1 00:00:05,207 --> 00:00:11,950 Today we are heading straight into the heart of high-energy density physics. 2 00:00:11,950 --> 00:00:19,432 The place where lasers crush matter, astrophysics meets the lab, and Bayesian inference becomes indispensable. 3 00:00:19,432 --> 00:00:32,365 My guest is Ethan Smith, a PhD candidate at the University of Rochester, working at the intersection of plasma spectroscopy, diagnostics, technology, and Bayesian data. 4 00:00:32,365 --> 00:00:33,445 analytics. 5 00:00:33,565 --> 00:00:45,205 studies what happens when you use some of the world's most powerful lasers to squeeze matter to extreme conditions, the same conditions you would find inside planets, stars, or 6 00:00:45,205 --> 00:00:45,885 supernovae. 7 00:00:45,885 --> 00:00:57,105 This is an episode for anyone who loves physics, who loves patient methods, or who just wants to hear how scientific discovery actually happens behind the scenes. 8 00:00:57,105 --> 00:01:01,982 This is Learning Vision Statistics, episode 146, recorded. 9 00:01:01,982 --> 00:01:04,640 October 7, 2025. 10 00:01:21,043 --> 00:01:30,164 Welcome to Learning Bayesian Statistics, a podcast about Bayesian inference, the methods, the projects, and the people who make it possible. 11 00:01:30,164 --> 00:01:32,364 I'm your host, Alex Andorra. 12 00:01:32,364 --> 00:01:35,850 You can follow me on Twitter at alex-underscore-andorra. 13 00:01:35,850 --> 00:01:36,670 like the country. 14 00:01:36,670 --> 00:01:40,902 For any info about the show, learnbasedats.com is Laplace to be. 15 00:01:40,902 --> 00:01:48,084 Show notes, becoming a corporate sponsor, unlocking Bayesian Merge, supporting the show on Patreon, everything is in there. 16 00:01:48,084 --> 00:01:49,994 That's learnbasedats.com. 17 00:01:49,994 --> 00:02:01,028 If you're interested in one-on-one mentorship, online courses, or statistical consulting, feel free to reach out and book a call at topmate.io slash Alex underscore and Dora. 18 00:02:01,028 --> 00:02:02,538 See you around, folks. 19 00:02:02,538 --> 00:02:04,429 and best patient wishes to you all. 20 00:02:04,429 --> 00:02:11,524 And if today's discussion sparked ideas for your business, well, our team at Pimc Labs can help bring them to life. 21 00:02:11,524 --> 00:02:13,987 Check us out at pimc-labs.com. 22 00:02:17,640 --> 00:02:21,921 Ethan Smith, welcome to Learning Basian Statistics. 23 00:02:21,921 --> 00:02:22,681 Thank you for having me. 24 00:02:22,681 --> 00:02:23,862 Thrill the beer. 25 00:02:24,402 --> 00:02:25,672 Yeah, it's great to have you here. 26 00:02:25,672 --> 00:02:31,284 Thanks a lot to JJ Ruby for putting us in contact. 27 00:02:31,284 --> 00:02:37,706 I hear you guys are doing some fun stuff at Rochester, doing some physics things. 28 00:02:37,706 --> 00:02:39,886 We'll definitely talk about that. 29 00:02:39,886 --> 00:02:46,794 First, both your first and last name are very challenging for me to say with an English accent as a French 30 00:02:46,794 --> 00:02:49,416 man, because I want to say Ethan Smith. 31 00:02:49,416 --> 00:02:53,068 Uh, and so it's very hard for me to say Ethan Smith. 32 00:02:53,068 --> 00:02:56,269 It's like twice to the age in a row. 33 00:02:56,269 --> 00:02:58,650 That's like, making my life hard. 34 00:02:59,251 --> 00:03:00,182 apologize for that. 35 00:03:00,182 --> 00:03:00,362 Yeah. 36 00:03:00,362 --> 00:03:01,752 That's, that's my bad. 37 00:03:02,912 --> 00:03:08,116 I do like, I do love the name Ethan that that sounds really good in English in French. 38 00:03:08,116 --> 00:03:13,559 I don't like it's etan, but Ethan sounds really like very classy. 39 00:03:13,659 --> 00:03:14,929 Yes, I agree. 40 00:03:14,929 --> 00:03:16,790 Unfortunately, it's a very common name. 41 00:03:16,790 --> 00:03:23,993 uh There is another Ethan Smith in my field and we go to lot of the same conferences and it causes a lot of issues. 42 00:03:23,993 --> 00:03:24,693 Yeah. 43 00:03:24,693 --> 00:03:25,433 Yeah. 44 00:03:25,433 --> 00:03:26,534 I am not surprised. 45 00:03:26,534 --> 00:03:36,598 will confess that while I was preparing for the episode, um looking you up on the internet was not easy because there was a lot of Ethan Smithies. 46 00:03:36,598 --> 00:03:42,130 And if you input Ethan Smiths Rochester, um 47 00:03:42,130 --> 00:03:47,104 There is actually one Ethan Smith who unfortunately died in a plane crash. 48 00:03:47,104 --> 00:03:49,105 Yeah, from Rochester, Minnesota. 49 00:03:49,146 --> 00:03:51,007 Yeah, exactly. 50 00:03:51,308 --> 00:03:52,408 I know him. 51 00:03:53,259 --> 00:03:53,790 damn. 52 00:03:53,790 --> 00:03:54,450 Yeah. 53 00:03:54,450 --> 00:03:55,761 So I was reading up on that. 54 00:03:55,761 --> 00:03:58,274 was like, damn, this is a very sad story. 55 00:03:58,274 --> 00:03:59,454 Not me though. 56 00:03:59,615 --> 00:04:01,176 Different Ethan. 57 00:04:01,176 --> 00:04:02,577 I'm still alive. 58 00:04:02,637 --> 00:04:04,498 Yeah, thankfully. 59 00:04:05,760 --> 00:04:08,442 So yeah, like actually, let's say talking about you. 60 00:04:08,442 --> 00:04:10,043 Let's start talking about you. 61 00:04:10,043 --> 00:04:11,264 um 62 00:04:11,526 --> 00:04:17,979 So always when I, when I start asking to the guests says, think, you know, because, uh, you listened to the show, you told me. 63 00:04:17,979 --> 00:04:23,341 And so I very, very grateful for that. 64 00:04:24,081 --> 00:04:30,134 and so I know, you know, you do a lot of cool stuff, um, listeners don't, you don't know yet. 65 00:04:30,134 --> 00:04:33,935 So yeah, just give us the, the origin story. 66 00:04:33,935 --> 00:04:38,027 What are you doing nowadays and how did you end up working on that? 67 00:04:38,107 --> 00:04:38,477 Yeah. 68 00:04:38,477 --> 00:04:41,188 So I'm getting, I'm finishing up my PhD. 69 00:04:41,240 --> 00:04:42,581 at the University of Rochester. 70 00:04:42,581 --> 00:04:58,036 um Specifically, I'm in the field of high energy density physics, uh which is sort of a uh niche subfield uh in physics where we use extremely high powered lasers to compress matter 71 00:04:58,036 --> 00:05:01,839 to some extremely distressing conditions. 72 00:05:01,980 --> 00:05:06,620 And it turns out when you compress matter with lasers, can... 73 00:05:06,620 --> 00:05:14,944 you know, recreate some of the states that you would find, for example, at the center of giant planets, or even in some cases at the center of stars. 74 00:05:14,944 --> 00:05:27,549 And so that lets you recreate these systems in the laboratory and directly study how they behave uh sort of at a much smaller scale than you can with observational astronomy. 75 00:05:27,549 --> 00:05:33,712 And so that lets you learn a lot about the material properties of these astrophysical objects. 76 00:05:33,956 --> 00:05:42,259 And it also lets you create some just really interesting states of matter that you can't find anywhere else on earth or even sometimes in the universe. 77 00:05:42,259 --> 00:05:50,563 um And so, we create these very hot, dense states of matter for a billionth of a second. 78 00:05:50,563 --> 00:06:02,328 And what the focus of my work is on is understanding how do you make a measurement, one, how do you make a measurement of a system that that extreme and exists for a very short 79 00:06:02,328 --> 00:06:03,366 amount of time? 80 00:06:03,366 --> 00:06:05,037 And two, how do you interpret those measurements? 81 00:06:05,037 --> 00:06:08,920 So measurements that we get out of these systems are often very integrated. 82 00:06:08,920 --> 00:06:17,711 ah know, there's this inverse problem of trying to take up, know, you have an image of this plasma that you've created and you have to try and back out what was the temperature 83 00:06:17,711 --> 00:06:23,129 and density of this, you know, miniature sun we've created in the laboratory. 84 00:06:23,530 --> 00:06:32,956 And so that's sort of the thrust of my PhD is trying to use data science techniques, including of course, know, Bayesian inference. 85 00:06:33,007 --> 00:06:40,523 to take this set of uh information-rich but complicated measurements and understand what's happening in these experiments. 86 00:06:40,523 --> 00:06:43,826 um And so we've learned a lot. 87 00:06:43,826 --> 00:06:45,007 uh J.J. 88 00:06:45,007 --> 00:06:53,414 Ruby, who was on the show before, was sort of got the ball rolling with thinking about using Bayesian inference in this context. 89 00:06:53,414 --> 00:06:58,968 And I've sort of inherited a lot of his work and carried it on to the next generation of grad student. 90 00:06:58,968 --> 00:07:01,520 And then I'm sure some grad student will come after me 91 00:07:01,590 --> 00:07:03,542 and carry it even further, hopefully. 92 00:07:03,542 --> 00:07:08,566 Oh, I didn't answer the second part of your question, which is how did I end up doing this? 93 00:07:08,566 --> 00:07:16,072 Which is a fair question, because it's very, when I tell people what I do, they're like, how did you even know that you could do that? 94 00:07:16,072 --> 00:07:20,676 ah And so I grew up in Rochester, New York, not Minnesota. 95 00:07:20,676 --> 00:07:26,431 ah And at the University of Rochester, we have two of the largest lasers in the world. 96 00:07:26,431 --> 00:07:30,792 They're actually the largest lasers at any academic facility anywhere. 97 00:07:30,792 --> 00:07:33,884 And those are the Omega 60 and Omega EP lasers. 98 00:07:33,884 --> 00:07:39,839 They're very impressive if you ever are in town and you want to come, you know, take a tour. 99 00:07:39,839 --> 00:07:43,241 Each of these lasers is about the size of a football field. 100 00:07:43,241 --> 00:07:51,687 uh And, you know, we take these massive lasers and focus them down onto a, tiny point. 101 00:07:51,687 --> 00:07:54,089 uh And it's very impressive. 102 00:07:54,089 --> 00:07:56,951 And so, you know, I went to school in the area. 103 00:07:56,951 --> 00:07:58,372 I went and toured. 104 00:07:58,372 --> 00:08:04,772 these facilities as part of my undergraduate research program. 105 00:08:04,812 --> 00:08:06,652 And so I was like, oh, that's really cool. 106 00:08:06,652 --> 00:08:07,752 I didn't really think anything of it. 107 00:08:07,752 --> 00:08:10,372 Like, of course this giant laser would be in Rochester. 108 00:08:10,552 --> 00:08:12,152 Yeah, like that makes sense. 109 00:08:12,152 --> 00:08:18,972 And then I was, you know, I decided to apply to grad school and I was, you know, considering the University of Rochester, but I didn't want to really stay. 110 00:08:18,972 --> 00:08:20,972 I wanted to go, you know, go somewhere else. 111 00:08:20,972 --> 00:08:27,772 But then I met Rip Collins, who is my current advisor at a conference in Fort Lauderdale of all places. 112 00:08:28,024 --> 00:08:37,739 And he told me about the cool science that they were doing with those lasers, know, all of the high energy density physics, the planetary science that you can do with these things. 113 00:08:37,739 --> 00:08:48,085 And that really, you know, sort of convinced me that, yeah, like this is, you know, this is even cooler than you might suspect by the fact that there's these football field size 114 00:08:48,085 --> 00:08:48,295 lasers. 115 00:08:48,295 --> 00:08:51,077 You can actually do really interesting science with these. 116 00:08:51,077 --> 00:08:57,732 And that sort of inspired me to go down that road and then, you know, ended up working with him for the past, ooh. 117 00:08:57,732 --> 00:09:00,032 on five years now. 118 00:09:00,172 --> 00:09:08,572 So it's been an experience, been a wild ride, doing a lot of laser experiments. 119 00:09:09,772 --> 00:09:10,212 Yeah. 120 00:09:10,212 --> 00:09:10,572 Okay. 121 00:09:10,572 --> 00:09:18,032 So you were attracted by the big shiny laser, basically. 122 00:09:18,032 --> 00:09:20,132 This was an important role. 123 00:09:20,352 --> 00:09:25,840 I'll say the lasers themselves are impressive, but I didn't have any interest in... 124 00:09:25,840 --> 00:09:38,051 using them because I wasn't super motivated by the main uh mission of the laboratory for laser energetics, which is where these football field sized lasers are, is the pursuit of 125 00:09:38,051 --> 00:09:54,154 fusion energy, which is great and I think a worthwhile endeavor, but it ah didn't move me because it's just about uh effectively making as many neutrons as you can from fusion. 126 00:09:54,312 --> 00:09:56,313 And you just want that number to go up. 127 00:09:56,313 --> 00:10:00,635 And that to me always felt like more of an engineering problem than a physics problem. 128 00:10:00,635 --> 00:10:09,559 ah And so when you go and tour these facilities, a lot of emphasis is placed on this fusion uh arm of the program. 129 00:10:09,720 --> 00:10:14,062 And so what really inspired me was the physics of it all, right? 130 00:10:14,963 --> 00:10:24,197 To make the number of neutrons go up, you need to really understand the physics of these very extreme systems that exist at millions or billions of atmospheres of pressure. 131 00:10:24,221 --> 00:10:32,595 And that's a very interesting and difficult problem from a theoretical standpoint and from an experimental standpoint and statistical analysis. 132 00:10:32,595 --> 00:10:34,505 And so that's what really got me interested. 133 00:10:34,505 --> 00:10:43,809 And then, you know, I met JJ through that and he got me interested in Bayesian inference, um which is an even more interesting subset of an interesting problem. 134 00:10:43,809 --> 00:10:51,302 So I would say the Bayesian inference at the end of the day is what got me more than the giant lasers. 135 00:10:51,522 --> 00:10:52,163 Hmm. 136 00:10:52,163 --> 00:10:52,653 Okay. 137 00:10:52,653 --> 00:10:53,713 Interesting. 138 00:10:54,003 --> 00:10:58,986 So thanks JJ first and second. 139 00:10:58,986 --> 00:11:07,620 oh So basically the, like the physics was here from a very, from the very beginning, right? 140 00:11:07,620 --> 00:11:17,716 So, um, yeah, can you also share your path into physics and, um, how did you first get interested in that? 141 00:11:17,716 --> 00:11:23,539 What was also your undergraduate training like and how do you think that shaped your 142 00:11:23,805 --> 00:11:25,886 your research direction today. 143 00:11:25,886 --> 00:11:27,096 Yeah, sure. 144 00:11:27,096 --> 00:11:30,298 So I took a pretty straightforward path. 145 00:11:30,658 --> 00:11:32,799 you know, took science classes in high school. 146 00:11:32,799 --> 00:11:33,899 I liked it. 147 00:11:34,240 --> 00:11:37,321 I came from a very liberal arts family. 148 00:11:37,321 --> 00:11:39,512 So my parents are both teachers. 149 00:11:39,512 --> 00:11:43,103 My dad taught history and my mom's an English professor. 150 00:11:43,163 --> 00:11:48,525 And so, you know, sort of teaching is the family business in my family. 151 00:11:48,786 --> 00:11:51,807 And so I had always assumed I would go teach. 152 00:11:51,847 --> 00:11:56,490 But I knew I didn't want to do English or history or any of these liberal arts things because I hated writing. 153 00:11:56,490 --> 00:12:01,192 uh And so I wanted to do, you know, I liked science and math a lot more. 154 00:12:01,192 --> 00:12:04,894 um And so my family was always like, well, what are you going to do with that? 155 00:12:04,894 --> 00:12:06,985 What are you going to, you know, a science degree? 156 00:12:06,985 --> 00:12:08,316 What use is that? 157 00:12:08,467 --> 00:12:11,838 You know, you know, and so I went into undergrad. 158 00:12:11,838 --> 00:12:17,361 went to uh SUNY Geneseo, which is about, you know, 40 minutes outside of Rochester. 