WEBVTT - Unlocking Our Quantum Future 0:00:02.680 --> 0:00:05.240 Hello, This is Malcolm Gladwell and you're listening to Smart 0:00:05.240 --> 0:00:08.719 Talks with IBM. Every year, tech Week brings thousands of 0:00:08.760 --> 0:00:12.160 people together to network and learn about what's emerging across 0:00:12.200 --> 0:00:16.560 the technology ecosystem, and at this year's conference in San Francisco, 0:00:16.960 --> 0:00:19.640 I had an amazing opportunity to sit down in front 0:00:19.680 --> 0:00:22.680 of a live audience with Jay Gambetta. Jay has been 0:00:22.720 --> 0:00:26.040 with IBM for years and was recently promoted to Director 0:00:26.079 --> 0:00:29.800 of Research. In this job, Jay has an important mission 0:00:30.240 --> 0:00:33.600 helping the company build the future of computing. In the 0:00:33.680 --> 0:00:36.280 last episode of Smart Talks, I began to learn about 0:00:36.320 --> 0:00:41.199 quantum computing from IBM Chairman and CEO Arvind Krishna. But 0:00:41.320 --> 0:00:44.720 this conversation I had with Jay went even deeper and 0:00:44.880 --> 0:00:48.520 convinced me that the development of quantum isn't just a fun, 0:00:48.920 --> 0:00:52.080 exciting new paradigm of computing. It may be one of 0:00:52.080 --> 0:01:02.480 the most important scientific achievements of my lifetime. Jay, Morning, Morning, 0:01:02.800 --> 0:01:05.840 Welcome to Smart Talks with IBM. Thank your special live 0:01:05.880 --> 0:01:09.480 recording here for tech Week, and congratulations. How long have 0:01:09.520 --> 0:01:11.640 you been Head of Research at IBM? 0:01:12.200 --> 0:01:15.880 Since October one? It's October tenth today, since nine days, 0:01:15.959 --> 0:01:16.479 nine days. 0:01:16.920 --> 0:01:19.360 Can you just talk a little about the position. This 0:01:19.520 --> 0:01:23.720 is one of the most important positions in research in 0:01:23.760 --> 0:01:24.200 the world. 0:01:24.720 --> 0:01:27.800 IBM research has been around for eighty years and it's 0:01:27.880 --> 0:01:32.920 done some tremendous technology, a lot of inventions and fundamentals 0:01:32.920 --> 0:01:38.040 for semiconductors, algorithms, AI. Yeah, I think if we look 0:01:38.080 --> 0:01:40.039 back to where a lot of the innovation and the 0:01:40.080 --> 0:01:42.680 technology of the world comes from, I think you can 0:01:42.680 --> 0:01:46.240 find Ibram's footprints on it, and you can find IBM research. 0:01:46.319 --> 0:01:49.560 So yeah, I'm very excited for the opportunity, but I'm 0:01:49.600 --> 0:01:52.160 also aware that there's big shoes to fill, and I'm 0:01:52.240 --> 0:01:56.960 looking forward to how we take IBM research forward. Obviously, 0:01:57.040 --> 0:02:00.000 I'm going to be bringing a lot of the quantum side, 0:02:00.000 --> 0:02:03.520 which we're going to talk about later. Beyond quantum, there's 0:02:03.600 --> 0:02:07.040 important work that needs to happen in AI hybrid cloud, 0:02:07.560 --> 0:02:09.920 and I think we're going to also enter into this 0:02:10.000 --> 0:02:13.400 new period of mathematics where we get to use quantum 0:02:13.440 --> 0:02:17.560 machines and also AI machines. And there's some really good 0:02:17.720 --> 0:02:19.920 hard mathematical questions to answer. 0:02:20.040 --> 0:02:21.640 How many people do you have working for you? 0:02:22.040 --> 0:02:25.760 I've been researchers in the three thousand researchers across many 0:02:25.760 --> 0:02:29.040 different labs around the world. Our main lab is in Yorktown, 0:02:29.120 --> 0:02:31.200 but then we have the lab actually out on the 0:02:31.280 --> 0:02:34.680 West coast in Armatan or sbl now, and then we 0:02:34.760 --> 0:02:38.080 have one in Zurich, Japan, and a few others around 0:02:38.120 --> 0:02:38.440 the world. 0:02:39.120 --> 0:02:41.600 Tell me a little bit before we get into quantum. 0:02:41.760 --> 0:02:45.040 I'm just curious about your path. So you're Australian. Yep, 0:02:45.440 --> 0:02:48.840 we were talking about earlier. Backstage. Your accent has become muted. 