --> 00:07:48,200 Speaker 4: how do we make a cubit? How do we make 143 00:07:48,200 --> 00:07:51,640 Speaker 4: a stable cubit? And that took about five years and 144 00:07:51,760 --> 00:07:54,760 Speaker 4: that took us up to two thousand and seven. And 145 00:07:55,120 --> 00:07:56,960 Speaker 4: I think the rest of the world looks and says 146 00:07:57,040 --> 00:08:00,480 Speaker 4: quantums like just blowing up. But it's actually been like 147 00:08:00,560 --> 00:08:03,920 Speaker 4: all most phases theory, showing that we got the algorithms, 148 00:08:04,280 --> 00:08:06,480 Speaker 4: how do we make a cubit? How do we couple 149 00:08:06,480 --> 00:08:09,280 Speaker 4: of the cubits together? And now we're in the scaling phase. 150 00:08:09,760 --> 00:08:14,240 Speaker 3: Describe for us, because many people in this room, me included, 151 00:08:14,760 --> 00:08:18,000 Speaker 3: have only a kind of surface level understanding of what 152 00:08:18,040 --> 00:08:20,320 Speaker 3: we mean when we use that phrase. What is the 153 00:08:20,360 --> 00:08:24,040 Speaker 3: difference between classical computing and quantum computing? What does that 154 00:08:24,080 --> 00:08:24,600 Speaker 3: word mean? 155 00:08:25,000 --> 00:08:28,320 Speaker 4: Yeah, so you can go down the physics way and 156 00:08:28,360 --> 00:08:31,320 Speaker 4: talk about supersition and entanglement, which we can go in later, 157 00:08:31,360 --> 00:08:34,599 Speaker 4: but actually feel it's a bit of a distraction. So 158 00:08:34,960 --> 00:08:39,120 Speaker 4: when you think of classical computers, what they were is 159 00:08:39,160 --> 00:08:42,520 Speaker 4: there were machines that were very good at adding numbers together, 160 00:08:43,040 --> 00:08:46,920 Speaker 4: like simple addition, and they really showed that they could 161 00:08:47,440 --> 00:08:50,640 Speaker 4: add these numbers together really really fast. And now with 162 00:08:50,960 --> 00:08:54,480 Speaker 4: GPUs and other AI accelerators, we can add those numbers 163 00:08:54,480 --> 00:08:58,360 Speaker 4: together in parallel, and so the whole classical computing can 164 00:08:58,400 --> 00:09:02,200 Speaker 4: come down to just arithmetic, just adding numbers together. It 165 00:09:02,280 --> 00:09:06,560 Speaker 4: turns out that there's a math that is the quantum 166 00:09:06,559 --> 00:09:07,520 Speaker 4: mechanics shown. 167 00:09:07,320 --> 00:09:07,800 Speaker 3: To be true. 168 00:09:08,320 --> 00:09:11,880 Speaker 4: It's more like a group theory type structure. And the 169 00:09:11,880 --> 00:09:15,080 Speaker 4: way quantum works is it has a different mathos are primitive, 170 00:09:15,520 --> 00:09:18,240 Speaker 4: and if we can exploit that new math and build 171 00:09:18,240 --> 00:09:20,560 Speaker 4: a machine that does it, it allows us to answer 172 00:09:20,559 --> 00:09:23,640 Speaker 4: different questions. And so think of it as a branching 173 00:09:23,760 --> 00:09:27,120 Speaker 4: from classical compute that is very good at adding just 174 00:09:27,200 --> 00:09:30,320 Speaker 4: numbers together to something that allows us to work with 175 00:09:30,400 --> 00:09:35,200 Speaker 4: an algebra that is much much harder to represent with addition. 176 00:09:35,920 --> 00:09:38,520 Speaker 4: And that algebra happens to be the same algebra that 177 00:09:38,600 --> 00:09:42,520 Speaker 4: defines the fundamental equations of nature shirting as equation. So 178 00:09:42,760 --> 00:09:44,800 Speaker 4: this is why you say it computes the same way 179 00:09:44,880 --> 00:09:48,360 Speaker 4: nature does. But there are many other interesting problems. So 180 00:09:48,760 --> 00:09:51,120 Speaker 4: the way I explain it to people is think of 181 00:09:51,200 --> 00:09:56,200 Speaker 4: it as bringing a new primitive to computer science and 182 00:09:56,400 --> 00:10:00,160 Speaker 4: allowing us to work how to go with it. And 183 00:10:00,240 --> 00:10:03,240 Speaker 4: I like the analogy. Well, actually, maybe go back. So 184 00:10:04,120 --> 00:10:06,080 Speaker 4: if you went back in time, so we're one hundred 185 00:10:06,120 --> 00:10:08,600 Speaker 4: years of quantum, and you went back in time and 186 00:10:08,640 --> 00:10:12,160 Speaker 4: you asked, what is the foundation? Is a chemistry or physics. 187 00:10:12,559 --> 00:10:14,719 Speaker 4: What would have probably the scientists of one hundred years 188 00:10:14,760 --> 00:10:17,040 Speaker 4: ago would have said is they would have said, you know, 189 00:10:17,120 --> 00:10:20,400 Speaker 4: chemistry is about the small, physics is about planets and 190 00:10:20,440 --> 00:10:24,640 Speaker 4: things like this. And one hundred years ago when Heisenberg 191 00:10:25,200 --> 00:10:30,239 Speaker 4: or Einstein, all the greats, Schrodinger himself invented quantum mechanics, 192 00:10:30,679 --> 00:10:33,679 Speaker 4: it was this concept that nature is discreete not continuous. 193 00:10:34,200 --> 00:10:37,480 Speaker 4: It actually brought all the physical sciences together. And now 194 00:10:37,600 --> 00:10:41,640 Speaker 4: quantum mechanics is like it is the foundation of the science. 195 00:10:42,480 --> 00:10:45,520 Speaker 4: And so now what quantum computing is by that analogy 196 00:10:45,600 --> 00:10:48,559 Speaker 4: is computer science. The foundation of the math is coming 197 00:10:48,559 --> 00:10:52,000 Speaker 4: together with the physical science to allow us to compute 198 00:10:52,480 --> 00:10:55,400 Speaker 4: using math that if you were to try to represent 199 00:10:55,440 --> 00:10:58,880 Speaker 4: it with classical computers, it takes exponential time. 200 00:10:59,360 --> 00:11:03,120 Speaker 3: Yeah, it was a classical computer and an expense in 201 00:11:03,160 --> 00:11:05,360 Speaker 3: a way that someone is well informed as I am 202 00:11:05,440 --> 00:11:09,440 Speaker 3: can understand it. A customer computer works primarily on problems 203 00:11:09,440 --> 00:11:12,920 Speaker 3: that can be easily represented in numerical form in numbers. Yes, 204 00:11:13,240 --> 00:11:15,920 Speaker 3: quantum allows you to step outside to a class of 205 00:11:15,960 --> 00:11:20,360 Speaker 3: problems that don't necessarily have a simple numerical representation. 206 00:11:20,600 --> 00:11:25,360 Speaker 4: Yeah, and so imagine I got some medicine or or 207 00:11:25,400 --> 00:11:28,440 Speaker 4: some set of operation, but call it A and I 208 00:11:28,520 --> 00:11:31,800 Speaker 4: then follow it by a different operation B if A 209 00:11:31,960 --> 00:11:35,200 Speaker 4: followed by B gave a different answer than B first 210 00:11:35,360 --> 00:11:38,480 Speaker 4: followed by A. So in mathematics we call that commuting. 211 00:11:38,679 --> 00:11:41,280 Speaker 4: But like you can think of a correlation there one 212 00:11:41,480 --> 00:11:44,400 Speaker 4: one gives you a different outcome to the other. That 213 00:11:44,480 --> 00:11:48,560 Speaker 4: means there's an algebra behind it. That Representing that algebra 214 00:11:48,640 --> 00:11:53,079 Speaker 4: traditionally on classical computers is really really hard, whereas that algebra, 215 00:11:53,440 --> 00:11:55,840 Speaker 4: if we can get creative, we can come up with 216 00:11:55,880 --> 00:11:58,760 Speaker 4: ways of representing that math. So we step as you say, 217 00:11:59,000 --> 00:12:02,800 Speaker 4: we step out aside of the simple math to a 218 00:12:02,840 --> 00:12:06,240 Speaker 4: new math to allow us to calculate interesting problems. 219 00:12:06,640 --> 00:12:10,360 Speaker 3: So quite in a sense, compliments it doesn't replace judicial. 220 00:12:10,520 --> 00:12:10,840 Speaker 3: That's good. 221 00:12:11,559 --> 00:12:13,320 Speaker 4: I think this is one of the this is you're 222 00:12:13,480 --> 00:12:16,199 Speaker 4: exactly on is people think quantum is going to be 223 00:12:16,320 --> 00:12:19,400 Speaker 4: replacing classical If your problem is good at adding numbers together, 224 00:12:19,679 --> 00:12:23,120 Speaker 4: you should just keep using classical computers. I think the 225 00:12:23,160 --> 00:12:26,480 Speaker 4: future is going to be heterogeneous accelerators, and it will 226 00:12:26,520 --> 00:12:29,640 Speaker 4: definitely have quantum as one. But in some sense, the 227 00:12:29,679 --> 00:12:32,800 Speaker 4: next generation of superstars are going to be those applied 228 00:12:32,840 --> 00:12:36,800 Speaker 4: mathematicians that know, how do I write a problem using 229 00:12:36,840 --> 00:12:40,240 Speaker 4: the simple math of classical computers or the more complicated 230 00:12:40,280 --> 00:12:43,960 Speaker 4: math for quantum computers, and how do I actually iterate 231 00:12:44,040 --> 00:12:46,760 Speaker 4: between them? And things like this? This is where I 232 00:12:46,760 --> 00:12:49,559 Speaker 4: think the next generation of students are going to come 233 00:12:49,640 --> 00:12:51,560 Speaker 4: up with much more novel ideas. I can give you 234 00:12:51,600 --> 00:12:54,080 Speaker 4: examples of what we want to do on quantum, but like, 235 00:12:54,120 --> 00:12:59,520 Speaker 4: you're giving them a fundamental, foundational new thing, and so 236 00:12:59,640 --> 00:13:03,520 Speaker 4: I'm optimistic that will do much better jobs than my generation. 237 00:13:03,640 --> 00:13:06,160 Speaker 3: Well, yeah, we're to get to some of the albums 238 00:13:06,160 --> 00:13:08,640 Speaker 3: in a moment, but I wanted you to the most 239 00:13:08,760 --> 00:13:11,600 Speaker 3: kind of down to you said, as a kid, you 240 00:13:11,679 --> 00:13:13,360 Speaker 3: thought you might want to be a mechanic because you'd 241 00:13:13,360 --> 00:13:17,080 Speaker 3: like to build things. Describe to me what it takes 242 00:13:17,120 --> 00:13:20,439 Speaker 3: to build a quantum computer, Like, what are you doing 243 00:13:20,520 --> 00:13:22,360 Speaker 3: that's different from building a classical computer. 