159 00:12:17,361 --> 00:12:18,682 So very far from home for me. 160 00:12:18,682 --> 00:12:21,143 ah 161 00:12:21,211 --> 00:12:29,657 I, you know, I had taken physics most recently, uh, as a, know, junior and senior in high school and I really liked it. 162 00:12:29,677 --> 00:12:33,700 So I said, well, we'll keep, we'll keep, you know, stick with it. 163 00:12:33,800 --> 00:12:38,204 And so I majored in it and I really liked in, in, uh, college as an undergrad. 164 00:12:38,204 --> 00:12:41,866 The plan was to go into, um, teach high school teaching. 165 00:12:41,866 --> 00:12:47,290 So I was gonna, you know, enroll in the education school there and then do that. 166 00:12:47,290 --> 00:12:48,883 But then I decided that. 167 00:12:48,883 --> 00:12:49,933 That's not what I wanted to do. 168 00:12:49,933 --> 00:12:58,869 And so I started um doing research, undergraduate research with a physics professor at Geneseo, Kurt Fletcher, shout out. 169 00:12:59,090 --> 00:13:06,735 And um that's, I mean, that's where I really learned, you know, that I could do science and that doing science is really fun. 170 00:13:06,735 --> 00:13:11,748 So that was a much more hands-on research project that I did in undergrad, turning wrenches. 171 00:13:11,748 --> 00:13:18,963 And we were building uh a spectrometer for, you know, ion backscattering studies or whatever. 172 00:13:19,317 --> 00:13:27,690 And, you know, that's, that's when I fell in love with science and doing science, both from a hands-on perspective and just, you know, thinking about science is, is a really 173 00:13:27,690 --> 00:13:30,711 interesting and rewarding pastime. 174 00:13:30,991 --> 00:13:33,392 And so then I was like, well, maybe I should do more of this. 175 00:13:33,392 --> 00:13:45,357 And so, you know, I applied to grad school, got in, and then I've, you know, been doing it for uh five and a half years now and probably going to do more of it after this. 176 00:13:45,577 --> 00:13:48,858 So I think that's, that's sort of the. 177 00:13:49,306 --> 00:13:50,726 the timeline for me. 178 00:13:50,726 --> 00:13:55,806 It's always just been like, ah, there's probably more science to know, more physics to do. 179 00:13:55,806 --> 00:13:59,538 And so it's been a rewarding process so far. 180 00:14:01,048 --> 00:14:01,558 Yeah, we'll see. 181 00:14:01,558 --> 00:14:02,399 Yeah. 182 00:14:02,399 --> 00:14:08,041 Um, that, makes total sense. 183 00:14:08,041 --> 00:14:13,693 I can, yeah, I can vouch for your, uh, your passion for the, the topic. 184 00:14:13,693 --> 00:14:17,924 I'm also going to bet that you're going to do that for, for a long time. 185 00:14:19,165 --> 00:14:30,850 and so actually what, what you already talked about is that your research lies and the intersection of high energy density physics. 186 00:14:31,098 --> 00:14:33,439 plasma spectroscopy and vision inference. 187 00:14:33,439 --> 00:14:42,053 So could you give us a big picture of what that looks like, of what your current portfolio project is? 188 00:14:42,053 --> 00:14:50,486 Because I'm guessing that lot of listeners are not familiar with high energy density physics and plasma spectroscopy. 189 00:14:50,486 --> 00:14:51,416 That's very fair. 190 00:14:51,416 --> 00:14:57,509 High energy density physics is an interesting one because it exists at the intersection of a lot of fields, right? 191 00:14:57,509 --> 00:15:00,282 Because you have these very hot, dense 192 00:15:00,282 --> 00:15:09,282 balls of plasma, you have plasma science going on, you have laser science because you're using these giant lasers, you have to know how to use those and how to compress things 193 00:15:09,282 --> 00:15:10,402 effectively. 194 00:15:10,782 --> 00:15:13,522 And then of course, you know, how do you make a measurement? 195 00:15:13,522 --> 00:15:21,422 Well, these things get really hot, they emit x-rays, they emit neutrons from fusion reactions, and so you have to understand nuclear physics, you have to understand atomic 196 00:15:21,422 --> 00:15:23,942 physics, and all these things. 197 00:15:24,102 --> 00:15:25,842 And so sort of 198 00:15:25,858 --> 00:15:32,662 At a high level, uh the kind of high energy density physics experiments I work on are spherical implosions. 199 00:15:32,922 --> 00:15:43,648 So you can think of uh the standard high energy density physical experiment is you take a planar foil, a piece of material, metal, and you just whack it on one side of the laser, 200 00:15:43,648 --> 00:15:46,310 drive a shockwave through it, and compress it that way. 201 00:15:46,310 --> 00:15:54,766 And that lets you get up to millions of atmospheres of pressure, which is uh unreasonable conditions. 202 00:15:55,066 --> 00:16:05,611 And it turns out that there's sort of a limit to how hard you can hit something with a laser ah before you just start, you know, dumping all this energy into nonlinear 203 00:16:05,611 --> 00:16:07,332 instabilities and whatnot. 204 00:16:07,332 --> 00:16:17,127 And so a way to get to higher pressure is to do it in convergent geometry, which essentially means you'd take your laser illumination from one side and you illuminate a 205 00:16:17,127 --> 00:16:19,197 spherical target from all sides. 206 00:16:19,318 --> 00:16:20,223 And so now... 207 00:16:20,223 --> 00:16:24,735 Instead of getting compression in one dimension, you're getting compressions in all three dimensions, right? 208 00:16:24,735 --> 00:16:33,690 And so that lets you, know, exponentially amplify the pressures you can reach up to, you know, billions, billions of atmospheres, hundreds of billions of atmospheres even. 209 00:16:33,970 --> 00:16:39,533 And so those, uh those convergent systems are even harder to measure. 210 00:16:39,533 --> 00:16:47,937 At least in a planar geometry, you can, you know, measure something from the backside of your target and maybe get a direct measurement of how fast it's going. 211 00:16:48,177 --> 00:16:49,876 But in these experiments, 212 00:16:49,876 --> 00:16:51,106 You know, it's complete. 213 00:16:51,106 --> 00:16:52,987 All your measurements are completely integrated. 214 00:16:52,987 --> 00:16:56,988 The X-rays that you measure from these have to travel through the target itself. 215 00:16:56,988 --> 00:17:02,049 So you have to already know what temperature and density the target's at to understand those X-rays. 216 00:17:02,189 --> 00:17:12,512 And so the, the sort of, main focus of, of my research is understanding what's happening in that 10 million degree ball of fire. 217 00:17:12,512 --> 00:17:17,174 You know, it's a spherical plasma that's emitting X-rays and neutrons and all this stuff. 218 00:17:17,174 --> 00:17:18,934 How do you understand how you image it? 219 00:17:18,934 --> 00:17:21,665 ah How do you understand the spectra that come out of it? 220 00:17:21,665 --> 00:17:21,906 Right? 221 00:17:21,906 --> 00:17:32,812 If you have bound electrons in there, they're emitting, you know, the characteristic emission like you see, uh I guess in science class, when you look at the neon lights, for 222 00:17:32,812 --> 00:17:34,382 example, the fluorescent lights. 223 00:17:34,382 --> 00:17:45,429 ah But how do those, you know, how do those lines change when you're compressed to 10 billion atmospheres of pressure and you're at 10 million degrees Kelvin? 224 00:17:45,429 --> 00:17:48,690 Like the atomic physics of that are challenging. 225 00:17:48,982 --> 00:17:53,824 But you can measure those transitions and you can learn a lot about the system from that. 226 00:17:53,984 --> 00:17:58,826 And so then the question is, how do you measure these systems? 227 00:17:58,826 --> 00:17:59,947 What do you measure? 228 00:17:59,947 --> 00:18:03,088 And then what measurements give you the most information? 229 00:18:03,088 --> 00:18:13,753 ah And so that's where the sort Bayesian inference comes in and working in an information content sort of paradigm where you take all these measurements, you take X-ray pictures, 230 00:18:13,753 --> 00:18:17,364 you take uh spectrographs of these systems. 231 00:18:17,962 --> 00:18:20,554 What information can you get and what's the sort of max? 232 00:18:20,554 --> 00:18:23,185 There's a certain number of measurements you can make, right? 233 00:18:23,185 --> 00:18:25,606 You're limited by space on the target chamber. 234 00:18:25,606 --> 00:18:38,734 ah And so understanding the information content of each measurement and what's the most effective combination is something that is very non-trivial and something that I'm looking 235 00:18:38,734 --> 00:18:40,074 into actively. 236 00:18:40,134 --> 00:18:46,798 And then the third piece of it is sort of what discoveries can we make with these tools, right? 237 00:18:47,366 --> 00:18:53,029 you know, we have these Bayesian statistics, we have all of these images and spectroscopy. 238 00:18:53,029 --> 00:18:57,252 uh What physics discoveries can we make and, you what can we do? 239 00:18:57,252 --> 00:19:04,695 How does that uh translate to our understanding of astrophysical systems and, you know, the universe at large? 240 00:19:05,116 --> 00:19:16,102 And I think that's really the most interesting part is we build all these really uh complicated tools to all these fancy statistics and whatnot. 241 00:19:16,214 --> 00:19:28,103 uh to sort of peer inside these extreme systems and understand really what's going on, not only in these systems, but in other extreme astrophysical systems. 242 00:19:28,103 --> 00:19:39,661 so there's a lot of interesting discoveries that are coming out of that about atomic physics, about nuclear physics, about what role does radiation play, right? 243 00:19:39,661 --> 00:19:43,493 You're producing an enormous amount of photons. 244 00:19:43,493 --> 00:19:44,252 uh 245 00:19:44,252 --> 00:19:46,883 just because you have this hot, dense ball of gas. 246 00:19:46,883 --> 00:19:51,526 And those, you can produce enough photons to affect the dynamics of the system. 247 00:19:51,526 --> 00:19:55,448 The radiation flux is so large that the system behaves differently. 248 00:19:55,448 --> 00:20:06,773 And understanding that is very non-trivial because a lot of these, you know, systems can't be created anywhere else uh except, you know, here in the lab and understanding that is 249 00:20:06,773 --> 00:20:07,394 very difficult. 250 00:20:07,394 --> 00:20:14,397 So all of this requires some level of, you know, statistical inference. 251 00:20:14,405 --> 00:20:23,310 And that's really what my, the big picture of my work is building the tools to do that statistical inference. 252 00:20:23,991 --> 00:20:32,265 Because you can make the analog very similar uh type of experiment in particle physics at particle colliders. 253 00:20:32,265 --> 00:20:34,917 You have very distressing conditions there as well. 254 00:20:34,917 --> 00:20:40,620 Although those are at order of magnitude higher energy than we can produce with our laser systems. 255 00:20:40,620 --> 00:20:43,081 um 256 00:20:43,323 --> 00:20:55,293 To understand what comes out of those violent particle collisions where you're at, know, GEV, TEV energies requires very complex statistical tools, very high level understanding 257 00:20:55,293 --> 00:21:00,347 of your detector and very good models of those physical systems. 258 00:21:00,428 --> 00:21:08,184 And, you know, they're sort of the particle physics field is a lot further along down this road than we are. 259 00:21:08,184 --> 00:21:12,047 And so we have to sort of, we're starting from a very much lower level. 260 00:21:12,101 --> 00:21:22,487 And so we have to understand uh from a fundamental level, the physics of the system, the detector, and how to do that inference rigorously. 261 00:21:24,111 --> 00:21:27,884 Yeah, this is just fascinating. 262 00:21:27,984 --> 00:21:33,409 So concretely, what does that look like to work with that kind of data? 263 00:21:33,409 --> 00:21:36,461 First, what does your data look like? 264 00:21:36,461 --> 00:21:37,712 Where do they come from? 265 00:21:37,712 --> 00:21:41,816 um What's the size of them? 266 00:21:41,816 --> 00:21:50,943 How does it look like to concretely work on these kind of projects before we dive a bit more into the modeling complexities? 267 00:21:50,943 --> 00:21:51,564 Yeah, sure. 268 00:21:51,564 --> 00:21:53,007 oh 269 00:21:53,007 --> 00:21:54,718 That's a good question. 270 00:21:54,718 --> 00:22:09,426 The data oftentimes looks like a uh dot on a camera, like a little spherical blob, or maybe if you're lucky, a series of spherical blobs that correspond to different times. 271 00:22:09,426 --> 00:22:22,743 um Or if you're even luckier, you could get maybe a streaked continuous record in time with one of these streak cameras that we've developed in the field. 272 00:22:22,961 --> 00:22:32,315 uh And what it looks like is trying to build a sort of generative model of these blobs. 273 00:22:32,315 --> 00:22:36,517 So using our best available knowledge of the physics, right? 274 00:22:36,517 --> 00:22:38,728 We think we understand hydrodynamics. 275 00:22:38,728 --> 00:22:43,940 We think maybe we understand how X-rays are produced in plasmas and how they propagate. 276 00:22:43,940 --> 00:22:52,413 That may or may not be true, but we can take our best available models and try and build 277 00:22:52,441 --> 00:23:01,145 a generative model of how the physics, we think the physics will evolve and how that will manifest in the observed emission. 278 00:23:01,586 --> 00:23:07,549 And then we need to also understand how our detector is constructing that blob. 279 00:23:07,549 --> 00:23:16,304 And so that requires knowing what the resolution is and the sensitivity to different energies of X-rays and as well as the statistics, right? 280 00:23:16,304 --> 00:23:18,395 And that's the most complicated part. 281 00:23:18,395 --> 00:23:22,175 And something that we're really working on is understanding 282 00:23:22,175 --> 00:23:26,886 you know, if you have a given X-ray flux, what signal, what distribution of signals are you going to measure? 283 00:23:26,886 --> 00:23:34,878 And, know, you have Poisson counting statistics for how many photons are absorbed, but then the rest of it is a complete mess because it's so complicated. 284 00:23:34,878 --> 00:23:38,848 And you're trying to make a measurement in a fraction of a billionth of a second. 285 00:23:38,969 --> 00:23:43,520 And so all hell breaks loose basically, the statistics get very complicated. 286 00:23:43,520 --> 00:23:46,991 There's correlations between neighboring data points. 287 00:23:46,991 --> 00:23:49,804 And so understanding that is a really 288 00:23:49,804 --> 00:23:52,804 sort of nuanced and complicated piece of it. 289 00:23:52,804 --> 00:24:02,404 But all of it is just to produce this little, you know, dot on your camera, basically, which has an intensity, it's got a radius, maybe it's got some shape, but it's very simple 290 00:24:02,404 --> 00:24:03,084 data. 291 00:24:03,084 --> 00:24:06,244 We're using technology from the 1800s. 292 00:24:06,244 --> 00:24:10,244 It's like pinhole cameras and geometric optics. 293 00:24:10,404 --> 00:24:13,482 It's not nearly as fancy as some of the... 294 00:24:13,482 --> 00:24:15,717 diagnostics that are being fielded in other fields. 295 00:24:15,717 --> 00:24:17,902 But that's largely because it can't be, right? 296 00:24:17,902 --> 00:24:25,256 It's very hard to have X-ray optics in a compact space because X-rays don't really refract the way that visible light does. 297 00:24:27,286 --> 00:24:30,419 Damn, yeah. 298 00:24:30,419 --> 00:24:34,381 This is fascinating. 