0:02:49.400 --> 0:02:52.120 You should crank it up because it's. 0:02:52.080 --> 0:02:54.880 Yeah, I'm slowly losing my Australian accent. I've been in 0:02:54.919 --> 0:02:57.920 the US since two thousand and four, so accent, you know, 0:02:58.000 --> 0:03:00.760 to sound very Australian. Yeah, but how do you practice it? 0:03:00.800 --> 0:03:04.160 Maybe I got to go back to Australia. Here more 0:03:04.200 --> 0:03:06.960 Australians say gooday, how's it going? Things like that? 0:03:08.040 --> 0:03:10.560 And you you didn't grow up thinking you're going to 0:03:10.600 --> 0:03:12.080 be a scientist one day, now. 0:03:12.160 --> 0:03:16.160 I grew up in a pretty normal life. My dreams 0:03:16.400 --> 0:03:19.079 as a kid was building things, so I was either 0:03:19.120 --> 0:03:22.160 going to be a carpenter or a mechanic. But I 0:03:22.200 --> 0:03:24.400 had some great teachers that inspired me to go to 0:03:24.520 --> 0:03:27.640 university and I didn't even know honestly what a scientist was. 0:03:27.960 --> 0:03:32.480 And then I found myself at university doing science, particular physics, 0:03:32.600 --> 0:03:34.760 and I ended up loving it. So you go from 0:03:34.800 --> 0:03:38.040 there to what do you do your PhD? So I 0:03:38.040 --> 0:03:40.960 did my undergrad in Australia. I did it actually in 0:03:41.240 --> 0:03:47.600 laser science, so I think I watched some TV show 0:03:47.640 --> 0:03:50.520 in lasers seemed interesting, so I wanted to learn about lasers, 0:03:51.080 --> 0:03:54.160 and then I realized in trying to understand lasers there 0:03:54.240 --> 0:03:57.560 was this quantum mechanics, and so I was like, all right, 0:03:58.360 --> 0:04:01.000 I want to actually understand this mechanics. So I did 0:04:01.000 --> 0:04:03.440 my equivalent of what you and the US gore masters. 0:04:03.800 --> 0:04:05.720 We call it honors in Australia, but we do a 0:04:05.760 --> 0:04:09.320 research project. I said, I wanted to shoot lasers into 0:04:09.360 --> 0:04:13.000 atoms and measure cross sections and I got really into 0:04:13.080 --> 0:04:15.640 quantum physics. So then I decided, all right, I don't 0:04:15.720 --> 0:04:19.719 understand this quantum physics. I want to do my PhD 0:04:19.839 --> 0:04:23.440 in interpretations of quantum mechanics. So I jumped in and said, 0:04:23.440 --> 0:04:26.200 all right, what is this quantum mechanics? Why is everyone 0:04:26.320 --> 0:04:29.919 arguing on these different interpretations? Then I finished my PhD 0:04:29.960 --> 0:04:33.560 in Australia doing that. Then I moved over at the 0:04:33.680 --> 0:04:38.799 end of my PhD interpretations, it's more people arguing about 0:04:38.800 --> 0:04:42.920 the equations whilst I think it's really important. I decided 0:04:43.000 --> 0:04:45.560 if it's going to be like a collapse equation versus 0:04:45.600 --> 0:04:50.200 many worlds, or a hidden variable model, or that just 0:04:50.279 --> 0:04:54.479 quantum mechanics decoheres because we don't see supersitions in the 0:04:54.600 --> 0:04:58.839 everyday world because it interacts with environment. The only way 0:04:58.839 --> 0:05:02.000 to answer that question was build a quantum computer. And 0:05:02.080 --> 0:05:04.160 so then I decided at the end of my PhD, 0:05:04.240 --> 0:05:07.080 I wanted to work out how to build a quantum computer. 0:05:07.640 --> 0:05:10.320 And then I left there and I went to Yale, 0:05:10.760 --> 0:05:14.000 and then at Yale, that's where I got into superconducting cubits, 0:05:14.360 --> 0:05:16.680 which just a few days ago one of the professors 0:05:16.680 --> 0:05:19.120 there just won the Nobel Price this year. 0:05:19.480 --> 0:05:22.960 Oh wow, I'm very interested in tracing because your career 0:05:24.200 --> 0:05:27.400 follows the arc of quantum computing in a certain way. 0:05:27.480 --> 0:05:30.520 Right at the time when you asked the question, what 0:05:30.560 --> 0:05:32.200 I really want to do is to figure out how 0:05:32.200 --> 0:05:35.120 to build a quantum computer. Where are we in quantum 0:05:35.200 --> 0:05:36.400 computing at that point? 