244 00:13:22,600 --> 00:13:25,240 Speaker 4: Yeah, so maybe I'll give you analogy and then i'll 245 00:13:25,240 --> 00:13:28,800 Speaker 4: go in so the way classical computers, we've got them 246 00:13:28,840 --> 00:13:33,960 Speaker 4: to get to smaller and smaller sizes like five seven animeters, 247 00:13:34,000 --> 00:13:38,960 Speaker 4: five animeters and things is actually inventing material to kill 248 00:13:39,040 --> 00:13:43,680 Speaker 4: quantum effects, so you actually put dielectrics and other things 249 00:13:43,720 --> 00:13:46,680 Speaker 4: in there. To kill the quantum tunneling effects, and you 250 00:13:46,760 --> 00:13:51,600 Speaker 4: want them to behave more classically. In the quantum world, 251 00:13:51,720 --> 00:13:53,960 Speaker 4: you want to get rid of all the classical effects, 252 00:13:54,360 --> 00:13:56,760 Speaker 4: So you want to get rid of the ability of 253 00:13:56,800 --> 00:13:59,680 Speaker 4: the cubits to interact with the environment. And in the 254 00:14:00,080 --> 00:14:02,160 Speaker 4: in the sort of technical world, we call it this 255 00:14:02,440 --> 00:14:05,400 Speaker 4: quantum conflict. The more ways you want to control the 256 00:14:05,480 --> 00:14:09,640 Speaker 4: quantum computer, you open it up to interacting with everything else, 257 00:14:09,880 --> 00:14:13,200 Speaker 4: like interacting with its environment. So the biggest challenge has 258 00:14:13,200 --> 00:14:17,800 Speaker 4: always been how do we give more control but don't 259 00:14:17,840 --> 00:14:21,280 Speaker 4: bring in other sources of noise. So I want to 260 00:14:21,320 --> 00:14:24,360 Speaker 4: be able to do gates on the cubit, but I 261 00:14:24,400 --> 00:14:27,760 Speaker 4: don't want it to decohere. I want to couple the cubits, 262 00:14:28,040 --> 00:14:30,400 Speaker 4: but I don't want them to couple to other things. 263 00:14:30,800 --> 00:14:35,120 Speaker 4: So the hardest challenge is the energy inside the cubits 264 00:14:35,240 --> 00:14:37,440 Speaker 4: is a nine gigahertz, and if your tames that by 265 00:14:37,640 --> 00:14:41,440 Speaker 4: HBO tend to the neggave thirty four with nine, you're 266 00:14:41,480 --> 00:14:44,640 Speaker 4: at a tender the negative twenty like three or something 267 00:14:44,720 --> 00:14:48,120 Speaker 4: in energy. That's a tiny amount of energy. So you're 268 00:14:48,160 --> 00:14:51,320 Speaker 4: trying to have a tiny, tiny amount of energy to control, 269 00:14:52,120 --> 00:14:55,160 Speaker 4: and you don't want that to interact with anything, So 270 00:14:55,200 --> 00:14:57,800 Speaker 4: you have to cool them down, you have to isolate them, 271 00:14:58,080 --> 00:15:01,360 Speaker 4: and you have to make the quantum effects dominate over 272 00:15:01,400 --> 00:15:02,480 Speaker 4: the classical effects. 273 00:15:03,200 --> 00:15:06,960 Speaker 3: So practically, if I'm trying to do that right now, 274 00:15:07,080 --> 00:15:08,120 Speaker 3: how big are these machines. 275 00:15:08,480 --> 00:15:10,880 Speaker 4: So the cubits themselves are not that big. So the 276 00:15:10,960 --> 00:15:15,600 Speaker 4: cubits themselves are like a few microns. But yeah, most 277 00:15:15,640 --> 00:15:17,960 Speaker 4: of the size so you can see some of our 278 00:15:18,240 --> 00:15:20,240 Speaker 4: I got the pleasure of showing you around to one 279 00:15:20,240 --> 00:15:22,400 Speaker 4: of the machines in Yorktown. You saw that they're like 280 00:15:23,040 --> 00:15:25,920 Speaker 4: twenty foot by twenty foot in size. Most of that 281 00:15:26,480 --> 00:15:30,800 Speaker 4: is all that equipment to isolate the cubit chip, which 282 00:15:30,880 --> 00:15:33,480 Speaker 4: is only a few millimeters when you put it together 283 00:15:33,880 --> 00:15:37,360 Speaker 4: from the rest of the environment. We will, as we 284 00:15:37,440 --> 00:15:41,040 Speaker 4: get better at that, miniaturize all the isolation. But that's 285 00:15:41,240 --> 00:15:44,800 Speaker 4: cooling it down to a few milli calvin, so about 286 00:15:44,800 --> 00:15:48,560 Speaker 4: a thousand times colder than outer space. It's isolating the 287 00:15:48,760 --> 00:15:52,280 Speaker 4: noise on any electrical signal so that no noise from 288 00:15:52,320 --> 00:15:55,720 Speaker 4: the outside world gets into the system. And so that's 289 00:15:55,720 --> 00:15:59,120 Speaker 4: a lot of isolators, filters, and things like that that 290 00:15:59,200 --> 00:16:01,920 Speaker 4: we've had to invent and to allow us to make 291 00:16:01,920 --> 00:16:03,800 Speaker 4: the quantum properties of this chip go. 292 00:16:04,160 --> 00:16:06,880 Speaker 3: It's like the Princess and the pe mounds and mounds 293 00:16:06,920 --> 00:16:10,520 Speaker 3: and mounds of mattresses trying to isolate the impact of 294 00:16:10,560 --> 00:16:12,720 Speaker 3: this little thing, and that maybe. 295 00:16:12,480 --> 00:16:14,360 Speaker 4: That's the best way to describe it. Yeah, and you've 296 00:16:14,360 --> 00:16:17,080 Speaker 4: got to keep it really really prestige. 297 00:16:17,360 --> 00:16:19,240 Speaker 3: But that when you show me. So in the in 298 00:16:19,320 --> 00:16:23,040 Speaker 3: the lobby of the Watson Research Center in New Yorktown, 299 00:16:23,240 --> 00:16:26,120 Speaker 3: which by the way, is just the coolest building. It's 300 00:16:26,120 --> 00:16:30,920 Speaker 3: like a it's like a modernist it's also master piece. Anyway, 301 00:16:31,080 --> 00:16:34,600 Speaker 3: in the lobby there's there are these is it two machines. 302 00:16:34,720 --> 00:16:39,120 Speaker 4: It's it's inside a container that has three machines machines. 303 00:16:39,600 --> 00:16:41,760 Speaker 3: So what can you can you tell me what would 304 00:16:41,760 --> 00:16:43,760 Speaker 3: one of those machines cost to build right now? 305 00:16:44,360 --> 00:16:48,560 Speaker 4: So typically we put them together in a way where 306 00:16:48,640 --> 00:16:51,760 Speaker 4: we upgrade them because we want to as I as 307 00:16:51,760 --> 00:16:54,000 Speaker 4: I was talking about before, one of the things we 308 00:16:54,040 --> 00:16:57,680 Speaker 4: want to do is always get algorithms done on our machines. 309 00:16:58,360 --> 00:17:01,360 Speaker 4: And I've got a roadmap of build bigger and bigger machines. 310 00:17:01,880 --> 00:17:05,320 Speaker 4: So usually one of those quantum processes today is out 311 00:17:05,400 --> 00:17:09,480 Speaker 4: of date in six months so we want to build 312 00:17:09,520 --> 00:17:13,480 Speaker 4: this future of computing that leverages quantum computing, where every 313 00:17:13,560 --> 00:17:19,239 Speaker 4: six months we've outdated a quantum processor. Eventually, hopefully we 314 00:17:19,280 --> 00:17:22,320 Speaker 4: get to a point where it's like stable and it 315 00:17:22,320 --> 00:17:25,680 Speaker 4: can be many years operating. But we want to get 316 00:17:25,840 --> 00:17:28,960 Speaker 4: as large a quantum computer in the hands of people 317 00:17:29,040 --> 00:17:31,199 Speaker 4: to explore the math as possible to come up with 318 00:17:31,240 --> 00:17:34,119 Speaker 4: those new algorithms. So we've had a philosophy of having 319 00:17:34,160 --> 00:17:37,959 Speaker 4: them open, working with universities and things like that. So 320 00:17:37,960 --> 00:17:40,159 Speaker 4: to answer a question of costs, yes, there's cost in 321 00:17:40,240 --> 00:17:43,360 Speaker 4: building the system, but we are operating in them much 322 00:17:43,400 --> 00:17:46,360 Speaker 4: more in a service model where people pay to use 323 00:17:46,400 --> 00:17:50,199 Speaker 4: the machine because we have to continuously calibrate it and 324 00:17:50,240 --> 00:17:55,400 Speaker 4: operate it and so depending on various different things. Professors, 325 00:17:55,440 --> 00:17:57,960 Speaker 4: we have a credits program where they get free access 326 00:17:58,600 --> 00:18:02,040 Speaker 4: some universities and animal prizes. They can buy premium access 327 00:18:02,080 --> 00:18:05,040 Speaker 4: and get more access. So think of not like a 328 00:18:05,119 --> 00:18:08,199 Speaker 4: cost of it, because it's almost like a continuum. I 329 00:18:08,280 --> 00:18:11,320 Speaker 4: want to make sure that the best quantum processors that 330 00:18:11,440 --> 00:18:14,320 Speaker 4: I can build get in the hands of students and 331 00:18:14,480 --> 00:18:18,280 Speaker 4: professors and interested enterprises that want to explore these machines 332 00:18:18,320 --> 00:18:21,879 Speaker 4: as fast as possible, and typically every six months we 333 00:18:22,000 --> 00:18:22,480 Speaker 4: upgrade it. 334 00:18:22,960 --> 00:18:24,920 Speaker 3: Yeah, you don't start over, you upgrade. 335 00:18:25,080 --> 00:18:30,359 Speaker 4: We upgrade various different pieces, the processor, the electronics. Some 336 00:18:30,560 --> 00:18:35,200 Speaker 4: upgrades are just simply replaced the processor. But as an example, 337 00:18:35,520 --> 00:18:38,080 Speaker 4: I think many people have probably seen photos of quantum 338 00:18:38,119 --> 00:18:41,040 Speaker 4: computers and you see this scary thing with all these 339 00:18:41,040 --> 00:18:44,520 Speaker 4: wires hanging down, as I've referred to as the chandelier, 340 00:18:44,560 --> 00:18:47,040 Speaker 4: and it's got all these wires with loops and things 341 00:18:47,080 --> 00:18:50,440 Speaker 4: like that. They're called co x keebles. When we first 342 00:18:50,440 --> 00:18:53,000 Speaker 4: put the quantum computer on the cloud in twenty sixteen, 343 00:18:53,440 --> 00:18:58,160 Speaker 4: you could probably only fit about fifty cubits inside one cryostat. 