299 00:24:35,042 --> 00:24:46,562 So maybe first, can you walk us through the main technical or methodological obstacles you face just even before starting to analyze and model this data? 300 00:24:46,562 --> 00:24:54,959 Like, what's the first step in your workflow when you're working on one of the concrete projects you've named at the beginning of the show? 301 00:24:54,959 --> 00:24:55,299 Sure. 302 00:24:55,299 --> 00:24:56,944 So the first step. 303 00:24:56,944 --> 00:25:03,430 is designing these large scale experiments to be done on these football field size lasers. 304 00:25:03,430 --> 00:25:09,415 And so that work usually starts about a little more than a year out. 305 00:25:09,415 --> 00:25:15,981 You propose an experiment, it gets accepted or rejected or iterated on by the committee. 306 00:25:16,221 --> 00:25:18,724 You get time on the laser, you get usually one day. 307 00:25:18,724 --> 00:25:23,680 ah And so then your day is scheduled for the next fiscal year and you have to 308 00:25:23,680 --> 00:25:24,650 design your experiment. 309 00:25:24,650 --> 00:25:27,481 So you got to say, all right, what's my target going to look like? 310 00:25:27,481 --> 00:25:29,061 How it's the laser pulse going to look like? 311 00:25:29,061 --> 00:25:30,842 How much energy do I want? 312 00:25:31,262 --> 00:25:34,023 What am I looking to actually observe here? 313 00:25:34,023 --> 00:25:37,684 And what measurement is going to give me the most information? 314 00:25:37,804 --> 00:25:40,724 And so that requires a lot of simulations. 315 00:25:40,985 --> 00:25:49,577 So we run these very advanced state of the art uh radiation hydrodynamics codes to sort of predict how the system is going to behave. 316 00:25:49,577 --> 00:25:53,908 It's never right because these systems are really complicated. 317 00:25:54,024 --> 00:25:59,269 And so the codes can never quite capture what's going to happen, but we can get an idea. 318 00:25:59,269 --> 00:26:04,293 uh And then a piece I've really been pushing is trying to do it predictively. 319 00:26:04,293 --> 00:26:16,024 So sort of as a byproduct of this whole Bayesian inference generative modeling, you have to have a full forward model of your detector. 320 00:26:16,024 --> 00:26:21,169 So that lets you then input your hydrodynamic simulation, for example. 321 00:26:21,169 --> 00:26:23,930 and predict what exact signal am I going to observe? 322 00:26:23,930 --> 00:26:26,171 it, you know, am I going to saturate my camera? 323 00:26:26,171 --> 00:26:30,312 Am I going to be able to resolve the signature, the spectral signature that I want? 324 00:26:30,373 --> 00:26:37,162 And so it involves iterating, you know, at that level to try and simulate your data and say, this is what I'm going to see. 325 00:26:37,162 --> 00:26:38,275 This is what I think I'm going to see. 326 00:26:38,275 --> 00:26:41,497 ah And then you do the experiment. 327 00:26:41,497 --> 00:26:42,543 It's completely different. 328 00:26:42,543 --> 00:26:45,238 And you're like, well, that's not what I thought was going to happen. 329 00:26:45,318 --> 00:26:50,820 And then you have to start trying to interpret that data, figure out why, what's wrong, what the simulation got wrong. 330 00:26:50,972 --> 00:27:04,358 and then try and build a better model that's able to capture the data and sort of model from, you know, physical principles, build a physical model of the system to model this 331 00:27:04,358 --> 00:27:07,299 data and interpret it and try and extract the most information out. 332 00:27:07,299 --> 00:27:16,753 So the technical challenge of designing the experiment is in its own right, you know, a challenge because doing large-scale science is challenging. 333 00:27:16,753 --> 00:27:20,284 It requires coordinating with a large team of people and 334 00:27:20,284 --> 00:27:27,550 doing a lot of very tedious paperwork to make sure that you can feel the experiment you want to feel. 335 00:27:28,171 --> 00:27:31,814 And then after that, the data is incredibly complicated. 336 00:27:31,814 --> 00:27:46,827 think, um you know, JJ famously uh only ever uh published data from one shot from his entire graduate career because these experiments are very difficult to do and then even 337 00:27:46,827 --> 00:27:48,388 more difficult to analyze. 338 00:27:48,451 --> 00:27:57,913 And so it's considered a very, successful uh PhD thesis if you can analyze at least one data set. 339 00:27:58,194 --> 00:28:09,326 And so being able to extend that and do it for multiple data sets is something that's even more challenging, even once you've built up the tools, because each individual shot of the 340 00:28:09,326 --> 00:28:11,727 laser is slightly different. 341 00:28:11,727 --> 00:28:15,978 And the nuance of that is something that is... 342 00:28:16,294 --> 00:28:17,845 Because anything can happen. 343 00:28:17,845 --> 00:28:20,577 You can always have some fluke occurrence. 344 00:28:20,577 --> 00:28:31,063 There's sort of an old apocryphal tale that someone was, someone had an experiment on the laser and a squirrel chewed through a cable on the roof and the whole thing just shut off. 345 00:28:31,063 --> 00:28:32,884 And for the, you know, that was it. 346 00:28:32,884 --> 00:28:36,075 They didn't get any data because the squirrel chewed through a cable on the roof. 347 00:28:36,075 --> 00:28:37,787 It's like, well, couldn't have predicted that. 348 00:28:37,787 --> 00:28:44,310 The code did not, uh did not contain the necessary squirrel physics um to predict that outcome. 349 00:28:44,588 --> 00:28:46,819 And so any, any number of things can go wrong. 350 00:28:46,819 --> 00:28:59,794 Um, but being able to, to build a robust model and statistical inference engine to interpret these data sets and get meaningful information out, uh, is, that's, I would say 351 00:28:59,794 --> 00:29:04,536 that's the largest, um, sort of technical challenge. 352 00:29:04,536 --> 00:29:10,949 two, these are, I mean, they're, they're, there's a lot of data, but it's also not a lot of data by modern standards, right? 353 00:29:10,949 --> 00:29:14,406 We can shoot the laser maybe one, you know, 354 00:29:14,406 --> 00:29:19,089 12 times in a single day, which is a lot for a laser that size. 355 00:29:19,310 --> 00:29:23,793 But you maybe get eight shots of usable data. 356 00:29:23,793 --> 00:29:31,618 And so it's very hard to scan through parameter space and answer the questions that you might want to answer. 357 00:29:31,979 --> 00:29:37,663 So it's a very complicated and difficult uh paradigm to operate in. 358 00:29:37,663 --> 00:29:40,764 But it's very fun and rewarding as well. 359 00:29:40,805 --> 00:29:41,665 Yeah. 360 00:29:41,926 --> 00:29:42,758 Yeah, again, guys. 361 00:29:42,758 --> 00:29:48,458 Uh, and how does, what's the timeline, you know, on these projects you're working on? 362 00:29:48,458 --> 00:29:55,918 Like for how long does it tend to, like, what's a range of duration you guys are working on these, on these projects? 363 00:29:55,918 --> 00:30:01,378 Because it sounds so daunting that like, I'm guessing you need a lot of time. 364 00:30:01,378 --> 00:30:01,578 Yeah. 365 00:30:01,578 --> 00:30:09,292 Well, I mean, the timeline for an experiment as you know, as I said, a little over a year, but then to actually analyze that, I mean, we're talking. 366 00:30:09,292 --> 00:30:10,903 the length of a PhD thesis, right? 367 00:30:10,903 --> 00:30:13,274 JJ analyzed one shot in his thesis. 368 00:30:13,274 --> 00:30:16,516 I'm going to analyze slightly more than one shot, but not by much. 369 00:30:16,516 --> 00:30:23,639 uh It takes, yeah, I would say it takes about five years to finish one of these projects. 370 00:30:23,639 --> 00:30:28,782 uh And a lot of that is because, you know, I'm a graduate student, I'm learning. 371 00:30:29,002 --> 00:30:34,625 There's a lot I didn't know, I didn't know when I started that I know that I don't know now. 372 00:30:34,625 --> 00:30:38,283 ah And so that definitely makes it 373 00:30:38,283 --> 00:30:39,683 hard to get going at the beginning. 374 00:30:39,683 --> 00:30:49,326 uh But when you're starting from scratch, right, you don't have any of the uh tools necessarily built out yet. 375 00:30:49,326 --> 00:30:51,427 It's very difficult to get going. 376 00:30:51,427 --> 00:31:02,690 so, you know, that's sort of, hopefully the, you know, I've built up enough tools that the, you know, my successor, whoever they may be, will be able to build on that just as I 377 00:31:02,690 --> 00:31:06,371 built on what JJ did and, he built on what came before him. 378 00:31:06,471 --> 00:31:07,395 And so. 379 00:31:07,395 --> 00:31:13,675 Hopefully it accelerates, it is very challenging to be able to build out these tools. 380 00:31:14,015 --> 00:31:21,455 Plus, I'm guessing a lot of people are working on these projects, Physics projects are usually very collaborative. 381 00:31:21,455 --> 00:31:23,135 So how does that look like for you? 382 00:31:23,135 --> 00:31:25,595 How many people are working with you on these projects? 383 00:31:25,595 --> 00:31:31,155 Maybe it's even inter-country collaborations. 384 00:31:31,275 --> 00:31:35,227 I think it's very interesting to also look a bit at that and... 385 00:31:35,393 --> 00:31:46,770 how it looks backstage because not a lot of projects are as high, as large scale and international as physics projects. 386 00:31:46,951 --> 00:31:48,752 Yeah, that's a good point. 387 00:31:48,752 --> 00:31:54,866 So one thing that's kind of unusual about the field of high-end density physics is it is very collaborative. 388 00:31:54,866 --> 00:31:57,878 We collaborate uh all over the world. 389 00:31:58,359 --> 00:32:05,513 There's laser systems that we go use all around the world in Europe, Japan, all over. 390 00:32:05,699 --> 00:32:14,359 Um, and so there is that collaboration, but at the same time, it's small enough scale that you sort of own your project. 391 00:32:14,359 --> 00:32:14,819 Right. 392 00:32:14,819 --> 00:32:16,639 So I proposed my experiments. 393 00:32:16,639 --> 00:32:19,279 had a lot of help fielding that experiment, right. 394 00:32:19,279 --> 00:32:22,379 From a, you know, diagnostic and laser side, right. 395 00:32:22,379 --> 00:32:24,619 I don't know anything about laser systems. 396 00:32:24,619 --> 00:32:29,359 Uh, so there's a whole team of laser professionals at the lab who run the laser. 397 00:32:29,419 --> 00:32:35,639 Um, and so, you know, there's a lot of help from them, but at the end of the day, you know, I designed it as, as a graduate student, I. 398 00:32:35,639 --> 00:32:41,504 Designed the target, what the laser is going to look like, what diagnostics we're going to field, and then I'm in charge of analyzing it. 399 00:32:41,504 --> 00:32:49,571 And so it's very collaborative in the sense that everyone works together to get good data, but you do own your own project at the end of the day. 400 00:32:49,571 --> 00:32:58,177 And there's a lot of freedom that comes with it to pursue what questions you think are interesting, um, and sort of push the, push the boundaries. 401 00:32:58,458 --> 00:33:03,762 But of course, you know, there's always, um, you can always go talk to experts in the field. 402 00:33:03,762 --> 00:33:04,923 um 403 00:33:05,089 --> 00:33:08,341 because there is that collaboration, those connections. 404 00:33:08,441 --> 00:33:21,781 But it's definitely not a collaboration in the sense that like CERN or Atlas are where you have, you know, a hundred people working towards the same uh goal. 405 00:33:21,781 --> 00:33:23,992 It's a lot more individualized. 406 00:33:25,053 --> 00:33:25,393 Yeah. 407 00:33:25,393 --> 00:33:25,694 Okay. 408 00:33:25,694 --> 00:33:26,064 Okay. 409 00:33:26,064 --> 00:33:26,474 I see. 410 00:33:26,474 --> 00:33:30,797 uh And so is there a particular mentor? 411 00:33:33,341 --> 00:33:45,299 that has actually been very important for you so far in your career that has helped you jumpstart your learning or maybe unblock you on a project? 412 00:33:45,759 --> 00:33:50,622 Yeah, well, I'll definitely shout out JJ Ruby here. 413 00:33:51,183 --> 00:33:54,565 He taught me everything I know about Bayesian inference. 414 00:33:54,565 --> 00:34:02,000 um But then, he went off to go work at Lawrence Livermore and now he's at the Houston Astros. 415 00:34:02,328 --> 00:34:03,358 but we still, we still talk. 416 00:34:03,358 --> 00:34:10,640 So he's, he's been a huge help in, um, you know, helping me in, cause he was in, he was in my shoes not long ago. 417 00:34:10,640 --> 00:34:21,823 So he has a lot of, uh, advice, but then, yeah, I, it's been, it's a lot of, um, community among the graduate students at the lab. 418 00:34:22,284 --> 00:34:25,825 and so I, there's a lot of peer to peer, mentoring. 419 00:34:25,825 --> 00:34:30,346 Um, I'll, I'll shout out David Bischel and Alex Chin. 420 00:34:30,402 --> 00:34:36,667 were two senior graduate students who sort of helped me out when I first came in and knew absolutely nothing. 421 00:34:36,667 --> 00:34:39,909 And, you know, they sort of taught me what they knew. 422 00:34:39,909 --> 00:34:43,131 ah And now I'm the senior graduate student. 423 00:34:43,131 --> 00:34:48,754 And so I'm trying to pay that forward and, you know, teach me what I learned and what they taught me. 424 00:34:48,955 --> 00:34:56,500 so there's this sort of, you know, sense of community among graduate students because we're all sort of in the same boat. 425 00:34:56,840 --> 00:34:58,301 Grad school is hard. 426 00:34:58,341 --> 00:34:59,922 And so it's very... 427 00:35:00,234 --> 00:35:01,436 It's very helpful to have that. 428 00:35:01,436 --> 00:35:06,425 em You have someone you can talk to who's been through this. 429 00:35:06,807 --> 00:35:11,254 So it's not just one mentor, it's a whole lab of them. 430 00:35:13,559 --> 00:35:16,741 Yeah, I can guess that. 431 00:35:16,741 --> 00:35:32,895 And yeah, of course, I will put also JJ's appearance on the show in the show notes for people who want to dig deeper because JJ definitely has a very interesting background and 432 00:35:32,895 --> 00:35:41,980 path that I think is going to be very valuable for listeners to listen to, especially if are interested in... 433 00:35:41,980 --> 00:35:45,672 Either physics or baseball analytics or both. 434 00:35:45,672 --> 00:35:48,393 Definitely give a listen to this episode. 435 00:35:48,393 --> 00:35:52,574 And now let's talk a bit more concretely. 436 00:35:52,574 --> 00:35:58,377 How do you incorporate patient inference in your work? 437 00:35:58,377 --> 00:36:05,299 Especially so the kind of high energy density physics uh data you have. 438 00:36:05,440 --> 00:36:11,696 What advantages or limitations arise from using patient techniques? 439 00:36:11,696 --> 00:36:15,867 Yeah, great question and one that people ask me all the time. 440 00:36:15,907 --> 00:36:17,358 They're like, well, what's the point? 441 00:36:17,358 --> 00:36:18,598 Who cares about Bayesian inference? 442 00:36:18,598 --> 00:36:19,908 Why are you doing this? 443 00:36:20,909 --> 00:36:28,391 To which I say Bayesian inference is really the tool that you need to answer a problem like this. 444 00:36:28,391 --> 00:36:34,372 at a very high level, right, you observe this data and you need to understand the system that created it. 445 00:36:34,372 --> 00:36:37,453 And so this is a, you know, it's an inverse problem. 