Fuld transskription af ASU-Science Prize winners talk transforming messy datasets into solutions [BONUS]

to make sense of messy data 41 00:01:49,290 --> 00:01:50,580 and change human lives. 42 00:01:59,940 --> 00:02:01,920 Welcome, Meha. Thank you for joining me. 43 00:02:01,920 --> 00:02:04,500 Thanks so much for the invitation. I'm really excited. 44 00:02:04,500 --> 00:02:06,900 Well, I'm really glad to have you here today joining us 45 00:02:06,900 --> 00:02:08,400 all the way from Australia. 46 00:02:08,400 --> 00:02:10,290 So thank you for making the trip. 47 00:02:10,290 --> 00:02:14,910 Your work sits right at that interface of climate change 48 00:02:14,910 --> 00:02:17,610 and agriculture, so obviously a vital importance. 49 00:02:17,610 --> 00:02:21,090 Our food security is a key worry, I think, 50 00:02:21,090 --> 00:02:23,070 for many countries right now. 51 00:02:23,070 --> 00:02:25,590 And your work focuses on the global south 52 00:02:25,590 --> 00:02:29,400 and a group that you call small holder agriculture. 53 00:02:29,400 --> 00:02:31,920 Can you define what small holder is for us? 54 00:02:31,920 --> 00:02:34,380 So the way that I define small holder is one 55 00:02:34,380 --> 00:02:35,640 is by field size. 56 00:02:35,640 --> 00:02:36,990 So typically these are fields 57 00:02:36,990 --> 00:02:39,750 that are less than two hectares in size 58 00:02:39,750 --> 00:02:42,810 and largely I'm also working with family farms. 59 00:02:42,810 --> 00:02:46,530 So these are, you know, whole families working together 60 00:02:46,530 --> 00:02:48,780 to produce agricultural production, both 61 00:02:48,780 --> 00:02:51,420 for home consumption but also to sell to the market. 62 00:02:51,420 --> 00:02:52,590 Right. And this is most 63 00:02:52,590 --> 00:02:54,750 of the production in the global south, is that right? 64 00:02:54,750 --> 00:02:59,230 Exactly, yeah. In India I think it's about 90% 65 00:02:59,230 --> 00:03:01,000 or more of production is under these. 66 00:03:01,000 --> 00:03:02,260 Wow. These systems. 67 00:03:02,260 --> 00:03:04,450 So your work, as I understand it, 68 00:03:04,450 --> 00:03:07,930 uses great technology like satellite imagery 69 00:03:07,930 --> 00:03:10,690 and AI to try to understand 70 00:03:10,690 --> 00:03:14,440 how climate change is affecting these food producers 71 00:03:14,440 --> 00:03:18,550 and how they can adjust to changing regime right now. 72 00:03:18,550 --> 00:03:20,590 Can you tell us more about like 73 00:03:20,590 --> 00:03:22,930 what has AI brought to this project? 74 00:03:22,930 --> 00:03:25,450 What is the addition that AI's been able to bring? 75 00:03:25,450 --> 00:03:28,810 So one challenge that we face in small holder systems is 76 00:03:28,810 --> 00:03:30,490 that they're typically data scarce. 77 00:03:30,490 --> 00:03:33,250 So unlike places like the US where you might have a lot 78 00:03:33,250 --> 00:03:35,080 of ground data collected by the government 79 00:03:35,080 --> 00:03:38,080 or different agencies, you don't always see that, see 80 00:03:38,080 --> 00:03:39,760 that in the regions where I work. 81 00:03:39,760 --> 00:03:42,940 So one thing that's been really great about AI is we're 82 00:03:42,940 --> 00:03:45,190 trying to use methods that can allow us 83 00:03:45,190 --> 00:03:49,120 to create high-accuracy models using less training data. 84 00:03:49,120 --> 00:03:51,670 Big picture data allow us to see patterns that were sort 85 00:03:51,670 --> 00:03:52,840 of previously invisible. 86 00:03:52,840 --> 00:03:54,970 And you have this nice mixed methods approach. 87 00:03:54,970 --> 00:03:57,550 We're using sort of small and large-scale data, 88 00:03:57,550 --> 00:03:59,080 but how do you get that information 89 00:03:59,080 --> 00:04:02,380 and solution sort of to farmers on the ground? 90 00:04:02,380 --> 00:04:06,190 So I largely actually work with organizations 91 00:04:06,190 --> 00:04:07,450 who are working with farmers. 92 00:04:07,450 --> 00:04:10,390 So I'm not personally directly providing the information 93 00:04:10,390 --> 00:04:12,850 to farmers, and that's for several reasons. 94 00:04:12,850 --> 00:04:16,210 So there are so many wonderful organizations on the ground 95 00:04:16,210 --> 00:04:19,570 who have these close relationships established with farmers, 96 00:04:19,570 --> 00:04:23,260 have wonderful kind of networks in place 97 00:04:23,260 --> 00:04:26,200 and information dissemination techniques in place to, 98 00:04:26,200 --> 00:04:28,840 to work with local farmers. 