344 00:18:58,320 --> 00:19:01,040 Speaker 4: We've had to upgrade all those cables so that we 345 00:19:01,080 --> 00:19:03,399 Speaker 4: can fit around one thousand. I want to get to 346 00:19:03,440 --> 00:19:07,280 Speaker 4: three thousand, and that's about miniaturizing it. So to answer 347 00:19:07,280 --> 00:19:09,680 Speaker 4: your question, an upgrade, it depends. It can be either 348 00:19:09,800 --> 00:19:12,879 Speaker 4: just the processor or it can be the complete insides. 349 00:19:13,400 --> 00:19:16,520 Speaker 4: And we're actually in our third generation of our electronics 350 00:19:16,560 --> 00:19:21,520 Speaker 4: to control the systems, to make them faster, less noise. Internally, 351 00:19:21,560 --> 00:19:25,960 Speaker 4: we've got exciting results of going to something like cold cryocemos, 352 00:19:26,280 --> 00:19:29,119 Speaker 4: so you can bring down the cost in terms of 353 00:19:29,240 --> 00:19:33,639 Speaker 4: energy of running these quantum computers almost negligible, and you 354 00:19:33,640 --> 00:19:37,240 Speaker 4: could imagine future quantum computers. I'm not going to require 355 00:19:37,320 --> 00:19:41,240 Speaker 4: much energy to run, so unlike classical compute that requires 356 00:19:41,240 --> 00:19:44,159 Speaker 4: lots of energy. The biggest machines that we envision is 357 00:19:44,160 --> 00:19:47,080 Speaker 4: only in the few megawatts. But we have to upgrade 358 00:19:47,119 --> 00:19:51,440 Speaker 4: to future controls that use less energy. So it depends. 359 00:19:51,600 --> 00:19:56,080 Speaker 4: It's my long answer, short answer to how it upgrades, 360 00:19:56,119 --> 00:19:57,239 Speaker 4: and it depends on what it is. 361 00:19:57,520 --> 00:20:00,639 Speaker 3: The only observation that I felt I could with making 362 00:20:00,680 --> 00:20:04,159 Speaker 3: when you showed me the quantum machine is it's gorgeous. 363 00:20:04,880 --> 00:20:05,600 Speaker 3: I look at art. 364 00:20:05,920 --> 00:20:09,399 Speaker 4: I've always believed that, and I think that there's an 365 00:20:09,440 --> 00:20:12,560 Speaker 4: IBM saying good design is good business. But we've always 366 00:20:12,880 --> 00:20:17,280 Speaker 4: taken pride in making sure what we build. I don't know. 367 00:20:17,400 --> 00:20:20,080 Speaker 4: I feel if you're going to build something that is 368 00:20:20,600 --> 00:20:24,199 Speaker 4: new that can change, you should take the time to 369 00:20:24,359 --> 00:20:26,080 Speaker 4: make sure it looks and feels good. 370 00:20:26,240 --> 00:20:28,960 Speaker 3: Will you donated to MoMA when you're through with that 371 00:20:29,040 --> 00:20:33,160 Speaker 3: particular Actually, I think we just put. 372 00:20:32,880 --> 00:20:36,480 Speaker 4: An old version of one of our insights with the 373 00:20:36,560 --> 00:20:40,840 Speaker 4: United Airlines and the AAPS, which is the American Physical Society, 374 00:20:40,880 --> 00:20:43,960 Speaker 4: and the University of Chicago. There's a replica right now. 375 00:20:44,080 --> 00:20:46,640 Speaker 4: If you fly into one of the terminals in Chicago, 376 00:20:47,160 --> 00:20:48,320 Speaker 4: you can walk and see one. 377 00:20:48,600 --> 00:20:51,199 Speaker 3: Oh really yeah, well the most advanced thing at all air. 378 00:20:51,280 --> 00:20:54,720 Speaker 4: I'm sure probably, but yeah, I hopefully, I think, yeah, 379 00:20:54,720 --> 00:20:57,919 Speaker 4: we're open to that. But yeah, I appreciate that you 380 00:20:57,960 --> 00:20:59,720 Speaker 4: love the design. It was beautiful. 381 00:21:00,200 --> 00:21:03,080 Speaker 3: So I last week I interviewed for another episode of 382 00:21:03,119 --> 00:21:08,400 Speaker 3: Smart TALX your CEO, Ivin Kushner, And when we got 383 00:21:08,440 --> 00:21:12,920 Speaker 3: to the quantum question, I mean, he's always alliant and brilliant, 384 00:21:13,760 --> 00:21:17,320 Speaker 3: but quantum, he's like lit up. I mean right in 385 00:21:17,400 --> 00:21:22,080 Speaker 3: thinking that IBM is much more invested in quantum than 386 00:21:22,080 --> 00:21:24,880 Speaker 3: anybody else. Is that a fair statement? Oh yeah, most definitely. 387 00:21:25,040 --> 00:21:28,359 Speaker 3: Why Why did IBM choose to kind of make this 388 00:21:28,440 --> 00:21:29,200 Speaker 3: such a priority. 389 00:21:29,560 --> 00:21:32,679 Speaker 4: So when I took to the history of the physics side, 390 00:21:33,480 --> 00:21:36,440 Speaker 4: there's this interesting thing in the history of computing. So 391 00:21:36,600 --> 00:21:41,240 Speaker 4: we build computer classical computers today using bits and sea moss, 392 00:21:41,240 --> 00:21:44,040 Speaker 4: and they consume energy. Do you know that there is 393 00:21:44,080 --> 00:21:47,960 Speaker 4: a way in classical where you can actually compute without 394 00:21:48,040 --> 00:21:51,480 Speaker 4: using energy. It's called reversal computing. Turns out to be 395 00:21:51,720 --> 00:21:56,400 Speaker 4: a terrible idea. It's not practical to build. But IBM 396 00:21:56,680 --> 00:22:00,760 Speaker 4: investigated that with Ralph Laura and Charlie ban It early 397 00:22:00,840 --> 00:22:04,600 Speaker 4: on and they proved the concept that reversible computing. The 398 00:22:04,640 --> 00:22:08,280 Speaker 4: first use of quantum information theory, one of the first 399 00:22:08,480 --> 00:22:12,239 Speaker 4: actually was from IBM. When I did my PhD, I 400 00:22:12,280 --> 00:22:16,199 Speaker 4: remember actually picking up this paper on quantum teleportation and 401 00:22:16,400 --> 00:22:18,840 Speaker 4: seeing IBM written there, and at the time I remember 402 00:22:18,840 --> 00:22:21,239 Speaker 4: thinking that they make PCs. Well, what the hell are 403 00:22:21,240 --> 00:22:26,399 Speaker 4: they doing this foundational paper on quantum teleportation? Why are 404 00:22:26,400 --> 00:22:29,240 Speaker 4: they doing it? So to answer your question, actually, IBM 405 00:22:29,400 --> 00:22:33,520 Speaker 4: was the first in quantum information science because it's the 406 00:22:33,600 --> 00:22:37,920 Speaker 4: fundamental of computation. Can we actually come up with compute 407 00:22:38,400 --> 00:22:42,200 Speaker 4: that we can go beyond the classical So way before 408 00:22:42,240 --> 00:22:45,719 Speaker 4: anyone was talking about it, they were doing fundamental theory. 409 00:22:46,359 --> 00:22:49,119 Speaker 4: And then as we've built it, we've always When I 410 00:22:49,160 --> 00:22:52,520 Speaker 4: first came there, the experimental team was small. In twenty eleven, 411 00:22:53,040 --> 00:22:56,960 Speaker 4: we've had a small team that we're focusing on single 412 00:22:57,040 --> 00:23:01,359 Speaker 4: cubitts coupling them. I think in two thy twelve was 413 00:23:01,359 --> 00:23:04,919 Speaker 4: the first time we showed really good two Cuba gates 414 00:23:05,720 --> 00:23:08,919 Speaker 4: and no one was talking about quantum computing then. And 415 00:23:08,960 --> 00:23:12,720 Speaker 4: then I remember in about twenty sixteen, I said to 416 00:23:13,080 --> 00:23:16,560 Speaker 4: actually Arvin was the director of research, then can we 417 00:23:16,680 --> 00:23:20,000 Speaker 4: actually put our quantum computer on the cloud? Well that's 418 00:23:20,040 --> 00:23:24,760 Speaker 4: probably twenty fifteen, and it was always supporting that. So 419 00:23:24,840 --> 00:23:27,280 Speaker 4: as we've done more and more we've been able to 420 00:23:27,320 --> 00:23:31,200 Speaker 4: do it. It's had this program going now, I agree 421 00:23:31,560 --> 00:23:34,840 Speaker 4: is very visible, like because we're in this scaling phase 422 00:23:35,240 --> 00:23:38,840 Speaker 4: and so we're invested to keep scaling it and to 423 00:23:38,840 --> 00:23:43,320 Speaker 4: get why is At IBM research, what we always do 424 00:23:43,560 --> 00:23:46,560 Speaker 4: is answer what is the future of computing? Whether it's 425 00:23:46,560 --> 00:23:50,520 Speaker 4: coming up with new algorithms, coming up with better AI, 426 00:23:50,840 --> 00:23:53,880 Speaker 4: coming up with quantum, or coming up with just how 427 00:23:53,920 --> 00:23:57,240 Speaker 4: do different accelerators go together. It's our DNA to answer 428 00:23:57,280 --> 00:23:58,960 Speaker 4: the question of what is the future? 429 00:23:59,119 --> 00:24:01,439 Speaker 3: Need a perfect problem for IBM because you kind of 430 00:24:01,480 --> 00:24:05,160 Speaker 3: need to have a legacy of building stuff, building actual 431 00:24:06,280 --> 00:24:07,320 Speaker 3: physical machines. 432 00:24:07,920 --> 00:24:11,840 Speaker 4: Yeah, that's why I came to IBM. I wanted the experience, 433 00:24:12,520 --> 00:24:17,080 Speaker 4: the culture of building hard things that others have not 434 00:24:17,160 --> 00:24:17,800 Speaker 4: done before. 435 00:24:19,119 --> 00:24:21,520 Speaker 3: Where do you imagine we are in the timeline of 436 00:24:21,520 --> 00:24:25,720 Speaker 3: this technology? It will come a point when it will mature. 437 00:24:26,800 --> 00:24:29,480 Speaker 3: My cell phone is a mature technology at this point. 438 00:24:29,720 --> 00:24:32,119 Speaker 3: How far are we from that point? With condom. 439 00:24:32,560 --> 00:24:35,160 Speaker 4: So I think there's various aspects of it. So we 440 00:24:35,200 --> 00:24:38,639 Speaker 4: sat in twenty and seventy we set our goal that 441 00:24:38,720 --> 00:24:41,480 Speaker 4: in twenty twenty three we would be able to build 442 00:24:41,480 --> 00:24:45,960 Speaker 4: a machine that was beyond classical computers to simulate it. 443 00:24:46,600 --> 00:24:49,520 Speaker 4: And we achieved that in twenty twenty three. So to 444 00:24:50,560 --> 00:24:52,480 Speaker 4: run a biggo we call it a quantum circuit. The 445 00:24:52,480 --> 00:24:54,600 Speaker 4: details of it don't matter, but to run a quantum 446 00:24:54,600 --> 00:24:58,800 Speaker 4: workload that if you were to simulate that workload how 447 00:24:58,840 --> 00:25:01,480 Speaker 4: a quantum computer operates, it's on a classical computer, you 448 00:25:01,480 --> 00:25:04,160 Speaker 4: couldn't do it. So we said that as our first 449 00:25:04,520 --> 00:25:07,440 Speaker 4: and now I've made it publicly that by twenty twenty 450 00:25:07,520 --> 00:25:11,280 Speaker 4: nine we'll build the first fault tolerant quantum computer. That is, 451 00:25:11,320 --> 00:25:15,960 Speaker 4: one that can completely handle the noise to the level 452 00:25:16,040 --> 00:25:19,720 Speaker 4: to allow you to run a very very large, large problem. 