446 00:36:37,453 --> 00:36:39,824 And so it's very difficult to just invert the data. 447 00:36:39,824 --> 00:36:41,058 You have to do it. 448 00:36:41,058 --> 00:36:42,999 in a generative framework. 449 00:36:42,999 --> 00:36:46,190 And so right away Bayesian inference is very well suited for that kind of thing. 450 00:36:46,190 --> 00:36:47,491 You don't have to do Bayesian inference. 451 00:36:47,491 --> 00:36:51,763 You can do forward modeling uh without Bayesian statistics. 452 00:36:51,763 --> 00:36:53,894 But it's just such a natural framework for it. 453 00:36:53,894 --> 00:37:02,447 And then on top of that, it allows you ah what I think is probably the most powerful thing is to include prior information in a way that you can't. 454 00:37:02,447 --> 00:37:06,593 So we have historical data, for example, going back. 455 00:37:06,593 --> 00:37:12,237 several decades to the older version, you know, the, the Omega laser system has been around since the nineties. 456 00:37:12,478 --> 00:37:16,902 So we have a wealth of historical data that helps us understand these systems. 457 00:37:16,902 --> 00:37:26,930 And so that you can use that as a form of prior knowledge to sort of inform, you know, your expectations of what's going to come out of what you're going to learn from these 458 00:37:26,930 --> 00:37:28,070 experiments. 459 00:37:28,171 --> 00:37:35,673 And so that, you know, provides a self-consistent way to sort of include that prior knowledge that will, you know, 460 00:37:35,673 --> 00:37:41,293 improve the statistical inference that you're able to make from these. 461 00:37:41,293 --> 00:37:47,533 Because you can take a lot of measurements, but they're not, they may not have as much information as you need them to have. 462 00:37:47,533 --> 00:37:51,553 And so being able to impose prior knowledge is also very valuable. 463 00:37:51,973 --> 00:38:03,473 And then also, you know, one thing we struggle with in our field is propagation of uncertainties, because you have very highly nonlinear systems. 464 00:38:03,605 --> 00:38:15,350 And sometimes you can't uh build a model that's able to uh sort of characterize your level of uncertainty and Bayesian inference folds that in completely self-consistently and 465 00:38:15,350 --> 00:38:16,597 correlations and whatnot. 466 00:38:16,597 --> 00:38:25,605 you're doing uh Markov chain Monte Carlo, and even if you're doing something like stochastic variational inference, you can still get a much better understanding of the 467 00:38:25,605 --> 00:38:30,199 uncertainty uh in your system based on your data, which is 468 00:38:30,617 --> 00:38:31,868 very non-trivial for these systems. 469 00:38:31,868 --> 00:38:39,142 So I would say that's the big three uh reasons why we use Bayesian inference to analyze the data. 470 00:38:39,142 --> 00:38:41,723 It's because I think it just makes sense. 471 00:38:41,743 --> 00:38:51,589 Yeah, mean, preaching to the choir, but yeah, completely relate, of course. 472 00:38:51,589 --> 00:38:59,243 And actually, can you describe one experiment or project that 473 00:38:59,531 --> 00:39:12,546 you're particularly proud of where you used Bayesian inference, um what you measured, how, why it's significant and what does the model look like? 474 00:39:13,187 --> 00:39:14,328 Yeah, sure. 475 00:39:14,328 --> 00:39:17,019 Well, it's still ongoing, so it's not done. 476 00:39:17,019 --> 00:39:21,651 I don't know if I'm allowed to be proud of it if uh it hasn't been published yet, but that's fine. 477 00:39:21,651 --> 00:39:25,152 um So one thing that... 478 00:39:25,152 --> 00:39:27,095 um 479 00:39:27,095 --> 00:39:40,197 This sort of the what's become my main project at this point is uh making a sort of measurement of the equation of state of a billion atmosphere pressure plasma. 480 00:39:40,318 --> 00:39:44,922 So the equation of state is just the relationship between pressure, temperature, intensity basically. 481 00:39:44,922 --> 00:39:51,065 You think of the ideal gas law, uh van der Waals equation of state maybe to name a few. 482 00:39:51,065 --> 00:39:52,122 And so 483 00:39:52,122 --> 00:39:54,233 It's a very important material property, right? 484 00:39:54,233 --> 00:40:00,066 Because it says, if you're at a temperature and density, what's the pressure, what's the energy density, and so forth. 485 00:40:00,326 --> 00:40:07,440 And so we've sort of developed the tools to do this at lower pressures, uh fairly routinely. 486 00:40:07,440 --> 00:40:12,273 You can measure the temperature, density, and pressure of a planar experiment, for example. 487 00:40:12,273 --> 00:40:17,665 So you can get all the very high pressure, but you can't, as I said, there's a limit to how hard you can hit something with a laser. 488 00:40:17,665 --> 00:40:21,197 So you can't measure the equation of state beyond 489 00:40:21,235 --> 00:40:24,287 you know, 10 million atmospheres or so. 490 00:40:24,768 --> 00:40:29,773 And so, you know, you have to go to a convergent geometry, but that's very difficult. 491 00:40:29,773 --> 00:40:41,613 And so, um sort of what's become my mission is making that measurement using Bayesian uh inference techniques and other statistical methods, know, machine learning and whatnot, to 492 00:40:41,613 --> 00:40:47,899 take this really, this big data set um that, you know, consists of 493 00:40:47,899 --> 00:40:59,467 multiple different X-ray cameras, spectrometers, all sorts of, you know, the full kitchen sink and analyzing that data set, this multi-messenger data set, uh comprehensively at the 494 00:40:59,467 --> 00:41:05,131 same time and sort of in the data space as much as possible. 495 00:41:05,131 --> 00:41:07,983 So sort of from first principles, if you will. 496 00:41:08,423 --> 00:41:14,338 And so, you know, this was a really complicated system just to build, right? 497 00:41:14,338 --> 00:41:16,749 You have to build models of all these detectors. 498 00:41:16,775 --> 00:41:30,819 and understand the statistics uh in some sense, and then do the inference, which is going to be a very high dimensional problem um requiring the very advanced sampling tools 499 00:41:30,819 --> 00:41:34,742 provided by PyMC and other libraries. 500 00:41:35,263 --> 00:41:41,509 And so what we were able to do is take that data set and 501 00:41:41,509 --> 00:41:45,762 throw this very complicated model at it, which is it's complicated, but it's simple, right? 502 00:41:45,762 --> 00:41:54,739 Because it's basically, you know, you have your temperature and you have your density and your pressure, and you have a very simple, but sort of complete physical picture of how 503 00:41:54,739 --> 00:42:06,047 they evolve in time and space that's sort of parameterized in terms of these physical quantities, but flexible enough to be able to fit the data while encoding, you know, our 504 00:42:06,047 --> 00:42:08,745 prior belief on the physics of the system. 505 00:42:08,745 --> 00:42:12,566 Like for example, you we think the temperature should go up and then down, right? 506 00:42:12,786 --> 00:42:15,027 And then stuff like that. 507 00:42:15,227 --> 00:42:24,119 And so doing that, building this model that, you know, treats the evolution of the system and then, you know, propagates that forward. 508 00:42:24,119 --> 00:42:25,620 What X-rays does that produce? 509 00:42:25,620 --> 00:42:27,010 How do we observe those? 510 00:42:27,010 --> 00:42:28,650 What would we measure if that were the case? 511 00:42:28,650 --> 00:42:33,172 Bayesian inference, you know, the likelihood of uh data given theta. 512 00:42:33,432 --> 00:42:38,005 And then we are able to sort of independently constrain the 513 00:42:38,005 --> 00:42:42,068 pressure, temperature, and density from the available suite of measurements. 514 00:42:42,068 --> 00:42:46,353 And so that gives you sort of separate measurements of what's the temperature and density and what's the pressure. 515 00:42:46,353 --> 00:42:49,435 And those are related by the equation of state. 516 00:42:49,836 --> 00:42:56,703 And so this gives you a very fancy way to measure the equation of state using these advanced statistical methods. 517 00:42:56,703 --> 00:43:03,018 And so we were able to do that for the first time, and that gives you sort of insight into what physics is happening, right? 518 00:43:03,018 --> 00:43:04,117 You can sort of... 519 00:43:04,117 --> 00:43:06,258 You can compare it to your models. 520 00:43:06,258 --> 00:43:12,170 can do further inference on those conditions that you infer and try and understand what's happening here. 521 00:43:12,170 --> 00:43:17,704 there sort of interactions that we aren't taking into account in the physics of the system? 522 00:43:17,844 --> 00:43:20,326 Is it just the ideal gas law? 523 00:43:20,326 --> 00:43:21,797 Is there something else going on? 524 00:43:21,797 --> 00:43:26,429 uh Are the photons that we're creating producing their own pressure? 525 00:43:26,729 --> 00:43:28,830 Let's ask a lot of interesting questions. 526 00:43:29,851 --> 00:43:32,912 to my knowledge, it's the first measurement uh out there. 527 00:43:32,912 --> 00:43:34,253 uh 528 00:43:34,321 --> 00:43:36,322 these billion atmosphere pressures. 529 00:43:36,982 --> 00:43:49,329 And so, you we're still working on uh putting that together for publication, but just, you know, the sheer lift of putting it together and doing it in a self-consistent way is 530 00:43:49,329 --> 00:43:59,774 something that I'm proud of and has really, you know, pushed uh what JJ did with his single diagnostic, single shot analysis to a, you know, a more complete framework. 531 00:43:59,774 --> 00:44:03,992 And then, you know, because I took multiple shots of data, 532 00:44:03,992 --> 00:44:06,894 I can, I think I've analyzed seven, seven shots now. 533 00:44:06,894 --> 00:44:16,562 Uh, so that's, you know, a vast improvement for, for the statistics, know, square root of N was so much more, uh, so much more certain. 534 00:44:16,943 --> 00:44:17,293 Yeah. 535 00:44:17,293 --> 00:44:17,608 Yeah. 536 00:44:17,608 --> 00:44:17,924 Yeah. 537 00:44:17,924 --> 00:44:18,224 Damn. 538 00:44:18,224 --> 00:44:18,945 Well done. 539 00:44:18,945 --> 00:44:26,351 Um, yeah, I know this is, this is such a, such a hard work, but very impressive. 540 00:44:26,351 --> 00:44:31,576 And how, so first practical question, because I'm always curious. 541 00:44:31,576 --> 00:44:32,586 What? 542 00:44:32,716 --> 00:44:36,777 What did you use to code the actual model and to sample? 543 00:44:36,777 --> 00:44:39,838 Did you use any already existing package? 544 00:44:39,838 --> 00:44:46,980 um Did you have to come up with your custom packages, maybe even custom sample? 545 00:44:46,980 --> 00:44:48,301 How does it work? 546 00:44:48,301 --> 00:44:49,821 What does it look like? 547 00:44:49,821 --> 00:44:57,604 So I am a big proponent of Python, doing everything in Python, which is uh an increasingly popular opinion, out. 548 00:44:57,604 --> 00:44:57,904 Python... 549 00:44:57,904 --> 00:44:58,990 uh 550 00:44:58,990 --> 00:45:04,564 When I first started, no one in my lab was using Python, which was crazy, except for JJ. 551 00:45:04,564 --> 00:45:08,727 It was, everyone is using like Matlab, Excel, Mathematica. 552 00:45:08,727 --> 00:45:09,888 Python was... 553 00:45:09,888 --> 00:45:12,629 Damn, this is A lot of people still use Excel. 554 00:45:12,629 --> 00:45:18,073 I will say, you in their defense, what they're able to do with Excel is nothing short of magic. 555 00:45:18,073 --> 00:45:18,283 mean... 556 00:45:18,283 --> 00:45:18,834 Yeah, yeah. 557 00:45:18,834 --> 00:45:22,536 I mean, at that level, you have to be an Excel... 558 00:45:22,536 --> 00:45:25,548 I mean, they're doing everything I can do in Python. 559 00:45:25,548 --> 00:45:26,639 They are doing an Excel. 560 00:45:26,639 --> 00:45:28,780 It just looks ridiculous. 561 00:45:29,413 --> 00:45:29,693 Yeah. 562 00:45:29,693 --> 00:45:30,233 Yeah. 563 00:45:30,233 --> 00:45:30,853 Damn. 564 00:45:30,853 --> 00:45:31,554 Wow. 565 00:45:31,554 --> 00:45:32,255 Okay. 566 00:45:32,255 --> 00:45:33,866 Um, but good choice. 567 00:45:33,866 --> 00:45:40,042 I, I, I completely support you, Ethan and JJ, this, this endeavor. 568 00:45:40,042 --> 00:45:48,109 uh, yeah, so, so we're big, big, uh, proponents of Python and the sort of big knock on Python is that it's slow. 569 00:45:48,479 --> 00:45:56,336 but that's actually becoming less and less of an issue when you have things like, you know, just in time compilation, uh with, Jack's. 570 00:45:56,852 --> 00:46:00,815 and other things that are, you automatic differentiation. 571 00:46:00,815 --> 00:46:08,301 All of that is sort of closing the gap between Python and these sort of machine, you know, Fortran and C++ and whatnot. 572 00:46:08,301 --> 00:46:18,970 And so the ease of use is so much better that, you know, coding in Python is really, think, I mean, you obviously agree, is the practical choice at this stage. 573 00:46:18,970 --> 00:46:21,656 uh And so, you know, 574 00:46:21,656 --> 00:46:23,518 A lot of it is pre-existing, right? 575 00:46:23,518 --> 00:46:25,500 There's the Jax libraries. 576 00:46:25,500 --> 00:46:27,381 A lot of code exists. 577 00:46:27,381 --> 00:46:32,225 I didn't have to recode NP.traps or anything like that. 578 00:46:32,446 --> 00:46:37,770 But there's a lot of very bespoke code that I had to rewrite in Jax. 579 00:46:37,770 --> 00:46:42,894 That's an implementation that doesn't exist that does exactly what I want it to do. 580 00:46:43,095 --> 00:46:48,918 so I had to do a lot of software development uh considering I... 581 00:46:48,918 --> 00:46:55,511 came into college or uh graduate school rather with zero, basically zero Python experience. 582 00:46:55,511 --> 00:47:04,874 And so I had to teach myself from the ground up uh how to code in Python and then how to code in Jax and then how to develop a large scale software package. 583 00:47:04,874 --> 00:47:09,836 ah It's not beautiful, but it does work. 584 00:47:09,836 --> 00:47:17,439 that's another thing that I guess I should add that I'm proud of is the sheer amount of code that I've written. 585 00:47:17,483 --> 00:47:18,774 uh for this project. 586 00:47:18,774 --> 00:47:20,465 It's a massive repository. 587 00:47:20,465 --> 00:47:24,237 uh Just to, you know, analyze a couple pieces of data. 588 00:47:24,237 --> 00:47:28,729 yeah, shout out to Python, shout out to Jax. 589 00:47:28,729 --> 00:47:35,753 um We used to use PyMC as our sampling library to do the inference. 590 00:47:35,874 --> 00:47:45,939 We switched recently to NumPyro because that was, you know, this is when NumPyro was sort of new and Jax was sort of cutting edge. 591 00:47:46,175 --> 00:47:49,896 And now think PyMC uses Jax at its backend. 592 00:47:50,156 --> 00:47:51,876 Yeah, I mean, you can. 593 00:47:52,877 --> 00:47:55,758 Yeah, we have the three backends. 594 00:47:55,758 --> 00:47:58,878 So C is the default one. 595 00:47:58,999 --> 00:48:01,049 You can use Numba and you can use Jax. 596 00:48:01,049 --> 00:48:08,081 So depending on the project you're working on, be useful to use Numba or it will be faster using Jax. 597 00:48:08,081 --> 00:48:10,462 It depends on your model. 598 00:48:10,882 --> 00:48:13,583 yeah, that's great. 599 00:48:14,565 --> 00:48:22,561 Anyway, you're using a very good open source package and you build on top of um it. 600 00:48:22,561 --> 00:48:24,762 What are you using as a sampler? 601 00:48:26,564 --> 00:48:28,585 latest MCMC? 602 00:48:29,286 --> 00:48:31,347 So yeah, it depends. 603 00:48:31,347 --> 00:48:35,850 eh Usually we use, know, not Snowy U-turn sampler. 