0:05:36.560 --> 0:05:39.440 Yeah, So that would have been nineteen ninety so there 0:05:39.480 --> 0:05:43.320 was Shaw's algorithm came out let's say ninety five. There 0:05:43.360 --> 0:05:46.359 was a lot of theory, and then the reason I 0:05:46.400 --> 0:05:50.599 went to Yale is because people had started to show 0:05:50.680 --> 0:05:55.359 that they could see quantum effects in electrical circuits. So 0:05:55.440 --> 0:05:58.760 these macroscopic objects they were starting to behave quantum mechanical 0:05:59.320 --> 0:06:02.320 There was a really significant breakthrough in nineteen ninety nine 0:06:02.400 --> 0:06:06.559 where Yazoo Nakamura in Japan showed that a cubit could 0:06:06.640 --> 0:06:10.160 exist in these electrical circuits. And then I found out 0:06:10.200 --> 0:06:12.280 the group at Yale were really trying to take these 0:06:12.320 --> 0:06:17.080 electrical circuits and couple them together. And so it was like, 0:06:17.320 --> 0:06:21.400 if I can build something using electrical circuits and they're big, 0:06:21.960 --> 0:06:24.680 that that's the best way that you cancide to test 0:06:24.760 --> 0:06:28.440 and understand whether quantum mechanics breaks down at a macroscopic 0:06:28.480 --> 0:06:31.560 scale or not. Can we actually make them behave as cubits? 0:06:31.960 --> 0:06:34.279 And I agree. When I came to Yale, the qubits 0:06:35.120 --> 0:06:39.360 were not very good. They were actually a couple of nanoseconds. 0:06:39.640 --> 0:06:43.680 They were unstable. Electron would jump onto the chip and 0:06:43.720 --> 0:06:46.279 then they would change all their configurations, so you have 0:06:46.360 --> 0:06:49.680 to restart your experiment. And so for the first time 0:06:49.720 --> 0:06:52.240 at Yale. It's kind of what the challenge there was, 0:06:52.400 --> 0:06:54.720 how do we make a cubit? How do we make 0:06:54.760 --> 0:06:58.200 a stable cubit? And that took about five years, and 0:06:58.279 --> 0:07:00.240 that took us up to two thousand and seven. And 0:07:01.640 --> 0:07:03.479 I think the rest of the world looks and says 0:07:03.560 --> 0:07:07.000 quantums like just blowing up, But it's actually been like 0:07:07.080 --> 0:07:10.960 almost phases theory, showing that we got the algorithms, how 0:07:10.960 --> 0:07:13.080 do we make a cubit? How do we couple of 0:07:13.080 --> 0:07:15.800 the cubits together? And now we're in the scaling phase. 0:07:16.280 --> 0:07:20.720 Describe for us because many people in this room, me included, 0:07:21.280 --> 0:07:24.560 have only a kind of surface level understanding of what 0:07:24.560 --> 0:07:26.840 we mean when we use that phrase. What is the 0:07:26.880 --> 0:07:30.560 difference between classical computing and quantum computing? What does that 0:07:30.640 --> 0:07:31.120 word mean? 0:07:31.520 --> 0:07:34.840 Yeah, so you can go down the physics way and 0:07:34.880 --> 0:07:37.840 talk about supersition and entanglement, which we can go in later, 0:07:37.920 --> 0:07:39.720 but I actually feel it's a bit of a distraction. 0:07:40.600 --> 0:07:45.320 So when you think of classical computers, what they were 0:07:45.560 --> 0:07:48.200 is there were machines that were very good at adding 0:07:48.320 --> 0:07:52.920 numbers together, like simple addition, and they really showed that 0:07:52.960 --> 0:07:56.760 they could add these numbers together really really fast, and 0:07:56.840 --> 0:08:00.440 now with GPUs and other AI accelerators, we can add 0:08:00.440 --> 0:08:04.160 those numbers together in parallel, and so the whole classical 0:08:04.200 --> 0:08:07.960 computing can come down to just arithmetic, just adding numbers together. 0:08:08.640 --> 0:08:12.680 It turns out that there's a math that is with 0:08:12.720 --> 0:08:15.480 the quantum mechanics shown to be true. It's more like 0:08:15.560 --> 0:08:19.440 a group theory type structure. And the way quantum works 0:08:19.600 --> 0:08:22.160 is it has a different math as are primitive and 0:08:22.200 --> 0:08:24.880 if we can exploit that new math and build a 0:08:24.920 --> 0:08:28.000 machine that does it, it allows us to answer different questions. 