99 00:04:28,840 --> 00:04:33,100 So really I work with those organizations to help them 100 00:04:34,120 --> 00:04:36,400 identify potential solutions, identify 101 00:04:36,400 --> 00:04:38,410 where they can be targeted at scale, 102 00:04:38,410 --> 00:04:40,810 and then they're the ones who actually implement. 103 00:04:40,810 --> 00:04:43,030 And it's amazing that you're networked in that way. 104 00:04:43,030 --> 00:04:45,310 I work with a lot of faculty in my role 105 00:04:45,310 --> 00:04:49,360 who have incredible ideas that they're pursuing 106 00:04:49,360 --> 00:04:52,385 in their laboratory, in their research, 107 00:04:52,385 --> 00:04:56,650 but they don't know how to translate that, if you will, how 108 00:04:56,650 --> 00:04:58,960 to take that into the user space 109 00:04:58,960 --> 00:05:00,820 and make sure that it's actually being implemented 110 00:05:00,820 --> 00:05:03,940 or having an impact, be that a new technological thing 111 00:05:03,940 --> 00:05:07,180 or a solution like you're finding that someone can apply. 112 00:05:07,180 --> 00:05:09,610 How did you get networked with the, with those folks 113 00:05:09,610 --> 00:05:11,350 that are in country on the ground? 114 00:05:11,350 --> 00:05:15,040 So I've been very, very fortunate to have been linked up 115 00:05:15,040 --> 00:05:17,500 with amazing organizations that are doing this work 116 00:05:17,500 --> 00:05:19,360 and are also really keen to collaborate. 117 00:05:19,360 --> 00:05:22,840 So the specific organizations that I

mostly work 118 00:05:22,840 --> 00:05:25,390 with now are through the CGIAR network, 119 00:05:25,390 --> 00:05:29,470 which is an international research organization working on 120 00:05:29,470 --> 00:05:32,140 agricultural development across the world. 121 00:05:32,140 --> 00:05:33,730 They're headquartered in Mexico, 122 00:05:33,730 --> 00:05:35,500 but then have local offices in many 123 00:05:35,500 --> 00:05:37,390 countries across the globe. 124 00:05:37,390 --> 00:05:40,090 And I became connected with this group during my postdoc. 125 00:05:40,090 --> 00:05:42,430 I did a postdoc at Stanford University 126 00:05:42,430 --> 00:05:44,140 and my advisor there had been working 127 00:05:44,140 --> 00:05:48,130 with this group broadly over the last 10, 15 years. 128 00:05:48,130 --> 00:05:49,750 And through him I got connected 129 00:05:49,750 --> 00:05:52,060 and it's been a really wonderful experience 130 00:05:52,060 --> 00:05:55,510 and the collaborations have just kind of flourished and 131 00:05:55,510 --> 00:05:56,920 and deepened since then. 132 00:05:56,920 --> 00:05:58,370 And you mentioned India earlier. 133 00:05:58,370 --> 00:05:59,840 Is it the region where you work? 134 00:05:59,840 --> 00:06:02,960 It's the primary focus of the essay that you wrote that is 135 00:06:02,960 --> 00:06:06,020 that we're featuring here. Why India? 136 00:06:06,020 --> 00:06:08,810 So India, I have both personal reasons 137 00:06:08,810 --> 00:06:11,270 and food security reasons that I look at it. 138 00:06:11,270 --> 00:06:14,180 So India is the largest producer of rice globally. 139 00:06:14,180 --> 00:06:15,890 It's the second largest producer of wheat. 140 00:06:15,890 --> 00:06:18,320 So it really is an agricultural powerhouse. 141 00:06:18,320 --> 00:06:20,030 At the same time, as I mentioned, 142 00:06:20,030 --> 00:06:21,440 it's largely being produced 143 00:06:21,440 --> 00:06:23,780 by these small holder production systems 144 00:06:23,780 --> 00:06:25,760 that are very vulnerable to climate change. 145 00:06:25,760 --> 00:06:28,070 And India is also facing significant 146 00:06:28,070 --> 00:06:29,240 environmental challenges. 147 00:06:29,240 --> 00:06:32,630 So India's the largest consumer of freshwater worldwide, 148 00:06:32,630 --> 00:06:35,210 largely being used for irrigation and agriculture 149 00:06:35,210 --> 00:06:38,060 and over 50% of its aquifers are depleted. 150 00:06:38,060 --> 00:06:40,130 At the same time, it's also a hotspot 151 00:06:40,130 --> 00:06:41,900 for crop residue burning, which leads 152 00:06:41,900 --> 00:06:44,480 to really low air quality events in 153 00:06:44,480 --> 00:06:46,130 India, usually in the fall. 154 00:06:46,130 --> 00:06:49,910 So it's really a hotspot to examine these questions 155 00:06:49,910 --> 00:06:52,460 of adaptation and environmental trade-offs. 156 00:06:52,460 --> 00:06:55,070 But then also from a personal reason 157 00:06:55,070 --> 00:06:56,180 or a personal perspective,