604 00:48:35,850 --> 00:48:42,921 um I mean, have to sort of coax it to do what you want on occasion, but it's... 605 00:48:42,921 --> 00:48:47,333 sort of the Ferrari of samplers, as far as we're concerned. 606 00:48:48,153 --> 00:48:56,517 Occasionally, they will, you when you drive a Ferrari on a road full of potholes, you will occasionally encounter a negative result. 607 00:48:56,597 --> 00:49:03,881 And so we may have to use, you know, something like sequential Monte Carlo SMC or some other more robust but slower. 608 00:49:03,881 --> 00:49:08,362 uh Nest and sampling is another one that I've used a lot. 609 00:49:09,675 --> 00:49:15,555 So it's unfortunate because you'd like there to be one sampler that always works, but that is never the case. 610 00:49:15,555 --> 00:49:18,255 And you always have to do something else. 611 00:49:18,715 --> 00:49:25,754 Sometimes when you get really desperate, you'll be like, all right, we've got to use a totally different sampling library to get what we want. 612 00:49:25,754 --> 00:49:27,815 But that's part of the and part of the fun. 613 00:49:27,815 --> 00:49:34,835 It's already working for most of the, like for 98 % of the cases, and since it will be amazing. 614 00:49:34,835 --> 00:49:37,715 I mean, that's, and so, yeah. 615 00:49:37,993 --> 00:49:41,796 And that's great to hear that you didn't have to code your custom sampler. 616 00:49:41,796 --> 00:49:42,106 Right. 617 00:49:42,106 --> 00:49:48,071 So you see folks, you don't need to, you don't need to write your custom samplers anymore. 618 00:49:48,071 --> 00:49:54,216 Especially I'm, I'm, I'm speaking to the economics, uh, among econometricians among us. 619 00:49:54,216 --> 00:49:57,719 Like don't stop writing your own custom gift sampler. 620 00:49:58,280 --> 00:50:06,746 Just, uh, just use the ones that are in PIMC or NumPyro or Stan or whatever you're using as a probabilistic programming language is going to work well. 621 00:50:06,746 --> 00:50:07,898 And there are now. 622 00:50:07,898 --> 00:50:11,190 More than the nut sampler, there are many more of them. 623 00:50:11,190 --> 00:50:12,671 I % agree. 624 00:50:13,012 --> 00:50:19,156 And I've been trying to proselytize a little bit around different research groups. 625 00:50:19,156 --> 00:50:31,594 I've got some Bayesian inference outreach uh presentations that I've been giving, trying to spread the word that, hey, Bayesian inference is not as hard as it used to be in the 626 00:50:31,594 --> 00:50:32,124 olden days. 627 00:50:32,124 --> 00:50:33,505 You don't have to write your own sampler. 628 00:50:33,505 --> 00:50:36,647 You can just use one of these excellent uh 629 00:50:36,647 --> 00:50:40,928 pre-made ones that are only getting better and more sophisticated. 630 00:50:41,108 --> 00:50:41,868 exactly. 631 00:50:41,868 --> 00:50:44,909 The barrier to entry is so much lower now. 632 00:50:44,909 --> 00:50:45,489 Exactly. 633 00:50:45,489 --> 00:50:46,740 It's way, way lower. 634 00:50:46,740 --> 00:51:03,185 So now you can spin up a patient model and sample from it very fast and don't even have to worry about the sampling issues because it's all taken care of for you. 635 00:51:03,185 --> 00:51:04,995 I would say it's 636 00:51:05,575 --> 00:51:07,056 It's a great time to leave. 637 00:51:07,056 --> 00:51:16,863 m And a lot of people are scared of Bayesian inference because it sounds fancy, but it's, know, really, you know, quite easy. 638 00:51:16,863 --> 00:51:22,266 just build your model, turn the crank, Bayes rule and get out your posterior. 639 00:51:22,346 --> 00:51:24,267 Yeah, exactly. 640 00:51:25,148 --> 00:51:26,539 Not a lot of things to do now. 641 00:51:26,539 --> 00:51:29,501 It's pretty much, pretty much all automated. 642 00:51:29,501 --> 00:51:30,162 It's solved. 643 00:51:30,162 --> 00:51:30,652 Yeah. 644 00:51:30,652 --> 00:51:33,233 Bayesian inference is it's all done. 645 00:51:34,154 --> 00:51:35,075 Yeah. 646 00:51:35,353 --> 00:51:46,703 Actually, do you have um any good practices that you want to share with people when they're working with big code bases as you are? 647 00:51:46,703 --> 00:51:50,749 um Well, in your case in Python, but can be, can be anything else? 648 00:51:50,749 --> 00:52:04,375 Like what are, what are the good practices you've found to not lose yourself basically when working on these, on these big code bases and also probably to help. 649 00:52:04,551 --> 00:52:07,953 uh collaboration as we were talking about. 650 00:52:08,134 --> 00:52:10,915 Well, the first thing is always comment your code. 651 00:52:10,915 --> 00:52:22,363 And I say this with full knowledge that I am not very good at this, but always, for your future self at the very least, you know, write down your thought process so you can come 652 00:52:22,363 --> 00:52:23,324 back to it. 653 00:52:23,744 --> 00:52:33,611 But then on a, you know, sort of more sophisticated level, uh there's a sort of white paper by Andrew Goman. 654 00:52:33,707 --> 00:52:40,993 I think it's called a principal Bayesian workflow that sort of lays out best practices for constructing a model. 655 00:52:40,993 --> 00:52:47,098 And so, you construct your model, you uh test the prior predictive distribution. 656 00:52:47,098 --> 00:52:49,150 You you say, okay, is my prior correct? 657 00:52:49,150 --> 00:52:50,600 And then you go. 658 00:52:50,681 --> 00:53:02,330 And I think one thing that's really useful, especially in, you know, the sort of work that I do is building a sort of synthetic experiment and, you know, making sure that you can 659 00:53:02,330 --> 00:53:03,131 get out 660 00:53:03,131 --> 00:53:05,022 the information you think you can get out, right? 661 00:53:05,022 --> 00:53:15,369 Because if you build this whole synthetic set of synthetic diagnostics and your model, and then you take the data and you try to analyze it and you actually don't get any 662 00:53:15,369 --> 00:53:16,400 information out. 663 00:53:16,400 --> 00:53:17,621 Well, great. 664 00:53:17,621 --> 00:53:22,905 You you wasted a bunch of time and money on that experiment and you didn't get what you wanted out. 665 00:53:22,905 --> 00:53:31,891 And so being able to sort of, before you've even done the experiment, predict what you're, what you're going to measure and then what information you can get out of that is hugely 666 00:53:31,891 --> 00:53:32,761 helpful. 667 00:53:33,023 --> 00:53:40,306 And really, you know, should be a prerequisite for doing any of these experiments and showing that you think it's going to work. 668 00:53:40,306 --> 00:53:49,239 I mean, there's value in just shooting a laser and seeing what happens, but at this stage, you know, the field is advanced enough that we should be doing this a little more 669 00:53:49,239 --> 00:53:50,510 predictively. 670 00:53:51,210 --> 00:53:56,312 And then two, sort of um benchmarking and testing your model along the way. 671 00:53:56,312 --> 00:54:01,652 And so one thing that I really struggled with is I sort of put the model together all at once. 672 00:54:01,652 --> 00:54:03,744 And I just pressed go and it didn't work. 673 00:54:03,744 --> 00:54:06,125 And I was like, well, what, why didn't it work? 674 00:54:06,166 --> 00:54:06,896 What's breaking? 675 00:54:06,896 --> 00:54:09,188 And so then I was like, well, okay. 676 00:54:09,188 --> 00:54:15,314 I took it apart and piece by piece, you know, and I tested each module, made sure it did what I thought it was doing. 677 00:54:15,314 --> 00:54:20,728 You know, ran the inference, you know, ran the synthetic experiment, but let's say, okay, let's say there's no noise. 678 00:54:20,728 --> 00:54:22,079 It's a perfect measurement. 679 00:54:22,079 --> 00:54:23,540 Can I, does this work? 680 00:54:23,540 --> 00:54:30,676 so building confidence in your model is very important and convincing yourself that it works so that you can convince others. 681 00:54:31,262 --> 00:54:33,543 And that helps you find a lot of the issues. 682 00:54:33,744 --> 00:54:39,327 Because you definitely can lose yourself ah in the model building process. 683 00:54:39,327 --> 00:54:43,879 As fun as it is, ah once the model gets too big, there's nothing you can do. 684 00:54:43,879 --> 00:54:44,769 gotta... 685 00:54:45,070 --> 00:54:46,851 You gotta just, you know, one brick at a time. 686 00:54:46,851 --> 00:54:52,213 And after you place each brick, make sure that that brick is a brick. 687 00:54:52,734 --> 00:54:53,594 Hmm. 688 00:54:54,315 --> 00:54:55,856 Yeah, yeah, okay. 689 00:54:55,856 --> 00:54:57,136 Yeah, I see what you mean. 690 00:54:57,136 --> 00:54:58,737 So basically... 691 00:54:59,638 --> 00:55:00,658 Having... 692 00:55:01,076 --> 00:55:07,630 A principled patient workflow, building block by block and testing it along the way. 693 00:55:07,630 --> 00:55:08,530 Is that correct? 694 00:55:08,530 --> 00:55:12,822 So how do you benchmark and test your model actually? 695 00:55:12,822 --> 00:55:18,916 How do you, what does a good model look like to you in this kind of project? 696 00:55:18,916 --> 00:55:21,217 How do you call a model good? 697 00:55:21,217 --> 00:55:22,282 How do you evaluate it? 698 00:55:22,282 --> 00:55:23,448 How do you benchmark it? 699 00:55:23,448 --> 00:55:30,338 I think it's a very important question, not only in your case, but in everybody's case, like all my models, I need to benchmark them. 700 00:55:30,338 --> 00:55:37,201 So yeah, I'm curious, what does that look like for you and what does that mean to have a good model? 701 00:55:37,382 --> 00:55:37,982 Right. 702 00:55:37,982 --> 00:55:45,796 So for me, for the specific use case that I work on, a good, well, all models are wrong, but some models are useful. 703 00:55:45,796 --> 00:55:51,309 So a good model is one that is useful that you can use to get useful information about your system out. 704 00:55:51,330 --> 00:55:59,906 And so what that usually looks like is we'll run a, you know, state of the art, a full physics simulation, you know, hydrodynamics code. 705 00:55:59,906 --> 00:56:10,826 simulate our experiment, generate synthetic data from that using our state-of-the-art detector models, and then do inference on that synthetic data and say, okay, so we know 706 00:56:10,826 --> 00:56:14,986 what the ground truth value is from our simulation. 707 00:56:14,986 --> 00:56:17,726 Can we recover that with our model? 708 00:56:17,726 --> 00:56:21,946 And oftentimes the model is a vastly simplified physical model, right? 709 00:56:21,946 --> 00:56:27,746 Because these state-of-the-art simulations take hours to sometimes days to run one simulation. 710 00:56:27,746 --> 00:56:28,626 So you can't 711 00:56:28,626 --> 00:56:33,787 really iterate over parameter space on that time scale. 712 00:56:33,888 --> 00:56:40,209 And so it's important to do this full physics and benchmark your reduced order model on that. 713 00:56:40,329 --> 00:56:45,611 And then if that works great, so maybe you benchmarked it against a 1D simulation, it worked great. 714 00:56:45,611 --> 00:56:49,832 You go up to a 2D simulation or a 3D simulation and see, it still work? 715 00:56:49,832 --> 00:56:51,132 Do we still get useful information? 716 00:56:51,132 --> 00:56:56,680 And so verifying that you can get out of this set of integrated measurements. 717 00:56:56,680 --> 00:56:58,932 Can you recover the ground truth value? 718 00:56:58,932 --> 00:57:01,341 In the limiting case, when you know what it is. 719 00:57:01,341 --> 00:57:07,997 And that builds confidence that the result you get out of the experimental data is meaningful and useful. 720 00:57:07,998 --> 00:57:14,682 And so that's a very important piece of model development for the specific use case I work on. 721 00:57:14,682 --> 00:57:21,246 I don't know what that really looks like in other fields where maybe in social sciences and whatnot. 722 00:57:21,246 --> 00:57:24,268 I don't know if you can simulate an entire economy. 723 00:57:24,268 --> 00:57:25,487 uh 724 00:57:25,487 --> 00:57:27,757 and the stock market, maybe you can. 725 00:57:30,198 --> 00:57:31,699 Yeah, you probably can. 726 00:57:31,699 --> 00:57:39,021 um I think Jesse Grabowski would be able to talk about that, but I know he does a lot of simulation work. 727 00:57:39,021 --> 00:57:46,343 mean, at some point also it looks a lot like physics simulations when you're simulating a full economy. 728 00:57:46,343 --> 00:57:51,994 uh There's a lot of interactions and dynamics. 729 00:57:51,994 --> 00:57:55,405 so state space models actually look 730 00:57:55,503 --> 00:58:01,718 Like it look a lot like um physics simulations. 731 00:58:01,798 --> 00:58:05,041 You can definitely use that in physics, I'm sure. 732 00:58:05,053 --> 00:58:10,286 a lot of physicists do go work on Wall Street after their PhDs. 733 00:58:10,286 --> 00:58:12,547 It's either baseball or Wall Street. 734 00:58:12,547 --> 00:58:12,868 Right. 735 00:58:12,868 --> 00:58:13,548 Yeah. 736 00:58:13,548 --> 00:58:17,871 I mean, you can really do anything with physics. 737 00:58:18,112 --> 00:58:21,634 yeah, that's what you could have answered your parents. 738 00:58:21,755 --> 00:58:24,237 Well, you can do anything with a physics degree. 739 00:58:24,349 --> 00:58:26,150 Well, I didn't know that at the time, but I know. 740 00:58:26,150 --> 00:58:27,450 Yeah. 741 00:58:27,450 --> 00:58:30,711 Now you can't, now you can't. 742 00:58:31,132 --> 00:58:41,916 And actually, since you talk about simulations, I know Andrew Gell-Mann, for instance, is a big proponent of simulating the data and testing the models on simulated data, which I 743 00:58:41,916 --> 00:58:44,558 do all the time also because, you know, I'm a good student. 744 00:58:44,558 --> 00:58:45,938 listen to Andrew. 745 00:58:46,979 --> 00:58:48,429 He knows way better than I do. 746 00:58:48,429 --> 00:58:51,481 How does that look like for you? 747 00:58:51,481 --> 00:58:52,709 You know, how do you... 748 00:58:52,709 --> 00:58:57,012 balance, simulation, modeling, experimenting your own work? 749 00:58:57,052 --> 00:59:06,538 And a bit more generally, which side, if there is one, theory or experiment, do you find more challenging or more rewarding? 750 00:59:07,399 --> 00:59:13,673 Yeah, well, so I do a fair amount of simulations, right? 751 00:59:13,673 --> 00:59:17,086 I use these large scale physics codes to simulate my experiments. 752 00:59:17,086 --> 00:59:21,018 ah And, you know, as I alluded to earlier, they're sort of always wrong. 753 00:59:21,018 --> 00:59:22,489 They're not predictive. 754 00:59:22,661 --> 00:59:30,769 which is a big problem, ah especially when you start thinking about using implosion experiments for fusion applications, right? 755 00:59:30,769 --> 00:59:40,718 You want to make more neutrons, and so you want to be able to use your simulations to say, what do I change to make more yield, more fusions? 756 00:59:40,718 --> 00:59:47,704 ah And so having a code that's not predictive is not particularly useful in that capacity. 757 00:59:47,904 --> 00:59:48,985 And so... 758 00:59:49,255 --> 00:59:57,789 For me, the use of the simulations is that it provides a sort of physically self-consistent system to test your models on. 759 00:59:58,269 --> 00:59:59,930 It's not correct. 760 00:59:59,930 --> 01:00:12,585 It's not a perfect picture of what's going to happen in your experiment, but it provides a sort of synthetic ah universe, a self-consistent universe for you to sort of test your 761 01:00:12,585 --> 01:00:13,696 models on. 762 01:00:13,716 --> 01:00:16,477 And so to me, I mean... 763 01:00:16,481 --> 01:00:19,804 I think any experimentalist would tell you that the experiment is king. 764 01:00:19,804 --> 01:00:24,407 Reality, what actually happens is the most important thing. 