0:08:28.560 --> 0:08:31.760 And so think of it as a branching from classical 0:08:31.800 --> 0:08:34.480 compute that is very good at adding just numbers together 0:08:34.880 --> 0:08:37.600 to something that allows us to work with an algebra 0:08:38.320 --> 0:08:42.560 that is much much harder to represent with addition. And 0:08:42.600 --> 0:08:45.640 that algebra happens to be the same algebra that defines 0:08:45.800 --> 0:08:49.360 the fundamental equations of nature shirting as equation. So this 0:08:49.440 --> 0:08:52.000 is why you say it computes the same way nature does. 0:08:52.320 --> 0:08:55.600 But there are many other interesting problems. So the way 0:08:55.640 --> 0:08:58.120 I explain it to people is think of it as 0:08:58.640 --> 0:09:03.880 bringing a new permiitive to computer science and allowing us 0:09:04.320 --> 0:09:06.959 to work out to go with it. And I like 0:09:07.000 --> 0:09:10.920 the analogy well, actually, maybe go back. So if you 0:09:10.960 --> 0:09:13.599 went back in time, so we're one hundred years of quantum, 0:09:13.960 --> 0:09:16.040 and you went back in time and you asked, what 0:09:16.200 --> 0:09:19.440 is the foundation is a chemistry or physics? What would 0:09:19.480 --> 0:09:21.800 have probably the scientists of one hundred years ago would 0:09:21.800 --> 0:09:24.120 have said is they would have said, you know, chemistry 0:09:24.160 --> 0:09:27.240 is about the small, physics is about planets and things 0:09:27.280 --> 0:09:32.680 like this. And one hundred years ago when Heisenberg or Einstein, 0:09:32.760 --> 0:09:37.320 all the greats shreading her himself invented quantum mechanics, it 0:09:37.360 --> 0:09:40.840 was this concept that nature is discrete, not continuous. It 0:09:40.880 --> 0:09:44.440 actually brought all the physical sciences together. And now quantum 0:09:44.440 --> 0:09:48.160 mechanics is like it is the foundation of the science. 0:09:49.000 --> 0:09:52.079 And so now what quantum computing is by that analogy 0:09:52.120 --> 0:09:55.079 is computer science. The foundation of the math is coming 0:09:55.080 --> 0:09:58.560 together with the physical science to allow us to compute 0:09:59.000 --> 0:10:01.920 using math that if you were to try to represent 0:10:01.960 --> 0:10:05.400 it with classical computers, it takes exponential time. 0:10:05.840 --> 0:10:09.640 Yeah, and it was a classical computer an expense in 0:10:09.679 --> 0:10:11.920 a way that someone is well informed as I am 0:10:11.960 --> 0:10:15.960 can understand it. A customer computer works primarily on problems 0:10:15.960 --> 0:10:19.440 that can be easily represented in numerical form in numbers. Yes, 0:10:19.760 --> 0:10:22.480 Quantum allows you to step outside to a class of 0:10:22.480 --> 0:10:26.880 problems that don't necessarily have a simple numerical representation. 0:10:27.120 --> 0:10:31.880 Yeah. And so imagine I got some medicine or or 0:10:31.920 --> 0:10:34.960 some set of operation, but call it A, and I 0:10:35.040 --> 0:10:38.320 then follow it by a different operation B if A 0:10:38.480 --> 0:10:41.720 followed by B gave a different answer than B first 0:10:41.880 --> 0:10:45.000 followed by A. So in mathematics we call that commuting. 0:10:45.200 --> 0:10:47.800 But like you can think of a correlation there one 0:10:48.000 --> 0:10:50.920 one gives you a different outcome to the other. That 0:10:51.000 --> 0:10:55.080 means there's an algebra behind it. That representing that algebra 0:10:55.160 --> 0:10:59.600 traditionally on classical computers is really really hard, whereas that algebra, 0:11:00.040 --> 0:11:02.360 if we can get creative, we can come up with 0:11:02.400 --> 0:11:05.280 ways of representing that math. So we step as you say, 0:11:05.520 --> 0:11:09.320 we step out aside of the simple math to a 0:11:09.360 --> 0:11:12.760 new math to allow us to calculate interesting problems. 0:11:13.200 --> 0:11:16.880 So quite in a sense, compliments it doesn't replace judicial. 