453 00:25:19,840 --> 00:25:22,240 Speaker 3: So an example of a large problem. 454 00:25:21,920 --> 00:25:25,480 Speaker 4: Yeah, a large quantum problem. So for around a couple 455 00:25:25,480 --> 00:25:29,080 Speaker 4: of one hundred cubits and one hundred million operations, you're 456 00:25:29,119 --> 00:25:34,640 Speaker 4: talking still interesting science problems like simulating a molecule, or 457 00:25:35,119 --> 00:25:42,520 Speaker 4: calculating a small optimization problem, or calculating, say some part 458 00:25:42,800 --> 00:25:45,600 Speaker 4: of a matrix update. In some type of differential so 459 00:25:45,600 --> 00:25:48,600 Speaker 4: it'll still be scientific, but it'll be at the point 460 00:25:48,680 --> 00:25:54,480 Speaker 4: where it's beyond, well beyond any classical approximate method. And 461 00:25:54,520 --> 00:25:57,440 Speaker 4: then I think that's twenty twenty nine. That's twenty twenty nine, 462 00:25:57,680 --> 00:25:58,960 Speaker 4: So we're four. 463 00:25:58,840 --> 00:26:01,600 Speaker 3: Years away from something that can start to handle. 464 00:26:01,920 --> 00:26:05,840 Speaker 4: Interesting problem, serious problems. I do believe the scientists will 465 00:26:05,840 --> 00:26:09,280 Speaker 4: find interesting heuristic problems before that, And so over the 466 00:26:09,359 --> 00:26:11,800 Speaker 4: next four years, you're going to continue to see more 467 00:26:11,880 --> 00:26:16,400 Speaker 4: and more let's call them heuristic not provable quantum problems 468 00:26:16,400 --> 00:26:19,440 Speaker 4: that run on quantum computers that come out. We're see 469 00:26:19,560 --> 00:26:22,359 Speaker 4: more and more come from many of our partners and ourselves. 470 00:26:22,640 --> 00:26:25,440 Speaker 4: Heuristic problems have value, but they have to be tested, 471 00:26:25,520 --> 00:26:27,240 Speaker 4: they have to stand up over time. You have to 472 00:26:27,320 --> 00:26:29,719 Speaker 4: run them many, many times, you have to try different ones, 473 00:26:30,160 --> 00:26:33,240 Speaker 4: and many times heuristic can lead to formal problems. So 474 00:26:33,280 --> 00:26:35,399 Speaker 4: you're going to see, because we're beyond now the point 475 00:26:35,440 --> 00:26:39,760 Speaker 4: that you can simulate these quantum computers with any classical computer. 476 00:26:39,880 --> 00:26:43,360 Speaker 4: They're kind of like a scientific tool. So they're exploring 477 00:26:43,440 --> 00:26:44,000 Speaker 4: the heuristic. 478 00:26:44,240 --> 00:26:45,840 Speaker 3: What do you have to get done between now and 479 00:26:45,880 --> 00:26:47,240 Speaker 3: twenty twenty nine to get there? 480 00:26:47,640 --> 00:26:49,760 Speaker 4: So we had to reinvent how we wanted to do 481 00:26:49,920 --> 00:26:53,879 Speaker 4: error correction. So we have to demonstrate modules and if 482 00:26:53,920 --> 00:26:57,600 Speaker 4: we can demonstrate these error corrected module and our goal 483 00:26:57,720 --> 00:26:59,960 Speaker 4: is actually it's called Kooko, but I name all that 484 00:27:00,080 --> 00:27:02,840 Speaker 4: chips after birds, so it's called Kuoko. Borrow is named 485 00:27:02,840 --> 00:27:05,480 Speaker 4: after an Australian vert. I think I still say Cuoko 486 00:27:05,520 --> 00:27:10,000 Speaker 4: Borrow the way Australians do. We need to then show 487 00:27:10,040 --> 00:27:12,159 Speaker 4: that we can make a single module and then we 488 00:27:12,200 --> 00:27:14,560 Speaker 4: want to connect two of those modules together, and I 489 00:27:14,640 --> 00:27:18,399 Speaker 4: call that one Cockatoo, which is another Australian vert. And 490 00:27:18,440 --> 00:27:21,160 Speaker 4: then if we can do that, so that's twenty six 491 00:27:21,680 --> 00:27:24,280 Speaker 4: and twenty seven, and then we want to scale them 492 00:27:24,480 --> 00:27:27,479 Speaker 4: scale those modules, and that we call Starling, and we 493 00:27:27,480 --> 00:27:30,240 Speaker 4: want to scale that in twenty twenty nine. So get 494 00:27:30,280 --> 00:27:33,959 Speaker 4: a module, join two modules together, and scale and so 495 00:27:34,040 --> 00:27:36,440 Speaker 4: each module is going to be around one thousand cubits. 496 00:27:37,480 --> 00:27:40,000 Speaker 3: The challenge to getting there is it finding the right 497 00:27:40,240 --> 00:27:43,760 Speaker 3: material or how would you describe what that's. 498 00:27:43,600 --> 00:27:46,159 Speaker 4: The beauty to be done. That's the beauty of it 499 00:27:46,200 --> 00:27:49,440 Speaker 4: is if we would have been here two years ago, 500 00:27:50,240 --> 00:27:52,919 Speaker 4: I couldn't tell you how it would be done. So 501 00:27:53,000 --> 00:27:55,520 Speaker 4: we had a huge breakthrough we came up with a 502 00:27:55,520 --> 00:27:59,760 Speaker 4: new code, a new quantumeric Russian code, and that code. 503 00:28:00,320 --> 00:28:02,760 Speaker 4: The biggest in part of that code that is the 504 00:28:02,800 --> 00:28:07,080 Speaker 4: most important is it's modular in nature. So previous codes, 505 00:28:07,280 --> 00:28:11,199 Speaker 4: without getting too technical, they were very monolithic and you 506 00:28:11,240 --> 00:28:13,320 Speaker 4: had to build a very big device and I wouldn't 507 00:28:13,320 --> 00:28:16,640 Speaker 4: have known we would have to invent tools like new 508 00:28:16,840 --> 00:28:20,119 Speaker 4: simos tools to do that. So we came up with 509 00:28:20,160 --> 00:28:23,800 Speaker 4: this new code. We started on twenty nineteen, we published 510 00:28:23,840 --> 00:28:25,720 Speaker 4: in twenty twenty four. We kind of had most of 511 00:28:25,800 --> 00:28:28,159 Speaker 4: things worked out in twenty twenty three. That's why we 512 00:28:28,200 --> 00:28:31,440 Speaker 4: got confident to release the thing. So the biggest breakthrough 513 00:28:31,480 --> 00:28:33,800 Speaker 4: we had is coming up with a code that's modular 514 00:28:33,840 --> 00:28:36,960 Speaker 4: in nature. And think of that as a like a blueprint. 515 00:28:37,280 --> 00:28:40,280 Speaker 4: And so now we have the blueprint and now we're 516 00:28:40,320 --> 00:28:45,000 Speaker 4: doing engineering tasks to implement every part of that blueprint. 517 00:28:45,200 --> 00:28:48,360 Speaker 3: And so the minute you had that breakthrough, then you 518 00:28:48,440 --> 00:28:51,240 Speaker 3: began to have confidence at something exactly these goals could 519 00:28:51,240 --> 00:28:51,680 Speaker 3: be met. 520 00:28:51,800 --> 00:28:55,520 Speaker 4: And then you can't. And then anyone that's done engineering 521 00:28:55,800 --> 00:28:57,480 Speaker 4: will know what I'm talking about when I say this 522 00:28:57,720 --> 00:29:01,920 Speaker 4: is cycles are learning. It takes so long from test 523 00:29:01,960 --> 00:29:06,040 Speaker 4: idea to build two tests in hardware, the cycles of 524 00:29:06,160 --> 00:29:08,480 Speaker 4: learning are much much lower than software, Like you can 525 00:29:08,520 --> 00:29:11,760 Speaker 4: be really, really faster in the software. So then we've 526 00:29:11,840 --> 00:29:15,520 Speaker 4: planned out our iterations over the next few years, and 527 00:29:15,600 --> 00:29:19,800 Speaker 4: so we have to successfully demonstrate them. I may slip 528 00:29:19,840 --> 00:29:24,600 Speaker 4: because sometimes you may estimate your time wrong, but we 529 00:29:24,760 --> 00:29:27,200 Speaker 4: now have exactly what we want to do for the 530 00:29:27,200 --> 00:29:27,960 Speaker 4: next four years. 531 00:29:28,080 --> 00:29:29,640 Speaker 3: I want to go back to that breakthrough for a moment. 532 00:29:29,800 --> 00:29:32,479 Speaker 3: What does the word breaks we mean in that context, Like, 533 00:29:32,760 --> 00:29:34,840 Speaker 3: it's not that you get a call in the morning 534 00:29:35,280 --> 00:29:36,720 Speaker 3: from somebody who says. 535 00:29:36,400 --> 00:29:36,840 Speaker 2: I did it? 536 00:29:37,360 --> 00:29:39,120 Speaker 3: Do you see it coming? Or is it a surprise 537 00:29:39,160 --> 00:29:39,960 Speaker 3: when they get there. 538 00:29:40,160 --> 00:29:43,800 Speaker 4: So the way this one worked is Sogo Brave, who's 539 00:29:44,320 --> 00:29:47,280 Speaker 4: an algorithm person at IBM, one of the smartest and 540 00:29:47,400 --> 00:29:48,200 Speaker 4: quantum information. 541 00:29:48,880 --> 00:29:52,480 Speaker 3: Don't mention his name to everyone. You'll come for him. 542 00:29:52,520 --> 00:29:55,400 Speaker 4: Everyone in quantum already knows his name. I don't think 543 00:29:55,400 --> 00:29:59,280 Speaker 4: there's an idea that has not originated from him in quantum. 