765 01:00:24,407 --> 01:00:32,913 And the theory should try to be predictive of what actually happens rather than, oftentimes what happens, you'll do the experiment, it doesn't match your initial 766 01:00:32,913 --> 01:00:33,484 simulation. 767 01:00:33,484 --> 01:00:37,759 And so you'll tune your simulation to sort of match it and you'll add some fudge factors, whatever. 768 01:00:37,759 --> 01:00:39,248 you're like, look, we discovered this thing. 769 01:00:39,248 --> 01:00:40,759 It's like, wow. 770 01:00:40,759 --> 01:00:43,861 It would have been nice if you could have predicted it beforehand. 771 01:00:44,217 --> 01:00:53,002 uh And so I sort of you know make fun of my theorist colleagues a little bit and could see know they have these predictions and it's like well It doesn't match doesn't match the 772 01:00:53,002 --> 01:00:57,795 experiment if not, then I don't you know, but then they'll always find that well Maybe your experiment is wrong. 773 01:00:57,795 --> 01:01:00,056 It's like but you know it happens. 774 01:01:00,056 --> 01:01:07,860 Okay, good this You know, we're all sharing the same experiential reality We all agree that we did the experiment and this is what happened. 775 01:01:07,860 --> 01:01:12,823 But it's a it's a yeah, it's an interesting uh inner play 776 01:01:12,845 --> 01:01:20,298 because both sides are saying the other one is wrong and that it has to be, we have quantum accurate density functional theory calculation. 777 01:01:20,298 --> 01:01:22,129 This is the truth. 778 01:01:22,569 --> 01:01:28,752 But every side has its, there are skeletons in everyone's closet. 779 01:01:29,112 --> 01:01:30,912 It's things that we don't understand. 780 01:01:30,913 --> 01:01:41,207 so, ah trying to sort of unify those two silos is an ongoing struggle and something that we've been trying to work on sort of. 781 01:01:41,207 --> 01:01:49,202 have theorists and experimentalists work together on experimental design, for example, uh to do an experiment that better answers the outstanding questions. 782 01:01:49,202 --> 01:01:59,409 um But as, you know, in so far as I go, my uh philosophy is always that the theory should try to be predictive of the experiment. 783 01:01:59,690 --> 01:02:10,077 And if it's not, then it's not a useful simulation outside of this, you know, idea of a synthetic experiment to benchmark your models. 784 01:02:10,077 --> 01:02:10,873 Yeah, yeah, yeah. 785 01:02:10,873 --> 01:02:13,513 Yeah, yeah, that makes sense. 786 01:02:13,993 --> 01:02:16,873 and yeah, I completely, completely agree with you. 787 01:02:16,873 --> 01:02:18,753 have a way of doing things for sure. 788 01:02:18,753 --> 01:02:20,613 Uh, that's also in my own work. 789 01:02:20,613 --> 01:02:23,513 How, how I do it in the end. 790 01:02:23,893 --> 01:02:28,313 Experiment and autism perpetuation are king. 791 01:02:28,313 --> 01:02:40,777 It's like if, if they don't go well, but the rest is looking good, then I, my, my belief in the model being an actuate depiction of reality is not. 792 01:02:40,777 --> 01:02:43,798 is definitely decreased by a lot. 793 01:02:43,818 --> 01:03:01,886 And something I am very curious about just to change gears a bit, a bit more general question, but it seems to me like high energy density physics can be contributing to a lot 794 01:03:01,886 --> 01:03:09,749 of different subfields of physics, uh astrophysics, planetary science, materials research even under extreme conditions. 795 01:03:09,749 --> 01:03:10,751 uh 796 01:03:10,751 --> 01:03:16,144 How do you see high energy density physics? 797 01:03:16,144 --> 01:03:17,045 It's hard for me. 798 01:03:17,045 --> 01:03:23,049 uh We usually just uh abbreviate it as HED physics, if that's easier. 799 01:03:23,049 --> 01:03:24,319 Okay, HED. 800 01:03:24,319 --> 01:03:26,090 That's going to be easier. 801 01:03:26,090 --> 01:03:35,936 So how do you see HED contributing to the different subfields and what are maybe some exciting projects that are looking into that currently? 802 01:03:35,936 --> 01:03:37,127 Yeah, absolutely. 803 01:03:37,127 --> 01:03:45,263 So, I I sort of alluded to this earlier, but you're recreating states that exist at the center of astrophysical objects. 804 01:03:45,563 --> 01:03:50,867 And so it turns astronomy from an observational science to one that you can do in the laboratory. 805 01:03:50,867 --> 01:04:00,434 ah And so one of the, this is sort of the focus of the Center for Matter Atomic Pressures, which is an NSF Frontier Center at the University of Rochester. 806 01:04:00,434 --> 01:04:03,336 um So we're, you know, 807 01:04:03,336 --> 01:04:09,882 It's using NSF money to get time on these lasers to study outstanding sort astrophysical questions. 808 01:04:10,122 --> 01:04:21,212 And sort of one of those uh questions is sort of dealing with the missibility of hydrogen and helium in these gas giant planets. 809 01:04:21,212 --> 01:04:29,992 So, you we know now from the Juno missions and other space probes that we don't really understand what's going on inside these gas giants inside Jupiter. 810 01:04:29,992 --> 01:04:32,854 particular, know, is the core, is it a solid core? 811 01:04:32,854 --> 01:04:34,194 Is it dilute? 812 01:04:34,215 --> 01:04:38,096 Turns out it's somewhere in the middle, where it's the sort of fuzzy core. 813 01:04:38,197 --> 01:04:43,640 And that sort of flies in the face of everything we thought we knew about planetary formation. 814 01:04:44,000 --> 01:04:55,747 And so, you know, these laser driven laboratory experiments are sort of, you know, positioned to help us understand uh things about how we, you know, how the component 815 01:04:55,747 --> 01:04:57,928 pieces of the center of Jupiter 816 01:04:57,928 --> 01:04:59,990 How do they behave under those immense pressures? 817 01:04:59,990 --> 01:05:05,714 You have ah millions and millions of atmospheres pressures at the center of Jupiter. 818 01:05:05,714 --> 01:05:08,857 And so how does hydrogen behave? 819 01:05:08,857 --> 01:05:11,199 It turns into a liquid metal. 820 01:05:11,199 --> 01:05:12,940 I mean, that's weird. 821 01:05:13,341 --> 01:05:14,511 We should know more about that. 822 01:05:14,511 --> 01:05:26,271 And so being able to create that state of matter in the laboratory and study how it behaves has huge implications for our understanding of these gas giants and other even 823 01:05:26,271 --> 01:05:27,772 terrestrial planets. 824 01:05:28,084 --> 01:05:29,495 ah Even our own Earth, right? 825 01:05:29,495 --> 01:05:36,927 There's a lot of really interesting geoscience being done at these laser facilities through CMAP and otherwise. 826 01:05:37,268 --> 01:05:40,349 And so that's one end. 827 01:05:40,349 --> 01:05:47,492 And then, know, Torit, you can go to these convergence systems like I have that can create stellar interior relevant conditions. 828 01:05:47,492 --> 01:05:56,346 And so you can start to understand, you know, our own sun, which there are still a lot of questions about, turns out, even though it's the closest star and we can make a lot of 829 01:05:56,346 --> 01:05:57,412 measurements of it. 830 01:05:57,412 --> 01:06:00,793 You can't really see, you can't send a probe into the center of the sun. 831 01:06:00,953 --> 01:06:03,534 Turns out it'll get incinerated. 832 01:06:03,534 --> 01:06:07,755 And so we don't know even what's, even there's some debate about what's in the sun. 833 01:06:07,755 --> 01:06:13,216 We're not a hundred percent sure what the metallicity, what elements are in the sun. 834 01:06:13,717 --> 01:06:22,529 And that stems largely from disagreement between spectroscopic measurements and uh helioseismological measurements. 835 01:06:22,529 --> 01:06:25,660 So that's measuring the oscillations of the sun. 836 01:06:25,956 --> 01:06:28,237 and inferring from that the composition. 837 01:06:28,477 --> 01:06:37,460 And so, you know, those are sort of two separate measurements, but they both require knowledge of, you know, high energy density plasmas and, you know, their equation of 838 01:06:37,460 --> 01:06:37,690 state. 839 01:06:37,690 --> 01:06:40,221 How does sound waves propagate through the sun? 840 01:06:40,221 --> 01:06:46,344 uh That requires knowledge of material properties of hydrogen plasmas at billions and billions of atmospheres. 841 01:06:46,344 --> 01:06:54,306 uh then, know, spectroscopy requires knowledge of atomic physics and distressing conditions, know, millions of degrees. 842 01:06:54,507 --> 01:06:55,487 And so, 843 01:06:55,731 --> 01:07:03,718 Being able to create these systems in the laboratories is hugely impactful for understanding the uh physics. 844 01:07:03,718 --> 01:07:05,579 Or rather, it could be. 845 01:07:05,740 --> 01:07:09,533 I don't think it's living up to its potential uh as of yet, right? 846 01:07:09,533 --> 01:07:20,382 And that large part of that is this, uh there's a large amount of siloing uh in physics in general, but especially in this field, because you have people who know how to do laser 847 01:07:20,382 --> 01:07:24,621 experiments, and you have people who know how to do astronomy or astrophysics. 848 01:07:24,621 --> 01:07:26,732 And they don't talk usually. 849 01:07:26,732 --> 01:07:28,773 They're in different subfields. 850 01:07:28,773 --> 01:07:34,617 They might meet at a conference, but they don't, you know, they're speaking different languages effectively. 851 01:07:34,617 --> 01:07:40,240 And so being able to, you know, bring them together and say, Hey, what are the astrophysical questions we can answer? 852 01:07:40,240 --> 01:07:43,381 You know, cause we're doing these experiments based on what we can do. 853 01:07:43,502 --> 01:07:47,034 And that's not always directly relevant to these astrophysical. 854 01:07:47,034 --> 01:07:53,079 you, you know, bringing in astrophysicists and say, you know, if you could snap your fingers and we could tell you one thing. 855 01:07:53,079 --> 01:07:56,120 about the matter at the center of Jupiter, what would it be? 856 01:07:56,120 --> 01:07:59,621 And then finding ways to answer that with laser experiments. 857 01:07:59,639 --> 01:08:11,984 And that's one of the missions of CMAP is, I think they're calling it in-reach as opposed to outreach, talking within the scientific community and trying to break down the walls of 858 01:08:11,984 --> 01:08:23,047 that silo because it's very difficult to merge disciplines when each one is so myopically focused on what they know and what they know how to do. 859 01:08:23,343 --> 01:08:35,799 uh And then I'll just say the last thing that we're sort of doing is sort of pushing ah the boundaries uh of these fundamental theories in physics, right? 860 01:08:35,799 --> 01:08:47,654 When you assemble matter to these extreme conditions, you start uh pushing towards the breakdown of these uh quantum mechanics, special relativity, not general relativity yet. 861 01:08:47,654 --> 01:08:48,834 We can't make black holes. 862 01:08:48,834 --> 01:08:50,095 Everyone always asks me. 863 01:08:50,095 --> 01:08:51,795 So can you make a black hole with this later? 864 01:08:51,795 --> 01:08:54,135 It's like, no, not even close, sorry. 865 01:08:54,295 --> 01:09:01,115 But we can sure get things hot enough where special relativity begins to play a role. 866 01:09:02,255 --> 01:09:13,515 And so what we're able to do by pushing the boundaries of these fundamental theories is to see where they break down. 867 01:09:13,635 --> 01:09:19,447 Quantum mechanics is sort of always considered in a vacuum, zero temperature, even isolated atom. 868 01:09:19,447 --> 01:09:23,458 What happens if you put that atom in a dense plasma that's 10 times the density of lead? 869 01:09:23,458 --> 01:09:30,070 It's very difficult to calculate, but we can do it in the laboratory and we can observe what happens. 870 01:09:30,590 --> 01:09:37,862 So I'll shout out work by my colleague, David Bischel, ah who is measuring, he's a spectroscopist. 871 01:09:37,862 --> 01:09:47,915 And so he's using implosions and doing spectroscopy on those and seeing how the presence of this incredibly dense plasma affects the energy levels of your atom. 872 01:09:48,140 --> 01:09:54,876 And it turns out that, you you think of the energy, the quantized energy levels of your atom is sort of fixed, right? 873 01:09:54,876 --> 01:09:57,187 know, 13.6 EV for hydrogen. 874 01:09:57,208 --> 01:10:00,851 But when you're in a dense plasma, they can actually shift. 875 01:10:00,851 --> 01:10:04,113 can move the energy levels of your atom. 876 01:10:04,113 --> 01:10:08,477 And that's, you know, kind of very complicated to understand and model. 877 01:10:08,477 --> 01:10:16,463 the free electrons are being sort of sucked inside by the nucleus and screening your energy levels so that the energy levels shift. 878 01:10:16,489 --> 01:10:24,806 And that's sort of changing how we think about quantum mechanics and energy levels and unbalanced states and free electrons and all that. 879 01:10:24,806 --> 01:10:29,720 And so there's a lot of interesting physics to be done with these giant lasers. 880 01:10:29,720 --> 01:10:37,736 And we're really only just scratching the surface, uh largely because we couldn't really interpret these measurements too easily until now. 881 01:10:37,797 --> 01:10:45,403 And we live in the information age now with all these diagnostics and Bayesian inference and machine learning. 882 01:10:46,111 --> 01:10:48,012 Lots of, lots of interesting stuff. 883 01:10:48,493 --> 01:10:49,074 Yeah. 884 01:10:49,074 --> 01:10:49,454 Yeah. 885 01:10:49,454 --> 01:10:51,676 This is, is absolutely incredible. 886 01:10:51,676 --> 01:10:54,718 Um, thanks for, for sharing that on the show. 887 01:10:54,718 --> 01:11:06,828 Definitely add any links that you think are going to be relevant, um, to the show notes because, I'm sure a lot of listeners are going to want to, uh, to learn more about the 888 01:11:06,828 --> 01:11:13,594 different, topics you just, I mean, research topics you just, you just talked about. 889 01:11:13,594 --> 01:11:15,235 Um. 890 01:11:15,519 --> 01:11:18,850 And so you've been generous with your time. 891 01:11:18,850 --> 01:11:20,621 So I'm going to start playing this out. 892 01:11:20,621 --> 01:11:34,885 I still have a lot of questions for you, but a main one is if, so you see, well, you just said we're living in a great time now when it comes to diagnostics. 893 01:11:34,885 --> 01:11:41,166 ah Do you see any new diagnostic technologies? 894 01:11:41,166 --> 01:11:43,569 ah Whether that's 895 01:11:43,569 --> 01:11:54,846 novel detectors or quantum sensors or anything else that you're excited about or keeping an eye on, not only for your work, but for your field in general? 896 01:11:55,087 --> 01:11:55,507 Yeah. 897 01:11:55,507 --> 01:11:59,580 Well, so, I mean, this is a great time for diagnostic development. 898 01:11:59,580 --> 01:12:03,933 We're, you know, at the, at the LLE, we're, we're developing new diagnostics all the time. 899 01:12:03,933 --> 01:12:09,417 You know, many, many PhD thesis are uh devoted to developing these new diagnostics. 900 01:12:09,417 --> 01:12:12,499 um In terms of. 901 01:12:12,713 --> 01:12:20,346 I think the most, you know, the one that I have my eye on uh is there at LEDA developing the next generation street camera. 