0:11:17.040 --> 0:11:17.400 That's good. 0:11:18.080 --> 0:11:19.840 I think this is one of the this is you're 0:11:20.000 --> 0:11:22.760 exactly on is people think quantum is going to be 0:11:22.840 --> 0:11:25.920 replacing classical If your problem is good at adding numbers together, 0:11:26.200 --> 0:11:29.640 you should just keep using classical computers. I think the 0:11:29.679 --> 0:11:33.000 future is going to be heterogeneous accelerators, and it will 0:11:33.040 --> 0:11:36.160 definitely have quantum as one. But in some sense, the 0:11:36.200 --> 0:11:39.319 next generation of superstars are going to be those applied 0:11:39.360 --> 0:11:43.320 mathematicians that know, how do I write a problem using 0:11:43.360 --> 0:11:46.760 the simple math of classical computers or the more complicated 0:11:46.800 --> 0:11:50.480 math for quantum computers, and how do I actually iterate 0:11:50.559 --> 0:11:53.280 between them? And things like this. This is where I 0:11:53.280 --> 0:11:56.079 think the next generation of students are going to come 0:11:56.160 --> 0:11:58.080 up with much more novel ideas. I can give you 0:11:58.120 --> 0:12:00.600 examples of what we want to do on quantum, but like, 0:12:00.640 --> 0:12:06.040 you're giving them a fundamental, foundational new thing, and so 0:12:06.200 --> 0:12:10.040 I'm optimistic that will do much better jobs than my generation. 0:12:10.160 --> 0:12:12.679 Well, yeah, we're to get to some of the albums 0:12:12.720 --> 0:12:15.160 in a moment. But I wanted you to the most 0:12:15.280 --> 0:12:18.120 kind of down that you said as a kid, you 0:12:18.240 --> 0:12:19.880 thought you might want to be a mechanic because you'd 0:12:19.880 --> 0:12:23.600 like to build things. Describe to me what it takes 0:12:23.640 --> 0:12:26.959 to build a quantum computer, Like, what are you doing 0:12:27.040 --> 0:12:28.880 that's different from building a classical computer. 0:12:29.120 --> 0:12:31.760 Yeah, so maybe I'll give you analogy and then I'll 0:12:31.760 --> 0:12:35.319 go in so the way classical computers, we've got them 0:12:35.360 --> 0:12:40.480 to get to smaller and smaller sizes like five seven animeters, 0:12:40.520 --> 0:12:45.480 five animeters and things is actually inventing material to kill 0:12:45.559 --> 0:12:50.200 quantum effects. So you actually put dielectrics and other things 0:12:50.240 --> 0:12:53.200 in there to kill the quantum tunneling effects, and you 0:12:53.280 --> 0:12:58.120 want them to behave more classically. In the quantum world. 0:12:58.240 --> 0:13:00.400 You want to get rid of all the classical effects, 0:13:00.920 --> 0:13:03.319 so you want to get rid of the ability of 0:13:03.360 --> 0:13:06.160 the cubits to interact with the environment. And in the 0:13:06.760 --> 0:13:09.920 sort of technical world we call it this quantum conflict. 0:13:10.400 --> 0:13:12.680 The more ways you want to control the quantum computer, 0:13:13.240 --> 0:13:16.480 you open it up to interacting with everything else, like 0:13:16.600 --> 0:13:20.080 interacting with its environment. So the biggest challenge has always 0:13:20.120 --> 0:13:24.560 been how do we give more control but don't bring 0:13:24.600 --> 0:13:27.920 in other sources of noise. So I want to be 0:13:27.960 --> 0:13:31.160 able to do gates on the cubit, but I don't 0:13:31.160 --> 0:13:34.280 want it to decohere. I want to couple the cubits, 0:13:34.559 --> 0:13:36.880 but I don't want them to couple to other things. 0:13:37.360 --> 0:13:41.640 So the hardest challenge is the energy inside the cubits 0:13:41.800 --> 0:13:44.000 is a nine gigahertz, and if your tames that by 0:13:44.160 --> 0:13:47.960 HBO tend to the neggive thirty four with nine, you're 0:13:48.000 --> 0:13:51.160 at a tender the negative twenty like three or something 0:13:51.240 --> 0:13:54.640 in energy. That's a tiny amount of energy. So you're 0:13:54.679 --> 0:13:57.840 trying to have a tiny, tiny amount of energy to control, 0:13:58.640 --> 0:14:01.679 and you don't want that to interact with anything. So 0:14:01.720 --> 0:14:04.360 you have to cool them down, you have to isolate them, 0:14:04.640 --> 0:14:07.880 and you have to make the quantum effects dominate over 0:14:07.920 --> 0:14:09.000 the classical effects. 