544 00:30:00,600 --> 00:30:04,120 Speaker 4: So we're looking at other codes and we'll go all right, 545 00:30:05,120 --> 00:30:08,880 Speaker 4: we've got to get serious about these codes. And others 546 00:30:09,000 --> 00:30:11,600 Speaker 4: were starting to propose to bring these and then we 547 00:30:11,600 --> 00:30:16,400 Speaker 4: call them LDPC codes from the classical space into the quantum, 548 00:30:16,880 --> 00:30:19,600 Speaker 4: and I asked him, we need to get ahead of 549 00:30:19,640 --> 00:30:22,320 Speaker 4: this and understand what they're doing it. He's like, the 550 00:30:22,360 --> 00:30:26,120 Speaker 4: most modest perfuse late, Jay, let me learn about them 551 00:30:26,200 --> 00:30:29,280 Speaker 4: and I'll generate a report for us and we'll read 552 00:30:29,320 --> 00:30:32,360 Speaker 4: through it. And then I said great. Then I don't know. 553 00:30:32,400 --> 00:30:34,800 Speaker 4: Six months later, he comes back with one hundred page 554 00:30:34,840 --> 00:30:38,800 Speaker 4: report on everyone everyone had done in LTPC codes. I'm like, awesome. 555 00:30:38,840 --> 00:30:42,320 Speaker 4: So I started then to read from them. And then 556 00:30:42,440 --> 00:30:45,440 Speaker 4: we said, all right, how do we under the assumptions 557 00:30:45,480 --> 00:30:48,400 Speaker 4: of the hardware we can build? Can we get an 558 00:30:48,560 --> 00:30:54,280 Speaker 4: LTPC code knowing what we can build? And that's a 559 00:30:54,320 --> 00:30:57,200 Speaker 4: great question, and so we put a small team together 560 00:30:57,600 --> 00:31:01,320 Speaker 4: to investigate and honestly took two to three years, and 561 00:31:01,720 --> 00:31:05,880 Speaker 4: we iterated and we used the constraints, so we had 562 00:31:05,920 --> 00:31:08,720 Speaker 4: the sort of theory, and then we had the constraints 563 00:31:08,760 --> 00:31:11,480 Speaker 4: of what we could build. And we iterated for a 564 00:31:11,520 --> 00:31:13,960 Speaker 4: few years, and then at the end of that we 565 00:31:14,000 --> 00:31:16,920 Speaker 4: came out with a solution that yes, it is possible 566 00:31:16,960 --> 00:31:20,080 Speaker 4: to meet all the constraints of the hardware and build 567 00:31:20,120 --> 00:31:21,400 Speaker 4: a code that will work. 568 00:31:22,000 --> 00:31:26,240 Speaker 3: I'm just curious about So you had this task, this problem, 569 00:31:26,280 --> 00:31:29,440 Speaker 3: you want to solve. And when you set out on 570 00:31:29,480 --> 00:31:31,760 Speaker 3: the task of trying to solve the problem, what's your 571 00:31:31,840 --> 00:31:34,440 Speaker 3: certainty level that you'll get a solution. 572 00:31:34,880 --> 00:31:38,200 Speaker 4: Well, that's the beauty of science for things where you 573 00:31:38,360 --> 00:31:42,200 Speaker 4: kind of have a few ideas. My philosophy is try 574 00:31:42,240 --> 00:31:44,800 Speaker 4: a few for the ones that need to be in 575 00:31:44,880 --> 00:31:49,760 Speaker 4: that like wow moment. It's honestly, you've got to set 576 00:31:49,800 --> 00:31:53,160 Speaker 4: the ambition really, really high, but know when to stop. 577 00:31:54,080 --> 00:31:56,120 Speaker 4: It was a great team that went together to get 578 00:31:56,160 --> 00:31:59,120 Speaker 4: that breakthrough, and we knew that we needed to come 579 00:31:59,200 --> 00:32:03,240 Speaker 4: up with a code that met the requirements of the experiment. 580 00:32:03,800 --> 00:32:07,480 Speaker 4: And I think what was different before then is the 581 00:32:07,480 --> 00:32:12,240 Speaker 4: theorists that were doing error correction codes didn't necessarily know 582 00:32:12,480 --> 00:32:16,000 Speaker 4: the constraints of experiments, so it was like really more 583 00:32:16,040 --> 00:32:18,520 Speaker 4: pen and paper. So this became one, all right, given 584 00:32:18,520 --> 00:32:22,280 Speaker 4: these sets of constraints, is it possible when LA's. 585 00:32:22,200 --> 00:32:25,280 Speaker 3: Questions about this? Sorry, I love these kind of moments 586 00:32:25,280 --> 00:32:28,920 Speaker 3: when things become clear. At the time the problem was solved, 587 00:32:29,480 --> 00:32:32,120 Speaker 3: were you aware of the implications of the solution or 588 00:32:32,600 --> 00:32:34,720 Speaker 3: did that takes you knew exactly what. 589 00:32:35,480 --> 00:32:39,120 Speaker 4: We set out exactly like either we were going to 590 00:32:39,200 --> 00:32:41,400 Speaker 4: have to work out how to cool down a very 591 00:32:41,480 --> 00:32:44,880 Speaker 4: large piece of silicon, which would require a lot of 592 00:32:44,920 --> 00:32:48,280 Speaker 4: engineering and building tools beyond what anyone has ever built 593 00:32:48,280 --> 00:32:53,120 Speaker 4: in the silicon semoss industry to implement the known codes, 594 00:32:53,760 --> 00:32:56,000 Speaker 4: or we had to come up with a different one, 595 00:32:56,240 --> 00:32:58,800 Speaker 4: and once I knew that we had one, that I 596 00:32:59,000 --> 00:33:03,120 Speaker 4: didn't need to re invent any tools to build. The 597 00:33:03,160 --> 00:33:04,680 Speaker 4: implications are clear how. 598 00:33:04,640 --> 00:33:07,719 Speaker 3: Much time elapsed between the time you heard the problem 599 00:33:07,800 --> 00:33:11,000 Speaker 3: was solved and the time you told Arvin Krishna, the CEO, 600 00:33:11,120 --> 00:33:12,360 Speaker 3: the problem was solved. 601 00:33:13,360 --> 00:33:15,840 Speaker 4: I'm sure the next time I spoke to him, I update, 602 00:33:15,880 --> 00:33:18,800 Speaker 4: but I don't remember. The beauty of Avin is he 603 00:33:18,840 --> 00:33:21,440 Speaker 4: trusts the scientists will do it, and so he doesn't 604 00:33:21,440 --> 00:33:24,280 Speaker 4: really check on us. We update him when it is, 605 00:33:24,320 --> 00:33:27,440 Speaker 4: and he he empowers us to do really hard problems. 606 00:33:27,640 --> 00:33:31,360 Speaker 3: Yeah, so let's talk about uses. I mean, they're really 607 00:33:31,440 --> 00:33:35,280 Speaker 3: like cool, big shiny machine. I think you'll get pay 608 00:33:35,320 --> 00:33:38,560 Speaker 3: twenty twenty nine. But there's all kinds of really interesting 609 00:33:38,600 --> 00:33:40,040 Speaker 3: problems you're already working on. 610 00:33:40,440 --> 00:33:45,680 Speaker 4: Yes, this is like another interesting area is I can 611 00:33:45,840 --> 00:33:49,400 Speaker 4: prove in pen and paper algorithms that we want to 612 00:33:49,480 --> 00:33:51,719 Speaker 4: run that. Like, it's not that we don't know what 613 00:33:51,760 --> 00:33:54,720 Speaker 4: to do with a quantum computer, there are hundreds of algorithms. 614 00:33:54,720 --> 00:33:57,280 Speaker 4: So you can go to I think it's called quantumzoo 615 00:33:57,320 --> 00:34:00,360 Speaker 4: dot com and you can see many many algorithms. People 616 00:34:00,360 --> 00:34:01,920 Speaker 4: are coming up with more of more of them that 617 00:34:02,000 --> 00:34:05,680 Speaker 4: they prove by pen and paper. Imagine, now we have 618 00:34:05,800 --> 00:34:10,760 Speaker 4: a machine that you can't simulate. How do you actually 619 00:34:11,560 --> 00:34:16,160 Speaker 4: discover algorithms in a scientific way? How do you look 620 00:34:16,239 --> 00:34:19,840 Speaker 4: and discover algorithms using a quantum computer. We're in this 621 00:34:20,000 --> 00:34:24,319 Speaker 4: exciting period right now, and so even though I can 622 00:34:24,400 --> 00:34:26,640 Speaker 4: prove these ones that we can run in the future, 623 00:34:27,160 --> 00:34:31,000 Speaker 4: there's a big white space between what the machines we 624 00:34:31,080 --> 00:34:33,080 Speaker 4: have and we're going to build and continue to do 625 00:34:33,480 --> 00:34:38,120 Speaker 4: and those ones that want the provable ones. And I'm 626 00:34:38,120 --> 00:34:43,480 Speaker 4: an optimistic person by nature. I think getting those machines 627 00:34:43,520 --> 00:34:45,920 Speaker 4: in the hands of students to explore and look at 628 00:34:45,920 --> 00:34:50,080 Speaker 4: heuristic algorithms, so looking at the equivalent of doing numerical 629 00:34:50,160 --> 00:34:55,959 Speaker 4: algorithms on computers, which there's many histories of numerical algorithms 630 00:34:56,040 --> 00:34:59,880 Speaker 4: being discovered on classical computers before we had formal pre 631 00:35:00,600 --> 00:35:03,520 Speaker 4: that we rely on today. People would even argue the 632 00:35:03,560 --> 00:35:07,200 Speaker 4: way AI works was driven numerically, even though we have 633 00:35:07,560 --> 00:35:11,520 Speaker 4: input into it. There are ones in optimization driven numerically. 634 00:35:12,040 --> 00:35:15,879 Speaker 4: We are entering that phase. So the computer scientists now 635 00:35:16,480 --> 00:35:20,160 Speaker 4: need to go play with these primitives. Our prediction is 636 00:35:20,880 --> 00:35:23,600 Speaker 4: over the next couple of years we're going to see 637 00:35:24,000 --> 00:35:28,799 Speaker 4: valuable numerical equivalent algorithms emerge. And where the scientists are 638 00:35:28,840 --> 00:35:32,719 Speaker 4: going is in four categories. One is simulating nature, so 639 00:35:32,840 --> 00:35:38,000 Speaker 4: looking at either Haanji physics, chemistry, light problems. As an example, 640 00:35:38,120 --> 00:35:41,279 Speaker 4: with our partners in Japan, they took one of our 641 00:35:41,320 --> 00:35:47,160 Speaker 4: quantum computers and for Gackle, a very large classical supercomputer, 642 00:35:47,520 --> 00:35:50,400 Speaker 4: and they ran a problem where quantum was just a 643 00:35:50,440 --> 00:35:53,640 Speaker 4: sub routine of the problem that was running on all 644 00:35:53,640 --> 00:35:55,480 Speaker 4: of for garcule, and they were able to look at 645 00:35:55,520 --> 00:35:58,440 Speaker 4: an interesting molecule, a molecule that if you would go 646 00:35:58,480 --> 00:36:00,319 Speaker 4: by pen and paper you would have said to take 647 00:36:00,400 --> 00:36:02,560 Speaker 4: me a very long time to run that. They were 648 00:36:02,600 --> 00:36:06,160 Speaker 4: able to run that quite accurately, heuristically, and already get 649 00:36:06,280 --> 00:36:09,319 Speaker 4: results that are comparable with the best classical methods. So 650 00:36:09,360 --> 00:36:11,719 Speaker 4: they are extremely excited because they want to push that 651 00:36:11,800 --> 00:36:13,920 Speaker 4: further and they're sort of showing that you can take 652 00:36:13,960 --> 00:36:18,120 Speaker 4: a classical supercomputer with quantum as a subroutine and start 653 00:36:18,160 --> 00:36:19,520 Speaker 4: to push the level they were. 654 00:36:19,920 --> 00:36:22,240 Speaker 3: This was trying to solve a medical problem. 