902 01:12:20,346 --> 01:12:27,986 uh And so a street camera, basically, it's, you know, just a detector that takes a signal, an X-ray signal and sweeps it in time. 903 01:12:27,986 --> 01:12:31,261 So you get a continuous record of your X-ray. 904 01:12:31,261 --> 01:12:36,593 So you can use that to take a time resolved spectrum ah or for example, a time resolved image. 905 01:12:36,593 --> 01:12:42,035 ah And so they're very useful diagnostics, but they always break and 906 01:12:42,411 --> 01:12:46,393 ah You know, there's a joke going around that there's a conservation of street camera. 907 01:12:46,393 --> 01:12:50,775 So, you know, if one is working, that means somewhere in the world, there's another street camera that is not. 908 01:12:50,895 --> 01:12:55,157 And that holds fairly true for me in my experience. 909 01:12:55,397 --> 01:12:59,119 And so they're working on the next generation that's going to be very robust. 910 01:12:59,119 --> 01:13:04,222 ah it's Sean McPoyle is the graduate student who's developing it. 911 01:13:04,222 --> 01:13:07,103 um And so he's doing a very, 912 01:13:07,409 --> 01:13:16,490 thorough bottom-up approach, know, simulating the electron optics that are required to build this thing and, you know, building the best possible one, optimizing it from a very, 913 01:13:16,490 --> 01:13:19,082 you know, from first principles approach. 914 01:13:19,082 --> 01:13:28,925 And so that's going to enable really precise ultrafast uh spectroscopy of these, know, uh billionth of a second lived plasmas. 915 01:13:28,925 --> 01:13:32,606 And that's going to be able to, you know, totally push the boundaries of what we can measure. 916 01:13:32,606 --> 01:13:34,316 Cause I mean, even, you know, 917 01:13:34,552 --> 01:13:38,243 the officials work looking at these moving energy levels. 918 01:13:38,543 --> 01:13:43,425 That measurement is by modern spectroscopic standpoints terrible. 919 01:13:43,425 --> 01:13:45,725 The resolution is not super great. 920 01:13:45,725 --> 01:13:50,707 The time resolution is not, you can see that the lines move, you don't get much more than that. 921 01:13:50,707 --> 01:13:58,945 ah And so this next generation is going to be much more, it's going to be higher resolution, it's going to be better contrast. 922 01:13:58,945 --> 01:14:00,839 And so that's something that I'm really excited about. 923 01:14:00,839 --> 01:14:03,340 It's called the BHX spectrometer. 924 01:14:03,622 --> 01:14:10,507 um The other thing that I'm really excited about, um let's see, we have time resolved X-ray diffraction. 925 01:14:10,507 --> 01:14:15,851 ah So one thing that we do a lot of, know, X-ray diffraction, standard in material science. 926 01:14:15,851 --> 01:14:25,908 um We also do it at this sort of nanosecond scale where we blast a sample with a laser and then we hit it with a bunch of X-rays and you can measure the diffraction peaks that tell 927 01:14:25,908 --> 01:14:27,039 you what structure it's in. 928 01:14:27,039 --> 01:14:32,312 um And so that's told us, you know, we've discovered new uh states of matter. 929 01:14:32,546 --> 01:14:33,847 using that technique. 930 01:14:33,847 --> 01:14:38,309 uh And so it, you know, but it's a single measurement. 931 01:14:38,309 --> 01:14:48,223 And so now we're developing the capabilities to do time resolved diffraction where you're measuring the movement of those Bragg diffraction peaks with time as you compress it. 932 01:14:48,223 --> 01:14:56,946 So you can see phase transitions happen, which is a really outstanding question in material science in general is how do phase transitions happen and how do they happen at 933 01:14:56,946 --> 01:14:58,537 extreme conditions? 934 01:14:58,537 --> 01:15:00,758 And so that's going to the very 935 01:15:00,780 --> 01:15:05,554 you know, watching that closely to see what insights that develops. 936 01:15:05,554 --> 01:15:13,911 And then, know, last thing is, you know, the sort of development of high rep rate, you know, diagnostic technologies. 937 01:15:13,911 --> 01:15:19,305 Cause right now we're in a, you know, we're sitting on sort of a pond of data. 938 01:15:19,305 --> 01:15:22,328 You know, we've been operating the Omega laser since the nineties. 939 01:15:22,328 --> 01:15:27,272 There's a lot of data, but we're taking 10 shots a day, one shot an hour basically. 940 01:15:27,372 --> 01:15:29,732 And that produces a lot of data, but it's, you know, 941 01:15:29,732 --> 01:15:32,812 In terms of, know, compared to particle physics, it's nothing. 942 01:15:32,812 --> 01:15:34,912 Like I could store most of the data. 943 01:15:34,912 --> 01:15:37,771 I can store all the data I've taken from my PhD on my laptop. 944 01:15:37,771 --> 01:15:39,812 And it's not a, not an issue. 945 01:15:40,192 --> 01:15:43,812 Um, but you know, the next generation of lasers coming out are going to be much faster. 946 01:15:43,812 --> 01:15:46,692 They're going to be, you know, dipole or a diode pumped. 947 01:15:47,072 --> 01:15:50,552 And then they're going to be able to fire, you know, on the Hertz scale. 948 01:15:50,552 --> 01:15:54,832 So we're in time, like one shot every five seconds or a minute even. 949 01:15:55,312 --> 01:15:56,411 And so. 950 01:15:56,547 --> 01:16:01,847 that is going to rapidly increase the amount of data that we have just in general. 951 01:16:02,007 --> 01:16:13,127 so understanding what to do with that is a challenge that I think is sort of uniquely suited for the sort of data science techniques that I've been developing and other 952 01:16:13,127 --> 01:16:15,947 collaborators have been working on to understand that. 953 01:16:15,947 --> 01:16:20,487 And then what can you do if you have a shot every five seconds? 954 01:16:20,487 --> 01:16:24,515 What do you do with that level of repetition rate? 955 01:16:24,515 --> 01:16:29,116 uh What science questions can you answer and then how do you process that data and deal with it? 956 01:16:29,116 --> 01:16:32,457 How do you even build a target that you can shoot that fast? 957 01:16:32,457 --> 01:16:36,598 And so there's a lot of, I think that's really the frontier of the field. 958 01:16:36,598 --> 01:16:46,701 It's where the field is headed is this high rep rate frontier uh to be able to start behaving more like other fields where you do a bunch of experiments and you can get very 959 01:16:46,701 --> 01:16:48,522 good statistical certainty. 960 01:16:48,522 --> 01:16:49,782 Yeah. 961 01:16:50,095 --> 01:16:54,283 Actually that reminds me earlier in the show, you were saying that 962 01:16:54,467 --> 01:17:02,021 Um, in one of your projects, the, the data changes with time and space. 963 01:17:02,332 --> 01:17:07,451 so I'm curious how, how are you modeling that if you already are? 964 01:17:09,091 --> 01:17:18,415 um Well, so, I mean, it goes at the end of the day, goes back to, you know, the hydrodynamic equations, Navier-Stokes, and then, you know, however you want to reduce it 965 01:17:18,415 --> 01:17:18,935 from there. 966 01:17:18,935 --> 01:17:30,220 um And so understanding how it evolves with time and space is, you know, a very non-trivial physics problem, especially, you know, when you're dealing with laser plasma 967 01:17:30,220 --> 01:17:37,163 coupling and then, you know, beyond just the basic physics questions of what's the equation of state of this thing. 968 01:17:37,219 --> 01:17:43,199 How does a laser, now you introduce a laser field and transport properties are even less well understood. 969 01:17:44,099 --> 01:17:51,779 And so, you know, I've sort of adopted of, you know, try and keep things as simple as possible. 970 01:17:51,779 --> 01:17:54,939 So Newton's, know, keep things at like a Newton's second law level, right? 971 01:17:54,939 --> 01:17:57,219 You balance your forces, right? 972 01:17:57,219 --> 01:18:02,699 The laser is exerting some force on your target and then that necessarily produces an acceleration. 973 01:18:02,699 --> 01:18:06,403 And so one nice thing about Bayesian inference is you can have these parameters. 974 01:18:06,403 --> 01:18:08,523 that you infer from your data. 975 01:18:08,523 --> 01:18:16,003 So if you have a trajectory of your imploding capsule, you can say, OK, well, I know the acceleration from these images that I've taken. 976 01:18:16,003 --> 01:18:19,063 I can see this thing converging. 977 01:18:19,063 --> 01:18:21,523 So I know it's going inward. 978 01:18:21,523 --> 01:18:23,783 There must be some force driving that acceleration. 979 01:18:23,783 --> 01:18:26,363 I can refer that from Bayesian inference. 980 01:18:26,363 --> 01:18:34,703 so trying to keep the models as simple as possible without so you don't have to solve the full Navier-Stokes equation with radiative transfer and all that. 981 01:18:36,675 --> 01:18:45,308 The level that we're at in the future, you know, we'd like to be able to use these state-of-the-art um Codes these radiation hydrodynamic codes as the model right because 982 01:18:45,308 --> 01:18:56,355 that's a complete picture of all the physics involved The problem is there's you know, they're slow, you know, they're taking hours to run one simulation and they're also you 983 01:18:56,355 --> 01:19:05,333 know, there's a lot of parameters in these very complicated codes as you can imagine and So, you know the sort of future is to build a sort of Bayesian version 984 01:19:05,499 --> 01:19:13,467 of one of these codes that's ideally written in Python, ah fully differentiable with respect to the parameters and much faster. 985 01:19:13,467 --> 01:19:19,433 And probably that looks something like building a neural network surrogate ah of one of these codes or something. 986 01:19:19,433 --> 01:19:27,581 um We'll see what the future holds, but that's the dream is to move to more and more complex physical models. 987 01:19:27,581 --> 01:19:28,700 uh 988 01:19:28,700 --> 01:19:32,301 sort of trying to follow the blueprint that the particle physics people do. 989 01:19:32,301 --> 01:19:36,322 They have this very complicated, complete physical model and they use that to make predictions. 990 01:19:36,322 --> 01:19:42,214 And so we'd like to be able to do the same and then infer using the best available models. 991 01:19:43,034 --> 01:19:46,095 for now, very simple, very simple pictures. 992 01:19:46,135 --> 01:19:46,770 Yeah. 993 01:19:46,770 --> 01:19:52,016 I mean, you have to start somewhere, especially with all the complexities you were talking about before. 994 01:19:52,537 --> 01:19:56,538 And the block building philosophy, you were... 995 01:19:56,538 --> 01:19:57,832 uh 996 01:19:57,832 --> 01:20:03,407 talking about, which I definitely agree with, because that way, know, what is working and what is not. 997 01:20:03,407 --> 01:20:04,618 So that's extremely helpful. 998 01:20:04,618 --> 01:20:10,173 Now I was asking because me, I hear spatial temporal and I think about Gaussian processes, of course. 999 01:20:10,173 --> 01:20:11,294 you know, like I was curious. 1000 01:20:11,294 --> 01:20:12,334 of course. 1001 01:20:12,675 --> 01:20:16,628 And that's part of the simple model, right? 1002 01:20:16,628 --> 01:20:18,730 When you don't know how something evolves in time, right? 1003 01:20:18,730 --> 01:20:19,981 The temperature is a great example. 1004 01:20:19,981 --> 01:20:25,756 That's, you know, depends on all of the thermal processes, conduction, radiative heating, you know, expansion. 1005 01:20:25,756 --> 01:20:28,567 All of those things go into that and it's very difficult to model that. 1006 01:20:28,567 --> 01:20:32,808 ah But you know, the temperature goes up and then down. 1007 01:20:32,808 --> 01:20:37,379 And so the perfect sort of vehicle for that is of course, modeling as a Gaussian process. 1008 01:20:37,379 --> 01:20:42,601 And that's an infinitely flexible tool for that and something we've we've leaned on a lot. 1009 01:20:42,601 --> 01:20:43,031 Yes. 1010 01:20:43,031 --> 01:20:45,771 I forgot you're big fan of Gaussian processes. 1011 01:20:45,771 --> 01:20:47,652 I should have led with that. 1012 01:20:48,112 --> 01:20:48,512 I am. 1013 01:20:48,512 --> 01:20:48,932 am. 1014 01:20:48,932 --> 01:20:55,684 So did you try actually some GPs on that or that actually doesn't work because 1015 01:20:55,684 --> 01:21:05,921 When the data is too big or even better justification would be, well, we know more structure than we can impose on a GP. 1016 01:21:06,341 --> 01:21:06,902 No, yeah. 1017 01:21:06,902 --> 01:21:09,703 Well, so we use Gaussian processes. 1018 01:21:09,914 --> 01:21:15,087 I mean, I use Gaussian processes at every available opportunity just because they're a very powerful tool. 1019 01:21:15,087 --> 01:21:19,871 So what we'll often do is model a piece of the model. 1020 01:21:19,871 --> 01:21:23,873 For example, the temperature history is a function of time as a Gaussian process, right? 1021 01:21:23,873 --> 01:21:24,988 m 1022 01:21:24,988 --> 01:21:28,430 That's sort of, know, there's stuff that we know more about, right? 1023 01:21:28,430 --> 01:21:35,583 That the Newton's second law piece, I was talking earlier, there's physics pieces that we know and when possible we use the physics. 1024 01:21:35,943 --> 01:21:41,225 But you know, when you can't, you want to use something that is flexible and able to capture the data. 1025 01:21:41,225 --> 01:21:45,467 ah Well, you know, we know the temperature is a continuous function of time. 1026 01:21:45,467 --> 01:21:49,469 And so Gaussian process is the perfect sort of prior to place on that. 1027 01:21:49,949 --> 01:21:51,760 And sort of, you know. 1028 01:21:51,760 --> 01:21:56,994 lets the data inform what that is while giving an estimate of the uncertainty. 1029 01:21:59,417 --> 01:22:00,898 Yeah. 1030 01:22:00,898 --> 01:22:01,668 Great. 1031 01:22:01,668 --> 01:22:04,040 This is a very interesting thing to do. 1032 01:22:04,230 --> 01:22:15,670 And both in Amparo and Pinesy, have very good GP modules that you can use flexibly um in your models without forcing the model to be a pure GP. 1033 01:22:15,670 --> 01:22:19,293 So this is something extremely, extremely helpful. 1034 01:22:19,293 --> 01:22:20,854 um 1035 01:22:21,794 --> 01:22:27,198 Maybe a few more questions before I ask you the last two questions. 1036 01:22:27,198 --> 01:22:34,042 I promise I won't be too long, but I'm curious now, playing this out, thinking a bit more about the future. 1037 01:22:34,042 --> 01:22:37,624 What's your vision for scaling up experiments? 1038 01:22:37,925 --> 01:22:42,127 Whether that's energy, repetition rate, fidelity. 1039 01:22:42,308 --> 01:22:51,233 In the next few years, what do you hope the field will get when it comes to scaling up experiments? 1040 01:22:51,521 --> 01:22:54,344 Yeah, well, so it's going to be a sort of two pronged approach. 1041 01:22:54,344 --> 01:23:01,609 think what we're already seeing is a move towards high rep rate at existing facilities. 1042 01:23:01,810 --> 01:23:08,295 So the sort of most prominent example is the dipole laser at the European Exfel in Hamburg, Germany. 1043 01:23:08,295 --> 01:23:16,102 And so they've built a reparated sort of Joule class laser that can do, you know, one shot. 1044 01:23:16,102 --> 01:23:19,945 It's supposed to be, I think one Hertz, but it's not quite there. 1045 01:23:19,993 --> 01:23:27,618 ah It can shoot every couple minutes or so, which is much better than the previous version, which had like an hour cool time or whatever. 1046 01:23:27,618 --> 01:23:36,083 And so now you can just rattle through a bunch of different targets and you can do someone's whole PhD thesis in an hour. 1047 01:23:36,083 --> 01:23:45,058 ah And so I think moving that direction, this high rep rate is very, um that's where we're heading. 