0:14:09.720 --> 0:14:13.480 So practically, if I'm trying to do that right now, 0:14:13.600 --> 0:14:14.640 how big are these machines? 0:14:15.000 --> 0:14:17.400 So the cubits themselves are not that big, So the 0:14:17.480 --> 0:14:22.120 cubits themselves are like a few microns. But yeah, most 0:14:22.160 --> 0:14:24.520 of the size so you can see some of our 0:14:24.760 --> 0:14:26.760 I got the pleasure of showing you around to one 0:14:26.760 --> 0:14:28.920 of the machines in Yorktown. You saw that they're like 0:14:29.560 --> 0:14:32.440 twenty foot by twenty foot in size. Most of that 0:14:33.040 --> 0:14:37.320 is all that equipment to isolate the cubit chip, which 0:14:37.400 --> 0:14:40.000 is only a few millimeters when you put it together 0:14:40.400 --> 0:14:43.880 from the rest of the environment. We will, as we 0:14:43.960 --> 0:14:47.560 get better at that, miniaturize all the isolation. But that's 0:14:47.760 --> 0:14:51.320 cooling it down to a few milli calvin, so about 0:14:51.320 --> 0:14:55.080 a thousand times colder than outer space. It's isolating the 0:14:55.280 --> 0:14:58.800 noise on any electrical signal so that no noise from 0:14:58.840 --> 0:15:02.240 the outside world gets into the system. And so that's 0:15:02.240 --> 0:15:05.680 a lot of isolators, filters and things like that that 0:15:05.720 --> 0:15:08.520 we've had to invent to allow us to make the 0:15:08.600 --> 0:15:10.320 quantum properties of this chip go. 0:15:10.720 --> 0:15:13.400 It's like the Princess and the pea. Mounds and mounds 0:15:13.440 --> 0:15:17.040 and mounds of mattresses trying to isolate the impact of 0:15:17.080 --> 0:15:18.280 this little thing. 0:15:18.320 --> 0:15:20.640 And that maybe that's the best way to describe it. Yeah, 0:15:20.640 --> 0:15:23.600 and you've got to keep it really really prestige. 0:15:23.880 --> 0:15:25.760 But that when you show me so in the in 0:15:25.840 --> 0:15:29.560 the lobby of the Watson Research Center in New Yorktown, 0:15:29.760 --> 0:15:32.640 which by the way, is just the coolest building. It's 0:15:32.640 --> 0:15:37.440 like a it's like a modernist it's awesome master piece. Anyway, 0:15:37.600 --> 0:15:41.120 in the lobby there's there are these is it two machines. 0:15:41.240 --> 0:15:44.640 It's it's inside a container that has three machines. 0:15:44.720 --> 0:15:47.920 Three machines. So what can you can you tell me 0:15:47.960 --> 0:15:50.280 what would one of those machines cost to build? Right now? 0:15:50.920 --> 0:15:55.080 So typically we put them together in a way where 0:15:55.160 --> 0:15:58.280 we upgrade them because we want to as I as 0:15:58.280 --> 0:16:00.760 I was talking about before, one the things we want 0:16:00.800 --> 0:16:04.200 to do is always get algorithms done on our machines, 0:16:04.880 --> 0:16:07.920 and I've got a roadmap of building bigger and bigger machines. 0:16:08.400 --> 0:16:11.880 So usually one of those quantum processes today is out 0:16:11.920 --> 0:16:16.000 of date in six months. So we want to build 0:16:16.040 --> 0:16:20.000 this future of computing that leverages quantum computing where every 0:16:20.080 --> 0:16:25.760 six months we've outdated a quantum processor. Eventually, hopefully we 0:16:25.800 --> 0:16:28.840 get to a point where it's like stable and it 0:16:28.840 --> 0:16:32.200 can be many years operating. But we want to get 0:16:32.360 --> 0:16:35.480 as large a quantum computer in the hands of people 0:16:35.560 --> 0:16:37.720 to explore the math as possible to come up with 0:16:37.760 --> 0:16:40.640 those new algorithms. So we've had a philosophy of having 0:16:40.680 --> 0:16:44.480 them open, working with universities and things like that. So 0:16:44.480 --> 0:16:46.680 to answer a question of costs, yes, there's cost in 0:16:46.760 --> 0:16:49.880 building the system, but we are operating in them much 0:16:49.920 --> 0:16:52.880 more in a service model where people pay to use 0:16:52.920 --> 