655 00:36:22,280 --> 00:36:26,640 Speaker 4: Is this one is a like most people don't realize, 656 00:36:26,680 --> 00:36:29,560 Speaker 4: Like iron sulfur, just something as simple as iron and sulfur, 657 00:36:29,920 --> 00:36:34,520 Speaker 4: we can't solve that exactly, Like iron sulfur, molecules are 658 00:36:34,560 --> 00:36:38,879 Speaker 4: too hard, so really small small molecules are really really hard, 659 00:36:38,960 --> 00:36:41,799 Speaker 4: too hard for classical computers to solve. People think we 660 00:36:41,840 --> 00:36:44,080 Speaker 4: can solve a lot of things. It actually turns out 661 00:36:44,080 --> 00:36:45,240 Speaker 4: we can't solve very much. 662 00:36:45,480 --> 00:36:48,359 Speaker 3: You say solve its instance, you know precisely how that 663 00:36:48,400 --> 00:36:50,520 Speaker 3: molecule works and it's constructed. 664 00:36:50,520 --> 00:36:54,600 Speaker 4: No precisely what the energy levels of that molecule is 665 00:36:54,680 --> 00:36:57,440 Speaker 4: and how they come together, and then be able to 666 00:36:57,440 --> 00:36:59,799 Speaker 4: do that on a classical computer and compare it talk. 667 00:37:00,360 --> 00:37:04,240 Speaker 3: It would be really really useful to know that specifically, because. 668 00:37:04,280 --> 00:37:07,320 Speaker 4: If you can have energy levels, then you can estimate 669 00:37:07,600 --> 00:37:10,839 Speaker 4: reaction rates. If you can estimate reaction rates, you can 670 00:37:10,840 --> 00:37:14,680 Speaker 4: see how different types of chemicals will react. That can 671 00:37:14,719 --> 00:37:18,400 Speaker 4: then lead to better informing eventually how to build materials 672 00:37:18,760 --> 00:37:21,080 Speaker 4: or even drug design. I just want to be careful 673 00:37:21,080 --> 00:37:23,160 Speaker 4: and not say, oh, we're going to solve drug design 674 00:37:23,200 --> 00:37:28,319 Speaker 4: or that because there's many scientific steps to make that so. 675 00:37:28,960 --> 00:37:31,200 Speaker 4: And so what quantum gives you as a different tool 676 00:37:31,239 --> 00:37:34,480 Speaker 4: to give you more accuracy and then lead to making 677 00:37:34,560 --> 00:37:35,759 Speaker 4: the different methods work. 678 00:37:36,360 --> 00:37:41,200 Speaker 3: You can subcontract out aspects of a problem quantum right now, 679 00:37:41,239 --> 00:37:44,200 Speaker 3: and that just gets you further along than you would 680 00:37:44,200 --> 00:37:44,440 Speaker 3: have been. 681 00:37:44,719 --> 00:37:48,440 Speaker 4: So at the moment, even this result still does not 682 00:37:48,560 --> 00:37:53,600 Speaker 4: beat the best approximate classical method. It's comparable. So the 683 00:37:53,880 --> 00:37:57,319 Speaker 4: art of chemistry for the last hundred years has been 684 00:37:57,360 --> 00:38:01,239 Speaker 4: about approximating. So what we've done is we have got 685 00:38:01,680 --> 00:38:05,760 Speaker 4: very very good at coming up with ways of approximating nature. 686 00:38:06,680 --> 00:38:09,000 Speaker 4: And a lot of the things that we do and 687 00:38:09,040 --> 00:38:12,560 Speaker 4: we exploit and we use to estimate approximations. They don't 688 00:38:12,560 --> 00:38:15,720 Speaker 4: a stimulate nature of the way nature is. They approximate it. 689 00:38:15,760 --> 00:38:19,960 Speaker 4: And there's I could list many different acronyms of different 690 00:38:20,000 --> 00:38:24,600 Speaker 4: methods that go into approximating nature. What quantum gives us 691 00:38:24,680 --> 00:38:28,680 Speaker 4: is to eventually get beyond that approximation and do it 692 00:38:28,760 --> 00:38:32,080 Speaker 4: the way nature works. And so we aren't beating those 693 00:38:32,120 --> 00:38:35,000 Speaker 4: approximation methods, and this is why I think, this is 694 00:38:35,000 --> 00:38:37,120 Speaker 4: why it's still in the science. But they're getting comparable, 695 00:38:37,480 --> 00:38:40,800 Speaker 4: getting comparable with a new tool where the previous tool 696 00:38:40,960 --> 00:38:44,960 Speaker 4: is a dead end makes scientists very excited. Yeah, that 697 00:38:45,040 --> 00:38:47,880 Speaker 4: nuance is where it is, and so that's in machine learning. 698 00:38:48,000 --> 00:38:52,680 Speaker 4: Sorry Hamiltonian. Then there's examples in differential equations, So can 699 00:38:52,719 --> 00:38:55,680 Speaker 4: I actually come up with differential equations and solve them? 700 00:38:56,080 --> 00:38:57,880 Speaker 4: And if I can solve them, you could look at 701 00:38:58,239 --> 00:39:02,320 Speaker 4: things like an obvious Stokes goes into weather. There's financial 702 00:39:02,360 --> 00:39:06,160 Speaker 4: differential equations that you can better predict. So differential equations, 703 00:39:06,239 --> 00:39:09,080 Speaker 4: there's many different examples there. And then I would say 704 00:39:09,080 --> 00:39:12,680 Speaker 4: that two others are optimization and then there's quantum versions 705 00:39:12,680 --> 00:39:15,399 Speaker 4: of machine learning that are very exciting as well. 706 00:39:16,040 --> 00:39:19,160 Speaker 3: Cleveland Clinic one of the organizations that you guys have 707 00:39:19,239 --> 00:39:22,000 Speaker 3: worked with. Why would the Cleveland Clinic be calling you up? 708 00:39:22,360 --> 00:39:25,239 Speaker 4: Because that problem that they want to look at. So 709 00:39:26,120 --> 00:39:29,400 Speaker 4: they've also done similar problem to the recent lab. So 710 00:39:29,400 --> 00:39:32,680 Speaker 4: they've taken that method now and they've looked at molecules 711 00:39:32,680 --> 00:39:37,200 Speaker 4: that matter for drug design. So they're fundamentally looking at 712 00:39:37,239 --> 00:39:42,040 Speaker 4: those molecules that matter for eventually replacing some of the steps. 713 00:39:42,400 --> 00:39:46,080 Speaker 4: So they're investing to see how reliable it can be done. 714 00:39:46,160 --> 00:39:48,880 Speaker 4: And so there's a scientist there that's done many iterations 715 00:39:48,920 --> 00:39:52,040 Speaker 4: now using the techniques that were done first with the 716 00:39:52,040 --> 00:39:56,040 Speaker 4: team in Japan, they've now replicated that for new molecules 717 00:39:56,520 --> 00:40:01,080 Speaker 4: that are essential primitives for eventually designed drugs and things 718 00:40:01,080 --> 00:40:03,120 Speaker 4: that may matter for medical Yeah. 719 00:40:03,239 --> 00:40:08,360 Speaker 3: And also there's some finance firms yep, HBC, Van Good yep, 720 00:40:08,840 --> 00:40:09,960 Speaker 3: and their interest is. 721 00:40:09,920 --> 00:40:13,560 Speaker 4: What so that was the differential equation and optimization. So 722 00:40:14,200 --> 00:40:18,400 Speaker 4: if you are doing very large calculations like risk portfolio, 723 00:40:19,000 --> 00:40:21,440 Speaker 4: or if you want to model the black Shaw's equation 724 00:40:21,600 --> 00:40:23,759 Speaker 4: or things like this that are fundamental for them to 725 00:40:24,000 --> 00:40:27,040 Speaker 4: make better predictions, come up with better trades and things 726 00:40:27,080 --> 00:40:31,399 Speaker 4: like this. That is a very hard computational task. And 727 00:40:31,480 --> 00:40:35,240 Speaker 4: so rather than quantum replacing that whole problem, can quantum 728 00:40:35,280 --> 00:40:39,200 Speaker 4: be a subroutine in there? And what HSBC showed is 729 00:40:39,239 --> 00:40:41,680 Speaker 4: they showed they could take their real data, they could 730 00:40:41,719 --> 00:40:45,239 Speaker 4: take their real classical method and they just replaced a 731 00:40:45,280 --> 00:40:48,160 Speaker 4: tiny part of it. They replaced a tiny part of 732 00:40:48,200 --> 00:40:51,319 Speaker 4: it with a quantum subroutine that allowed them to come 733 00:40:51,400 --> 00:40:54,759 Speaker 4: up with better predictions of the weights that then when 734 00:40:54,800 --> 00:40:57,680 Speaker 4: they were to compare trial A versus Trial B, it 735 00:40:57,760 --> 00:41:02,200 Speaker 4: was thirty four percent better at algorithmic tun and that's 736 00:41:02,200 --> 00:41:03,160 Speaker 4: a big deal for them. 737 00:41:03,560 --> 00:41:04,120 Speaker 3: It's huge. 738 00:41:04,239 --> 00:41:07,520 Speaker 4: Yes, Now do they need to do more trials? Do 739 00:41:07,600 --> 00:41:10,160 Speaker 4: they need to see is this a heuristic algorithm? Do 740 00:41:10,239 --> 00:41:13,200 Speaker 4: we need to be careful? Is there other classical algorithms 741 00:41:13,239 --> 00:41:15,520 Speaker 4: that go into these are great questions that are now 742 00:41:16,040 --> 00:41:20,359 Speaker 4: being investigated. So think of this period of heuristic algorithms 743 00:41:20,920 --> 00:41:24,680 Speaker 4: is really a period of scientific discovery using these machines, 744 00:41:25,640 --> 00:41:28,560 Speaker 4: knowing that we want to continue and build the ones 745 00:41:28,640 --> 00:41:32,160 Speaker 4: which have determinist their algorithms that can run. 746 00:41:33,000 --> 00:41:36,280 Speaker 3: Do the people who would profit the most space starting 747 00:41:36,280 --> 00:41:41,640 Speaker 3: to run quantum experiments realize that they would profit so 748 00:41:41,719 --> 00:41:45,080 Speaker 3: much from running quantum experience And does the world know this? 749 00:41:45,719 --> 00:41:49,520 Speaker 3: You've given us a couple of specific examples, but generally speaking, 750 00:41:49,520 --> 00:41:51,680 Speaker 3: there must be a very large universe of people who 751 00:41:51,719 --> 00:41:54,760 Speaker 3: could gain from at least starting to play in the space. 