1048 01:23:45,058 --> 01:23:46,861 And I think that's a good direction to head. 1049 01:23:46,861 --> 01:23:51,082 The other piece is going to be, we're going to need to build new facilities. 1050 01:23:51,082 --> 01:23:56,024 And, you know, the Omega laser is, as I mentioned several times, very old. 1051 01:23:56,024 --> 01:23:57,104 It's from the nineties. 1052 01:23:57,104 --> 01:23:59,305 It's using nineties laser technology. 1053 01:23:59,305 --> 01:24:00,345 It's showing its age. 1054 01:24:00,345 --> 01:24:02,906 um And it's a really good facility. 1055 01:24:02,906 --> 01:24:03,846 It does really good work. 1056 01:24:03,846 --> 01:24:11,828 um But we're going to have to start thinking about the next generation of laser drivers to be able to sort of keep pushing the boundaries. 1057 01:24:11,828 --> 01:24:14,209 And so obviously the national ignition facility. 1058 01:24:14,209 --> 01:24:17,151 out at Livermore is uh mega joules. 1059 01:24:17,151 --> 01:24:21,973 So it's 10 times bigger than the Omega 60 laser, which is only 30 kilojoules. 1060 01:24:22,273 --> 01:24:30,869 And so that serves its own purpose, but you can only shoot that once a day because it's so massive, the lamps take forever to cool down. 1061 01:24:30,869 --> 01:24:38,292 And then you create so many neutrons from your fusion reactions that it uh makes the whole chamber radioactive. 1062 01:24:38,292 --> 01:24:43,929 so there, think, is a middle ground here sort of between the 1063 01:24:43,929 --> 01:24:53,883 kilojoule class laser that Omega is now and the mega joule laser at NIF, you know, probably like a hundred kilojoules is probably the sweet spot uh where it's also, you you 1064 01:24:53,883 --> 01:25:02,907 were, you're able to rep rate that and do all of, all of the, the cool science that the Omega 60 laser does now uh and more, right? 1065 01:25:02,907 --> 01:25:11,141 Cause you have a little bit more energy, but it being able to do that at scale at rep rate, I think is going to do, would be a very worthwhile endeavor. 1066 01:25:11,141 --> 01:25:21,420 And we'll see there, you know, they're in discussions to build the next generation of Omega 60 which is sort of tentatively named the unimaginative Omega next Which I think is 1067 01:25:21,420 --> 01:25:33,160 a terrible name and needs to be changed before we start proposing it to the government but um Yeah, I think that's that's the way the way forward is reparate and you know more 1068 01:25:33,160 --> 01:25:38,825 energy But you know, I don't think we need to go, you know, people are talking about building nif to 1069 01:25:39,126 --> 01:25:44,270 uh Right, because now that they got ignition over there, we need to build a 10-megajoule laser. 1070 01:25:44,270 --> 01:25:50,213 And that's great, but I think there's a lot more interesting scientific questions that can be answered with a 100-kilojoule laser. 1071 01:25:50,213 --> 01:25:54,877 uh Like Omega 60, but Omega next. 1072 01:25:54,877 --> 01:25:55,897 Damn, yeah. 1073 01:25:55,897 --> 01:26:00,981 Yeah, that sounds like a powerful beast to build. 1074 01:26:00,981 --> 01:26:06,384 um Now, one last question before the last two questions. 1075 01:26:06,384 --> 01:26:07,845 um I know... 1076 01:26:08,327 --> 01:26:16,368 So you're, you're maybe starting to teach it or at least you're, you're doing some Bayesian outreach as you were saying. 1077 01:26:16,368 --> 01:26:24,121 So I'm curious what advice do you younger scientists entering the field? 1078 01:26:24,441 --> 01:26:29,363 What skills, mindset, training do you think are most critical? 1079 01:26:30,063 --> 01:26:30,413 Yeah. 1080 01:26:30,413 --> 01:26:37,778 Well, I think it's one thing I wish, you know, people had told me, and I think a misconception about 1081 01:26:37,778 --> 01:26:43,872 you know, grad school in general, is I tell people I'm doing, you know, I PhD in physics and I'm like, you must be so smart. 1082 01:26:43,872 --> 01:26:45,453 You must be a genius. 1083 01:26:45,573 --> 01:26:48,155 And it's like, I don't think that's necessarily true. 1084 01:26:48,155 --> 01:26:52,398 I think more than anything, it's mental toughness is what you need to cause it. 1085 01:26:52,398 --> 01:26:54,319 Grad school is a grind. 1086 01:26:54,479 --> 01:26:57,321 And there are, there are days when you're like, I hate this. 1087 01:26:57,321 --> 01:26:58,001 I want to quit. 1088 01:26:58,001 --> 01:27:02,945 Um, and you know, there are, you know, just as many days where like, you know, this is great. 1089 01:27:02,945 --> 01:27:06,597 Like I, I feel like I'm unraveling truths of the universe. 1090 01:27:06,799 --> 01:27:13,271 And being able to sort of manage those highs and lows takes a remarkable amount of mental toughness. 1091 01:27:13,271 --> 01:27:21,883 And so what I've seen over the course of my career is graduate students who are successful are ones who are able to grind it out. 1092 01:27:22,123 --> 01:27:24,444 And it helps to have a sense of community. 1093 01:27:24,444 --> 01:27:30,585 And so that would be my other piece of advice is make friends with your lab mates and fellow grad students. 1094 01:27:30,926 --> 01:27:34,746 it'll get better, allegedly. 1095 01:27:36,903 --> 01:27:44,430 it's, it's important to, to have mental toughness and not lose sort of the big picture, uh, side of it all. 1096 01:27:44,430 --> 01:27:44,620 Right. 1097 01:27:44,620 --> 01:27:51,697 Cause it, you know, well, you're doing cutting edge science in, it's, it's very, uh, you know, remarkable in that sense. 1098 01:27:51,697 --> 01:27:57,512 And so being able to not lose sight of that while maintaining your, your toughness is, is I think that's the key to success. 1099 01:27:57,512 --> 01:27:59,163 That's, that's my advice. 1100 01:28:01,073 --> 01:28:01,373 Yeah. 1101 01:28:01,373 --> 01:28:02,243 Yeah. 1102 01:28:02,243 --> 01:28:16,834 So resilience and compounding effect of small daily grind is, definitely something you recommend and definitely push to with my own students. 1103 01:28:16,834 --> 01:28:20,617 Like, because I get the same, yeah, I guess the same reactions. 1104 01:28:20,617 --> 01:28:27,372 Um, each time I talk to people about what I do and it's always like, oh yeah, I'm not good at math. 1105 01:28:27,372 --> 01:28:29,183 know, I've never, it's not for me. 1106 01:28:29,183 --> 01:28:29,939 Um. 1107 01:28:29,939 --> 01:28:38,379 And I think trying to develop more of that growth mindset where it's well, you know, I'm not sure it's for me either. 1108 01:28:38,379 --> 01:28:41,579 You know, it's just, I, I love it and I love learning new things. 1109 01:28:41,579 --> 01:28:48,379 And I think working on difficult problems is the best way to learn new things because otherwise I get bored. 1110 01:28:48,379 --> 01:28:56,811 But this is extremely uncomfortable and hard because by definition, if you're always at the frontier of the field you work in. 1111 01:28:57,115 --> 01:28:59,016 You're working on things that aren't solved yet. 1112 01:28:59,016 --> 01:29:10,224 And so you don't have that feeling of, uh, yeah, being comfortable because you never know exactly how you're going to solve what you're working on. 1113 01:29:10,224 --> 01:29:12,665 And so, even if there is a solution. 1114 01:29:12,946 --> 01:29:13,516 Yeah. 1115 01:29:13,516 --> 01:29:14,657 Well, so that's even worse. 1116 01:29:14,657 --> 01:29:15,327 Yeah. 1117 01:29:15,327 --> 01:29:18,290 That's not up in that Pandora's box, but yeah, yeah, no, for sure. 1118 01:29:18,290 --> 01:29:21,055 Um, yeah, you don't even know if there is a solution. 1119 01:29:21,055 --> 01:29:26,715 Um, and, and so, yeah, I think having. 1120 01:29:27,347 --> 01:29:33,727 loving what you're working on because it gives you something you really value. 1121 01:29:34,607 --> 01:29:47,287 And just being able to persistently get point one percent better every day is going to give you a very valuable compounding effect. 1122 01:29:47,907 --> 01:29:49,207 100 % agree. 1123 01:29:50,767 --> 01:29:51,187 Awesome. 1124 01:29:51,187 --> 01:29:57,013 Well, Ethan, I think it's time to let you go because as you've seen in the Google doc app. 1125 01:29:57,013 --> 01:30:00,546 I have more questions, but let's call it a show. 1126 01:30:00,546 --> 01:30:06,090 um But before letting you go, of course, you know, I'm going to ask you the last two questions. 1127 01:30:06,090 --> 01:30:07,932 I ask every guest at the end of the show. 1128 01:30:07,932 --> 01:30:14,997 ah So first one, if you had unlimited time and resources, which problem would you try to solve? 1129 01:30:15,518 --> 01:30:18,100 So, all right, this is a bit of a niche one. 1130 01:30:18,921 --> 01:30:25,886 But I'm very interested in this idea of picnonuclear fusion. 1131 01:30:26,438 --> 01:30:28,610 And so, you know, most people are familiar with thermonuclear fusion. 1132 01:30:28,610 --> 01:30:29,921 That's what happens in the sun, right? 1133 01:30:29,921 --> 01:30:33,443 have, you know, hydrogen molecules, very hot, they crash into each other and they fuse. 1134 01:30:33,443 --> 01:30:45,111 ah There's another mechanism that's believed to occur in some super dense astrophysical objects, like in white dwarfs, for example, where you have sufficient density that you 1135 01:30:45,111 --> 01:30:52,456 have, you know, two atoms that are close enough together that they spontaneously fuse, not due to, you know, high velocity collisions, but just because of proximity. 1136 01:30:52,776 --> 01:30:55,818 And if it, you know, it sounds a lot like a cold fusion. 1137 01:30:55,948 --> 01:30:59,941 But this is a little bit more uh scientifically sound. 1138 01:30:59,941 --> 01:31:09,929 And so there's a longstanding uh belief of my advisor that you can do this on a laser facility like the National Ignition Facility. 1139 01:31:10,290 --> 01:31:13,812 But you would need to make it incredibly high density. 1140 01:31:14,453 --> 01:31:18,497 incredibly, you'd have to keep it cold enough that thermonuclear fusion doesn't happen. 1141 01:31:18,497 --> 01:31:25,272 And so if the government gave me unlimited time and money, I would just devote the entire time of the National Ignition Facility. 1142 01:31:25,272 --> 01:31:28,343 to trying to make as dense of an object as I could. 1143 01:31:28,343 --> 01:31:31,334 And maybe we'd get the black hole that everyone always asks me about. 1144 01:31:31,334 --> 01:31:38,786 But ideally, you'd show this peak no nuclear fusion is, I think, one of the coolest things. 1145 01:31:38,786 --> 01:31:45,368 um And it's a largely unsolved problem because we have no idea what's happening in those white wars. 1146 01:31:45,368 --> 01:31:49,209 And we can't yet make that plasma in the laboratory. 1147 01:31:49,209 --> 01:31:50,629 So that's the problem. 1148 01:31:50,629 --> 01:31:54,090 It's personal pet project. 1149 01:31:55,926 --> 01:31:56,346 Okay. 1150 01:31:56,346 --> 01:31:58,127 Yeah. 1151 01:31:58,127 --> 01:31:58,887 I love that. 1152 01:31:58,887 --> 01:32:05,789 um niche answers are definitely encouraged to that question. 1153 01:32:05,789 --> 01:32:08,230 em Second question. 1154 01:32:08,230 --> 01:32:15,371 If you could have dinner with any great scientific mind, dead alive, alive or fictional, who would it be? 1155 01:32:15,432 --> 01:32:24,394 So this is going to be controversial given the, he was kind of a villain in Oppenheimer, but I'm going to go with Ed Teller. 1156 01:32:24,702 --> 01:32:36,562 Uh, just because, you know, he's, he's sort of a controversial figure, uh, in, scientific history, but I think he, I think he had a very interesting view on science and sort of, 1157 01:32:36,562 --> 01:32:37,837 uh, interesting mind. 1158 01:32:37,837 --> 01:32:39,728 And I would love to sort of pick his brain. 1159 01:32:39,728 --> 01:32:50,593 He also, um, you know, a lot of the work that he did, uh, you know, post Manhattan project set laid the groundwork for, you know, the high energy density physics as we know it 1160 01:32:50,593 --> 01:32:51,353 today, right? 1161 01:32:51,353 --> 01:32:54,190 He was using nuclear bombs and set of giant lasers. 1162 01:32:54,190 --> 01:32:55,451 uh as a driver. 1163 01:32:55,451 --> 01:33:00,423 um But it would be interesting to sort of pick his brain about the foundations of the field. 1164 01:33:00,423 --> 01:33:07,898 And of course, he helped invent the uh Metropolis-Hastings uh algorithm that is the foundation of all Bayesian inference. 1165 01:33:07,898 --> 01:33:10,119 I'd be interested to talk to him about that. 1166 01:33:10,119 --> 01:33:10,619 Damn. 1167 01:33:10,619 --> 01:33:11,220 Yeah. 1168 01:33:11,220 --> 01:33:12,700 Yeah, great answer. 1169 01:33:12,700 --> 01:33:17,043 And I think you're the first one to actually answer that. 1170 01:33:17,043 --> 01:33:18,564 I would not be surprised. 1171 01:33:18,564 --> 01:33:20,304 I don't think that's a very popular answer. 1172 01:33:20,304 --> 01:33:23,453 um 1173 01:33:23,453 --> 01:33:27,214 Awesome, well, Ethan, thanks again for being on the show. 1174 01:33:27,214 --> 01:33:28,595 That was fantastic. 1175 01:33:28,595 --> 01:33:31,666 You were able to cover so much ground. 1176 01:33:31,906 --> 01:33:34,687 I am very happy about that. 1177 01:33:34,807 --> 01:33:36,478 You're doing fascinating work. 1178 01:33:36,478 --> 01:33:40,489 So I am extremely happy to have you on the show. 1179 01:33:40,489 --> 01:33:45,151 Let's make sure to add all these great resources to the show notes. 1180 01:33:45,151 --> 01:33:47,332 I've already added a few of them, folks. 1181 01:33:47,332 --> 01:33:52,654 So make sure to check them out if you want to dig deeper. 1182 01:33:53,104 --> 01:33:57,622 Ethan, thanks again for taking the time and being on this show. 1183 01:33:57,622 --> 01:33:58,809 Thank you so much for having me. 1184 01:33:58,809 --> 01:34:02,690 And thanks again to JJ Ruby for putting this in. 1185 01:34:07,028 --> 01:34:10,721 This has been another episode of Learning Bajan Statistics. 1186 01:34:10,721 --> 01:34:21,230 Be sure to rate, review and follow the show on your favorite podcatcher and visit learnbastats.com for more resources about today's topics as well as access to more 1187 01:34:21,230 --> 01:34:25,313 episodes to help you reach true Bajan state of mind. 1188 01:34:25,313 --> 01:34:27,255 That's learnbastats.com. 1189 01:34:27,255 --> 01:34:32,119 Our theme music is Good Bajan by Baba Brinkman, fit MC Lars and Meghiraan. 1190 01:34:32,119 --> 01:34:35,261 Check out his awesome work at bababrinkman.com. 1191 01:34:35,261 --> 01:34:36,456 I'm your host. 1192 01:34:36,456 --> 01:34:37,517 Alex and Dora. 1193 01:34:37,517 --> 01:34:41,666 can follow me on Twitter at Alex underscore and Dora like the country. 1194 01:34:41,666 --> 01:34:48,935 You can support the show and unlock exclusive benefits by visiting Patreon.com slash LearnBasedDance. 1195 01:34:48,935 --> 01:34:51,316 Thank you so much for listening and for your support. 1196 01:34:51,316 --> 01:34:53,618 You're truly a good Bayesian. 1197 01:34:53,618 --> 01:35:00,422 Change your predictions after taking information and if you're thinking I'll be less than amazing. 1198 01:35:00,422 --> 01:35:03,725 Let's adjust those expectations. 1199 01:35:03,725 --> 01:35:06,234 Let me show you how to be a good Bayesian. 1200 01:35:06,234 --> 01:35:16,937 You change calculations after taking fresh data in Those predictions that your brain is making Let's get them on a solid foundation