0:16:56.720 the machine because we have to continuously calibrate it and 0:16:56.760 --> 0:17:01.920 operate it and so depending on very different things. Professors, 0:17:01.960 --> 0:17:04.480 we have a credits program where they get free access 0:17:05.119 --> 0:17:08.639 some universities and enterprises, they can buy premium access and 0:17:08.680 --> 0:17:11.960 get more access. So think of not like a cost 0:17:12.040 --> 0:17:15.000 of it, because it's almost like a continuum. I want 0:17:15.080 --> 0:17:18.040 to make sure that the best quantum processors that I 0:17:18.080 --> 0:17:21.720 can build get in the hands of students and professors 0:17:21.760 --> 0:17:25.000 and interested enterprises that want to explore these machines as 0:17:25.040 --> 0:17:29.800 fast as possible. And typically every six months we upgrade it. Yeah, 0:17:29.960 --> 0:17:33.920 you don't start over, you upgrade. We upgrade various different pieces, 0:17:33.960 --> 0:17:38.800 the processor, the electronics. Some upgrades are just simply replaced 0:17:38.840 --> 0:17:42.920 the processor. But as an example, I think many people 0:17:42.920 --> 0:17:45.520 have probably seen photos of quantum computers and you see 0:17:45.520 --> 0:17:49.000 this scary thing with all these wires hanging down, as 0:17:49.119 --> 0:17:51.600 I've referred to as the chandelier, and it's got all 0:17:51.600 --> 0:17:54.840 these wires with loops and things like that. They're called 0:17:55.080 --> 0:17:57.879 co x cables. When we first put the quantum computer 0:17:57.960 --> 0:18:00.960 on the cloud in twenty sixteen, you could probably only 0:18:01.000 --> 0:18:05.439 fit about fifty cubits inside one cryostat. We've had to 0:18:05.560 --> 0:18:08.480 upgrade all those cables so that we can fit around 0:18:08.560 --> 0:18:11.359 one thousand. I want to get to three thousand, and 0:18:11.400 --> 0:18:14.639 that's about miniaturizing. So to answer your question, an upgrade, 0:18:14.680 --> 0:18:17.840 it depends. It can be just the processor or it 0:18:17.840 --> 0:18:20.920 can be the complete insides. And we're actually in our 0:18:20.960 --> 0:18:24.800 third generation of our electronics to control the systems, to 0:18:24.880 --> 0:18:29.440 make them faster, less noise. Internally, We've got exciting results 0:18:29.520 --> 0:18:33.280 of going to something like cold cryocemos. So you can 0:18:33.320 --> 0:18:36.920 bring down the cost in terms of energy of running 0:18:36.960 --> 0:18:40.800 these quantum computers almost to negligible, and you could imagine 0:18:40.800 --> 0:18:44.480 future quantum computers. I'm not going to require much energy 0:18:44.480 --> 0:18:48.520 to run, so unlike classical compute that requires lots of energy. 0:18:48.840 --> 0:18:51.080 The biggest machines that we envision is only in the 0:18:51.160 --> 0:18:55.119 few megawatts. But we have to upgrade to future controls 0:18:55.160 --> 0:18:59.640 that use less energy. So it depends it's my long 0:18:59.720 --> 0:19:03.199 end short answer to how it upgrades, and it depends 0:19:03.200 --> 0:19:03.760 on what it is. 0:19:04.040 --> 0:19:06.879 The only observation that I felt I was capable of 0:19:06.880 --> 0:19:10.960 making when you showed me the quantum machine is it's gorgeous. 0:19:11.440 --> 0:19:12.120 I look at art. 0:19:12.440 --> 0:19:15.959 I've always believed that, and I think that there's an 0:19:15.960 --> 0:19:19.080 IBM saying good design is good business. But we've always 0:19:19.400 --> 0:19:23.760 taken pride in making sure what we build. I don't know, 0:19:23.920 --> 0:19:26.600 I feel if you're going to build something that is 0:19:27.119 --> 0:19:30.720 new that can change, you should take the time to 0:19:30.880 --> 0:19:32.600 make sure it looks and feels good. 0:19:32.760 --> 0:19:35.520 Will you donated to MoMA when you're through with that 0:19:35.560 --> 0:19:40.879 particular actually I think we just put an old version 0:19:40.920 --> 0:19:44.880 of one of our insights with the United Airlines and 0:19:45.000 --> 0:19:47.560 the AAPS, which is the American Physical Society, and the 0:19:47.640 --> 0:19:48.640 University of Chicago. 