752 00:41:55,360 --> 00:41:59,680 Speaker 4: So the enterprises that use computation as key for their 753 00:42:00,840 --> 00:42:04,840 Speaker 4: understand the limits of classical computation and they're very interested 754 00:42:04,880 --> 00:42:09,400 Speaker 4: to get started. The universities are very interested. Could we 755 00:42:09,440 --> 00:42:13,920 Speaker 4: get more students doing more algorithms one hundred percent. Some 756 00:42:14,080 --> 00:42:17,160 Speaker 4: of the limitations on the rate of algorithm discovery is 757 00:42:17,200 --> 00:42:21,080 Speaker 4: because people are thinking through the classical way of writing algorithms. 758 00:42:21,120 --> 00:42:23,680 Speaker 4: My belief is yes, So this is why we want 759 00:42:23,719 --> 00:42:25,960 Speaker 4: to get more and more students and things, because it's 760 00:42:26,040 --> 00:42:29,080 Speaker 4: just starting. But I would say in general most people 761 00:42:29,160 --> 00:42:32,359 Speaker 4: are aware of it. Could we get more, could we 762 00:42:32,440 --> 00:42:33,239 Speaker 4: accelerate it? 763 00:42:33,480 --> 00:42:33,719 Speaker 3: Yes? 764 00:42:33,920 --> 00:42:36,040 Speaker 4: Do we need to make better hardware, do we need 765 00:42:36,080 --> 00:42:38,600 Speaker 4: to come up with better libraries, yes? Do we need 766 00:42:38,640 --> 00:42:42,080 Speaker 4: better software yes, But it's all happening over the next 767 00:42:42,080 --> 00:42:42,720 Speaker 4: few years. 768 00:42:43,080 --> 00:42:45,279 Speaker 3: Is it hard to get someone who's spent their entire 769 00:42:45,320 --> 00:42:48,560 Speaker 3: life thinking in terms of solving problems to classical means 770 00:42:48,719 --> 00:42:50,600 Speaker 3: to make the transition to this new paradigm. 771 00:42:51,360 --> 00:42:54,160 Speaker 4: There's a lot of examples when you approach something with 772 00:42:54,280 --> 00:42:57,680 Speaker 4: the classical intuition, it's not the right way to do 773 00:42:57,719 --> 00:43:00,879 Speaker 4: it when you approach it through the quantum. But if 774 00:43:00,920 --> 00:43:05,200 Speaker 4: people are being taught to understand the fundamentals of the math, 775 00:43:05,719 --> 00:43:09,320 Speaker 4: then a lot of the techniques carry across. I don't 776 00:43:09,360 --> 00:43:13,240 Speaker 4: recommend people need to learn about entanglement or supersition because 777 00:43:14,040 --> 00:43:18,280 Speaker 4: whilst the physicists will argue like spooky action a distance 778 00:43:18,320 --> 00:43:21,520 Speaker 4: and all these type of things, entanglement is the power. Yes, 779 00:43:21,600 --> 00:43:24,799 Speaker 4: that's how physicists are labeled. How quantum is different. But 780 00:43:24,880 --> 00:43:28,280 Speaker 4: I would say, do we need some physicists really worrying 781 00:43:28,440 --> 00:43:31,680 Speaker 4: thinking about that? Yes, but we need more applied mathematicians 782 00:43:31,680 --> 00:43:34,960 Speaker 4: that are realizing they can use this as a as 783 00:43:35,000 --> 00:43:36,640 Speaker 4: a different way of looking at the problems. 784 00:43:36,840 --> 00:43:38,680 Speaker 3: Yeah. And when I asked you one question, know is 785 00:43:39,400 --> 00:43:42,680 Speaker 3: we're describing a a It's more than a new technology. 786 00:43:42,680 --> 00:43:45,840 Speaker 3: We're talking about a new paradigm. It's a way of 787 00:43:45,880 --> 00:43:49,480 Speaker 3: thinking about problems. Can you compare this to kind of 788 00:43:49,680 --> 00:43:54,919 Speaker 3: previous technological paradigms. I'm thinking at the last couple hundred years, 789 00:43:54,920 --> 00:43:58,520 Speaker 3: what does this rank in terms of a new field 790 00:43:58,600 --> 00:43:59,480 Speaker 3: that we've opened up. 791 00:44:00,120 --> 00:44:02,160 Speaker 4: It's a hard question to answer, but I often say 792 00:44:02,360 --> 00:44:05,600 Speaker 4: the history of computing, this will be the first time 793 00:44:06,200 --> 00:44:10,880 Speaker 4: that computation is branched between classical and quantum. I like 794 00:44:11,239 --> 00:44:14,759 Speaker 4: thinking reading a lot in the past. One of the 795 00:44:14,800 --> 00:44:18,279 Speaker 4: things that I think was a way we changed as 796 00:44:18,280 --> 00:44:22,560 Speaker 4: a society was the invention of zero. Before zero, math 797 00:44:22,800 --> 00:44:27,360 Speaker 4: was limited. Realizing that numbers have a number a zero 798 00:44:27,880 --> 00:44:30,720 Speaker 4: allowed us to develop a whole set of new mathematics 799 00:44:31,120 --> 00:44:35,320 Speaker 4: that then went on and defined like everything from waves 800 00:44:35,360 --> 00:44:39,120 Speaker 4: to calculus to all of that. Yes, we can describe 801 00:44:39,200 --> 00:44:41,560 Speaker 4: it with that same math, but when we describe it 802 00:44:41,600 --> 00:44:44,839 Speaker 4: with that math, it gets exponentially big and gets impractical 803 00:44:44,920 --> 00:44:47,960 Speaker 4: to do. Now we can actually work on it, I 804 00:44:47,960 --> 00:44:50,520 Speaker 4: would say, if I had to give you a quick answer, 805 00:44:50,640 --> 00:44:55,080 Speaker 4: maybe going all the way back to when we were accepted. 806 00:44:54,719 --> 00:44:56,880 Speaker 3: Zero, I thought you were going to say, like the airplane, 807 00:44:57,000 --> 00:44:59,840 Speaker 3: but in fact, yeah, you went several orders of magnitude 808 00:45:00,080 --> 00:45:00,359 Speaker 3: on that. 809 00:45:00,560 --> 00:45:03,799 Speaker 4: Yes, but but I think it's sort of fundamental. 810 00:45:04,120 --> 00:45:07,560 Speaker 3: This is absolutely fascinating. Thank you so much for chatting 811 00:45:07,560 --> 00:45:08,080 Speaker 3: with me about it. 812 00:45:08,160 --> 00:45:08,799 Speaker 4: Thank you for your time. 813 00:45:11,280 --> 00:45:14,440 Speaker 3: Hey, listeners. So normally we end this episode here, but 814 00:45:14,480 --> 00:45:17,800 Speaker 3: the Tech Week attendees asked Jay some really great questions, 815 00:45:18,239 --> 00:45:21,000 Speaker 3: questions I wish I'd asked, so we wanted to include 816 00:45:21,040 --> 00:45:22,680 Speaker 3: those here. Enjoy. 817 00:45:24,480 --> 00:45:26,759 Speaker 5: Hi, J, thank you so much for the great presentation. 818 00:45:27,239 --> 00:45:30,840 Speaker 5: My name is Trixie Apiado. I work for Willis Towers Watson, 819 00:45:30,880 --> 00:45:34,839 Speaker 5: an insurance broker. I help seisos identify and quantify their 820 00:45:34,880 --> 00:45:38,880 Speaker 5: cyber risk so they can prepare for threats before they happen, 821 00:45:39,280 --> 00:45:42,000 Speaker 5: and so quantum threats keep me up at night. You 822 00:45:42,120 --> 00:45:46,120 Speaker 5: mentioned so many good problems that quantum can solve. It 823 00:45:46,160 --> 00:45:50,839 Speaker 5: can also break encryptions in our classical computer systems. So 824 00:45:51,960 --> 00:45:56,080 Speaker 5: what safeguards or policies do you implement in your teams 825 00:45:56,440 --> 00:46:00,920 Speaker 5: to build quantum capabilities responsibly? And what can we do 826 00:46:01,680 --> 00:46:05,040 Speaker 5: for people in this room as builders and users to 827 00:46:05,160 --> 00:46:09,560 Speaker 5: secure our data in systems before quantum computers become more 828 00:46:09,680 --> 00:46:11,960 Speaker 5: energy efficient, cheaper, and more available. 829 00:46:12,840 --> 00:46:16,120 Speaker 4: So it's a great question. So yes, one of the 830 00:46:16,160 --> 00:46:20,520 Speaker 4: algorithms for quantum computing is to break our traditional encryption. 831 00:46:21,239 --> 00:46:25,080 Speaker 4: So at IBM Research, we were aware of this from 832 00:46:25,160 --> 00:46:29,640 Speaker 4: day one. We've come up with algorithms that we believe 833 00:46:29,840 --> 00:46:32,759 Speaker 4: and have very strong evidence will not be broken by 834 00:46:32,760 --> 00:46:36,120 Speaker 4: a quantum or classical computer, and this has selected them. 835 00:46:36,600 --> 00:46:42,000 Speaker 4: So first the scientific technical question, security is saved. There 836 00:46:42,000 --> 00:46:46,879 Speaker 4: are algorithms that exist that we can implement that neither 837 00:46:47,040 --> 00:46:50,880 Speaker 4: a quantum or classical computer can break. So the technical 838 00:46:50,920 --> 00:46:55,319 Speaker 4: answer is we're all okay. The more complicated answer is 839 00:46:55,360 --> 00:46:59,920 Speaker 4: a social and society answer. Encryption was built in classical 840 00:47:00,040 --> 00:47:03,520 Speaker 4: computing in a way that was never thought of being upgraded. 841 00:47:04,320 --> 00:47:08,400 Speaker 4: It's mixed everywhere. Some of it is downstream, some of 842 00:47:08,440 --> 00:47:11,040 Speaker 4: it is like software that you may use. Some of 843 00:47:11,080 --> 00:47:15,319 Speaker 4: it is software that you've developed. And I get that 844 00:47:15,440 --> 00:47:17,279 Speaker 4: if you've got a product and you want to have 845 00:47:17,400 --> 00:47:20,359 Speaker 4: it secure for the next ten years, you probably want 846 00:47:20,360 --> 00:47:22,839 Speaker 4: to think about how you're going to upgrade it, or 847 00:47:23,000 --> 00:47:26,200 Speaker 4: if you have data that needs to be secure for 848 00:47:26,400 --> 00:47:29,680 Speaker 4: the next ten years, it needs to upgrade to new encryption. 