0:19:48.920 --> 0:19:51.800 There's a replica right now. If you fly into one 0:19:51.800 --> 0:19:54.840 of the terminals in Chicago, you can walk and see one. 0:19:55.160 --> 0:19:57.719 Oh really, yeah, well the most advanced thing at oh air. 0:19:57.800 --> 0:20:01.440 I'm sure probably, but yeah, hopefully. I think, yeah, we're 0:20:01.480 --> 0:20:04.439 open to that. But I yeah, I appreciate that you 0:20:04.520 --> 0:20:08.000 love the design. I was beautiful. So I last week 0:20:08.040 --> 0:20:11.720 I interviewed for another episode of Smart Tox your CEO, 0:20:12.000 --> 0:20:16.080 Irvin Kushner, And when we got to the quantum question. 0:20:16.200 --> 0:20:20.840 I mean, he's always alliant and brilliant, and but quantum, 0:20:20.840 --> 0:20:24.280 he's like lit up. I mean right in thinking that 0:20:24.359 --> 0:20:29.280 IBM is much more invested in quantum than anybody else. 0:20:29.320 --> 0:20:32.600 Is that a fair statement? Oh yeah, most definitely. Why 0:20:32.359 --> 0:20:35.120 Why did IBM choose to kind of make this such 0:20:35.119 --> 0:20:35.760 a priority. 0:20:36.119 --> 0:20:39.119 So when I took to the history of the physics side, 0:20:40.000 --> 0:20:42.960 there's this interesting thing in the history of computing. So 0:20:43.119 --> 0:20:47.760 we build computer classical computers today using bits and see moss, 0:20:47.760 --> 0:20:50.560 and they consume energy. Do you know that there is 0:20:50.600 --> 0:20:54.480 a way in classical where you can actually compute without 0:20:54.560 --> 0:20:58.040 using energy. It's called reversal computing. Turns out to be 0:20:58.240 --> 0:21:03.960 a terrible idea, not practical to build, but IBM investigated 0:21:04.000 --> 0:21:07.720 that with Ralph Laura and Charlie Bennett early on, and 0:21:07.760 --> 0:21:11.880 they proved the concept that reversible computing. The first use 0:21:12.000 --> 0:21:15.560 of quantum information theory. One of the first actually was 0:21:15.600 --> 0:21:19.560 from IBM. When I did my PhD, I remember actually 0:21:19.640 --> 0:21:23.600 picking up this paper on quantum teleportation and seeing IBM 0:21:23.600 --> 0:21:25.760 written there, and at the time I remember thinking that 0:21:25.880 --> 0:21:28.240 they make PCs. Well, what the hell are they doing 0:21:29.000 --> 0:21:33.479 this foundational paper on quantum teleportation? Why are they doing it? 0:21:33.920 --> 0:21:36.760 So to answer your question, actually, IBM was the first 0:21:37.400 --> 0:21:41.560 in quantum information science because it's the fundamental of computation. 0:21:42.200 --> 0:21:46.160 Can we actually come up with compute that we can 0:21:46.200 --> 0:21:49.800 go beyond the classical So way before anyone was talking 0:21:49.840 --> 0:21:53.640 about it, they were doing fundamental theory. And then as 0:21:53.680 --> 0:21:56.400 we've built it, we've always When I first came there, 0:21:56.440 --> 0:21:59.880 the experimental team was small. In twenty eleven, we've had 0:21:59.880 --> 0:22:05.240 a small team that we're focusing on single cubitts coupling 0:22:05.240 --> 0:22:08.320 in them. I think in twenty twelve was the first 0:22:08.359 --> 0:22:12.840 time we showed really good two Cuba gates and no 0:22:12.920 --> 0:22:16.280 one was talking about quantum computing then. And then I 0:22:16.320 --> 0:22:20.479 remember in about twenty sixteen I said to actually Arvin 0:22:20.560 --> 0:22:24.040 was the director of research, then can we actually put 0:22:24.359 --> 0:22:27.760 our quantum computer on the cloud? Well that's probably twenty fifteen, 0:22:28.080 --> 0:22:32.040 and it was always supporting that. So as we've done 0:22:32.080 --> 0:22:35.040 more and more we've been able to do it. It's 0:22:35.080 --> 0:22:38.960 had this program going now, I agree, is very visible, 0:22:39.400 --> 0:22:42.440 like because we're in this scaling phase and so we're 0:22:42.480 --> 0:22:46.520 invested to keep scaling it and to get why is 0:22:47.080 --> 0:22:50.840