849 00:47:30,320 --> 00:47:33,360 Speaker 4: So the real challenge is more of a social business 850 00:47:33,400 --> 00:47:37,880 Speaker 4: problem of how do we actually transition from old encryption 851 00:47:38,000 --> 00:47:41,120 Speaker 4: to new encryption knowing this is going to happen. So 852 00:47:41,239 --> 00:47:43,759 Speaker 4: we at IBM have been very proactive on this. We've 853 00:47:43,800 --> 00:47:47,160 Speaker 4: developed tools where we can determine where encryption is used, 854 00:47:47,600 --> 00:47:51,600 Speaker 4: We've developed tools which can show you how to replace it, 855 00:47:51,760 --> 00:47:55,960 Speaker 4: and we early on have made sure the Mainframe when 856 00:47:56,040 --> 00:47:58,560 Speaker 4: we made these algorithms. So I think it was Z 857 00:47:58,760 --> 00:48:01,600 Speaker 4: sixteen that was the first version of the Mainframe to 858 00:48:01,800 --> 00:48:06,799 Speaker 4: have these quantum safe algorithms implemented. So my answer to 859 00:48:06,840 --> 00:48:10,520 Speaker 4: your question is, yes, there's a real problem, but it's 860 00:48:10,520 --> 00:48:13,560 Speaker 4: not a technical problem. It's a social and business problem. 861 00:48:13,560 --> 00:48:17,040 Speaker 4: And I'm not minimizing that. I understand that that is 862 00:48:17,080 --> 00:48:20,040 Speaker 4: a lot of work you need to start now. You 863 00:48:20,120 --> 00:48:22,200 Speaker 4: need to come up and do a you need to 864 00:48:22,239 --> 00:48:25,279 Speaker 4: make it part of your IT transformation. You need to 865 00:48:25,680 --> 00:48:30,239 Speaker 4: get onto it. And I realize I realize it's not 866 00:48:30,280 --> 00:48:32,960 Speaker 4: going to take zero time because it's not an easy 867 00:48:33,000 --> 00:48:36,279 Speaker 4: problem to do. So the short answer is one we 868 00:48:36,400 --> 00:48:39,440 Speaker 4: developed algorithms that we can't and we're developing tools to 869 00:48:39,480 --> 00:48:40,640 Speaker 4: help you in that transformation. 870 00:48:41,120 --> 00:48:44,560 Speaker 6: Thank you so much, Thank you. My name is Emma. 871 00:48:44,680 --> 00:48:48,279 Speaker 6: I'm a product manager at Expedia working on software side 872 00:48:48,280 --> 00:48:51,480 Speaker 6: of things. My question is around the non technical roles 873 00:48:51,560 --> 00:48:55,319 Speaker 6: outside of the researchers, the mathematicians, the builders. How can 874 00:48:55,440 --> 00:48:58,680 Speaker 6: the rest of us, whether it be policymakers, those in 875 00:48:58,719 --> 00:49:02,080 Speaker 6: the legal fields, those thinking about what use cases quantum 876 00:49:02,160 --> 00:49:04,440 Speaker 6: can solve for in the future, what should we be 877 00:49:04,520 --> 00:49:06,879 Speaker 6: thinking about and how can we prepare for that. 878 00:49:07,200 --> 00:49:09,319 Speaker 4: It's a good question. I think this is part of 879 00:49:09,400 --> 00:49:13,279 Speaker 4: the requirement of the scientists to being able to articulate 880 00:49:13,960 --> 00:49:16,560 Speaker 4: where they are. We need a forum for those type 881 00:49:16,560 --> 00:49:19,640 Speaker 4: of discussions. I think a lot of this can fit 882 00:49:19,760 --> 00:49:23,040 Speaker 4: within the forums that we already have for classical and AI, 883 00:49:23,640 --> 00:49:25,640 Speaker 4: and I think we need to just be asking how 884 00:49:25,680 --> 00:49:28,680 Speaker 4: do we actually bring them into them. Because I don't 885 00:49:28,960 --> 00:49:32,160 Speaker 4: think of quantum as a replacement of compute. I think 886 00:49:32,160 --> 00:49:35,160 Speaker 4: of it as an accelerator that expands what is possible, 887 00:49:35,840 --> 00:49:38,600 Speaker 4: and I think we can ask those questions in those forums, 888 00:49:39,000 --> 00:49:41,480 Speaker 4: are we doing enough now? I think I agree with you. No, 889 00:49:41,760 --> 00:49:43,040 Speaker 4: I don't know the answer to it. 890 00:49:43,920 --> 00:49:47,200 Speaker 6: I think it's a really interesting perspective because those existing 891 00:49:47,239 --> 00:49:50,680 Speaker 6: forums do start to bring in those other fields as well, 892 00:49:50,960 --> 00:49:53,840 Speaker 6: so it could warrant the same sort of discussion and. 893 00:49:54,160 --> 00:49:59,880 Speaker 4: Active And I understand those forums right now AI is 894 00:50:00,080 --> 00:50:03,360 Speaker 4: probably dominating, and it should be like we are going 895 00:50:03,400 --> 00:50:07,440 Speaker 4: through a period of time where AI is impacting society. 896 00:50:08,000 --> 00:50:11,160 Speaker 4: The technology is impacting society in big ways. So I 897 00:50:11,239 --> 00:50:14,840 Speaker 4: totally understand that most of their focus should be on AI, 898 00:50:15,040 --> 00:50:17,680 Speaker 4: but we should start to ask where is quantum in 899 00:50:18,120 --> 00:50:18,719 Speaker 4: that as well? 900 00:50:19,840 --> 00:50:23,640 Speaker 7: Hi, I'm Gobi and I'm a graduating PhD student at 901 00:50:23,680 --> 00:50:27,640 Speaker 7: Northwestern and also a member of south Park Commons, which 902 00:50:27,680 --> 00:50:31,240 Speaker 7: is a fund here. You mentioned earlier that some problems 903 00:50:31,239 --> 00:50:34,239 Speaker 7: are best solved by classical versus some problems are best 904 00:50:34,280 --> 00:50:36,799 Speaker 7: solved by quantum. When we're thinking about this, if we're 905 00:50:36,840 --> 00:50:39,239 Speaker 7: not experts in quantum, but we're thinking about this from 906 00:50:39,239 --> 00:50:41,759 Speaker 7: an AI perspective, could you just clarify when we think 907 00:50:41,760 --> 00:50:46,120 Speaker 7: about quantum, what is deterministic and what is not deterministic. 908 00:50:46,640 --> 00:50:49,120 Speaker 4: I think the future of computing we've got to get 909 00:50:49,160 --> 00:50:52,520 Speaker 4: our heads around is that not everything is deterministic, and 910 00:50:52,560 --> 00:50:55,080 Speaker 4: it's much more going to be probilistic. How do you 911 00:50:55,239 --> 00:50:58,440 Speaker 4: handle error bars? How do you put confidence? I think 912 00:50:58,520 --> 00:51:01,440 Speaker 4: a lot of those questions which you're referring to in 913 00:51:01,480 --> 00:51:05,200 Speaker 4: AI are going to completely apply in quantum. I actually 914 00:51:05,200 --> 00:51:10,799 Speaker 4: think it's a mistake to compare AI verse quantum. I 915 00:51:10,880 --> 00:51:15,480 Speaker 4: actually think of quantum as much its quantum verse classical compute, 916 00:51:15,520 --> 00:51:18,640 Speaker 4: and AI is going to come across on top. So 917 00:51:18,840 --> 00:51:21,560 Speaker 4: as we go forward and we get a better understanding, 918 00:51:21,719 --> 00:51:25,200 Speaker 4: I'm not going to say quantum is going to replace 919 00:51:25,239 --> 00:51:28,239 Speaker 4: the classical compute that enables AI, but I think some 920 00:51:28,320 --> 00:51:30,520 Speaker 4: of the math you do in AI will be able 921 00:51:30,560 --> 00:51:33,319 Speaker 4: to go to both. So what can we formally prove? 922 00:51:34,120 --> 00:51:37,000 Speaker 4: I can come up with a problem where I take 923 00:51:37,040 --> 00:51:39,480 Speaker 4: a circle and I color half of it red and 924 00:51:39,520 --> 00:51:42,359 Speaker 4: half of it of blue, and then I say I'm going 925 00:51:42,440 --> 00:51:46,399 Speaker 4: to apply an operation that takes those dots make it. Say, 926 00:51:46,480 --> 00:51:48,839 Speaker 4: let's say ten dots over here red, ten dots over 927 00:51:48,880 --> 00:51:51,520 Speaker 4: here blue, and I'm going to wind them around many 928 00:51:51,600 --> 00:51:54,719 Speaker 4: many times. I can then show you that if you 929 00:51:54,800 --> 00:51:57,720 Speaker 4: feed that into a classical computer, it's a classical random 930 00:51:57,800 --> 00:52:01,399 Speaker 4: number generator. You can give your as much data as 931 00:52:01,440 --> 00:52:03,960 Speaker 4: you want, you will never be able to say did 932 00:52:04,000 --> 00:52:06,919 Speaker 4: the red come from the left side or the right side. 933 00:52:06,960 --> 00:52:10,080 Speaker 4: You would take infinite data like it is like you 934 00:52:10,120 --> 00:52:13,440 Speaker 4: would have to break a classical random number generator. I 935 00:52:13,480 --> 00:52:16,960 Speaker 4: can show you a quantum algorithm that can do that deterministically. 936 00:52:18,040 --> 00:52:20,920 Speaker 4: So where we're thinking is when the data appears to 937 00:52:21,000 --> 00:52:26,120 Speaker 4: be completely unstructured or you looks essentially like a complete 938 00:52:26,200 --> 00:52:30,520 Speaker 4: random number to the classical methods, there are quantum methods 939 00:52:30,560 --> 00:52:33,240 Speaker 4: that can actually potentially find that structure. 940 00:52:37,760 --> 00:52:40,200 Speaker 3: That's it for this episode of Smart Talks with IBM. 941 00:52:40,680 --> 00:52:42,880 Speaker 3: If you haven't already, be sure to check out my 942 00:52:42,960 --> 00:52:48,360 Speaker 3: conversation with IBM Chairman and CEO Arvind Krishna, and stay tuned. 943 00:52:48,760 --> 00:52:53,759 Speaker 3: Another episode is coming soon. Smart Talks with IBM is 944 00:52:53,800 --> 00:52:57,920 Speaker 3: produced by Matt Romano, Amy Gains, McQuaid, Trina Menino, and 945 00:52:58,040 --> 00:53:02,720 Speaker 3: Jake Harper. Engineering by Bird Lawrence, Mastering by Sarah Buger, 946 00:53:03,080 --> 00:53:08,200 Speaker 3: music by Gramoscope, Strategy by Tatiana Lieberman, Cassidy Meyer and 947 00:53:08,320 --> 00:53:12,480 Speaker 3: Sofia Derlon. Smart Talks with IBM is a production of 948 00:53:12,560 --> 00:53:17,799 Speaker 3: Pushkin Industries and Ruby Studio at iHeartMedia to find more 949 00:53:17,800 --> 00:53:22,560 Speaker 3: Pushkin podcasts. Listen on the iHeartRadio app, Apple Podcasts, or 950 00:53:22,600 --> 00:53:27,080 Speaker 3: wherever you listen to podcasts. I'm Malcolm Godwell. This is 951 00:53:27,120 --> 00:53:31,600 Speaker 3: a paid advertisement from IBM. The conversations on this podcast 952 00:53:31,880 --> 00:53:42,120 Speaker 3: don't necessarily represent IBM's positions, strategies, or opinions.