158 00:06:56,180 --> 00:06:58,190 my family is originally from India. 159 00:06:58,190 --> 00:07:00,440 I was born in Canada and grew up in the US 160 00:07:00,440 --> 00:07:02,990 but I always had this tie to India 161 00:07:02,990 --> 00:07:05,720 and it's just been amazing to be able to go back and, 162 00:07:05,720 --> 00:07:06,740 and work in a community 163 00:07:06,740 --> 00:07:08,780 where my parents and family have grown up. 164 00:07:08,780 --> 00:07:10,730 And you spent significant time there 165 00:07:10,730 --> 00:07:13,490 between your undergrad and grad school, is that right? 166 00:07:13,490 --> 00:07:15,380 Exactly. Yeah. So my relationship 167 00:07:15,380 --> 00:07:17,570 with India when I was a child, every five years 168 00:07:17,570 --> 00:07:19,250 or so, we would go back in the summers 169 00:07:19,250 --> 00:07:21,770 because most of my extended family is still there. 170 00:07:21,770 --> 00:07:24,710 But I really didn't have that long kind 171 00:07:24,710 --> 00:07:27,230 of time spent in the field until, as you mentioned, 172 00:07:27,230 --> 00:07:30,320 after college I was fortunate enough to get a fellowship 173 00:07:30,320 --> 00:07:33,470 that allowed me to do independent research for a year. 174 00:07:33,470 --> 00:07:35,390 And then since then I've been going back 175 00:07:35,390 --> 00:07:36,680 to India at least once 176 00:07:36,680 --> 00:07:39,320 or twice a year to spend time on the ground 177 00:07:39,320 --> 00:07:41,780 and maintain connections with my local collaborators. 178 00:07:41,780 --> 00:07:43,340 That's, that's fantastic. 179 00:07:44,750 --> 00:07:46,190 One of the findings or 180 00:07:46,190 --> 00:07:48,920 or the outcomes that you talked about in your essay 181 00:07:49,790 --> 00:07:52,010 was related to water 182 00:07:52,010 --> 00:07:54,545 and irrigation, obviously in this, the, the, 183 00:07:54,545 --> 00:07:56,510 the plight faced by these small farmers. 184 00:07:56,510 --> 00:07:59,840 But one of the findings was so interesting in that some 185 00:07:59,840 --> 00:08:03,050 of the areas at a small scale, there was actually room 186 00:08:03,050 --> 00:08:04,310 for them to irrigate more. 187 00:08:04,310 --> 00:08:06,170 Yeah. Can you talk more about that? 188 00:08:06,170 --> 00:08:08,660 Yeah, completely. So just 189 00:08:08,660 --> 00:08:12,020 to provide a little bit more background on this project, 190 00:08:12,020 --> 00:08:13,970 I was spending a lot of time in the field trying 191 00:08:13,970 --> 00:08:16,040 to understand how farmers were, were responding 192 00:08:16,040 --> 00:08:17,330 to warming temperatures. 193 00:08:17,330 --> 00:08:18,890 And I kept hearing that one 194 00:08:18,890 --> 00:08:20,870 of the main adaptation strategies was 195 00:08:20,870 --> 00:08:22,580 to increase irrigation. 196 00:08:22,580 --> 00:08:24,020 And while this is beneficial 197 00:08:24,020 --> 00:08:26,330 for reducing negative impacts on yield, 198 00:08:26,330 --> 00:08:27,830 we became a little bit concerned 199 00:08:27,830 --> 00:08:29,510 because as I mentioned, India has this 200 00:08:29,510 --> 00:08:31,370 groundwater depletion problem. 201 00:08:31,370 --> 00:08:32,600 So we did an analysis 202 00:08:32,600 --> 00:08:35,450 where we created a satellite data product measuring 203 00:08:35,450 --> 00:08:36,500 irrigation use, 204 00:08:36,500 --> 00:08:40,460 and found that nationwide farmers are increasing the amount 205 00:08:40,460 --> 00:08:42,140 of irrigation and this is leading 206 00:08:42,140 --> 00:08:43,940 to more groundwater depletion. 207 00:08:43,940 --> 00:08:46,550 But there are parts of the country where there's room 208 00:08:46,550 --> 00:08:48,860 for additional irrigation use, 209 00:08:48,860 --> 00:08:52,220 primarily in the north northeastern part of the country. 210 00:08:52,220 --> 00:08:55,260 So we've started working with our partner organizations 211 00:08:55,260 --> 00:08:58,620 to think about what are some additional strategies 212 00:08:58,620 --> 00:09:01,950 that can be used in these areas to help farmers adapt, given 213 00:09:01,950 --> 00:09:03,780 that they can rely a bit more on irrigation. 214 00:09:03,780 --> 00:09:04,980 Yeah. And what an amazing outcome. 215 00:09:04,980 --> 00:09:06,270 And all over the world, 216 00:09:06,270 --> 00:09:08,460 and even here locally, I'm gonna put a plug in now 217 00:09:08,460 --> 00:09:09,570 for our podcast 218 00:09:09,570 --> 00:09:12,420 that dropped last week on water in the Colorado River here 219 00:09:12,420 --> 00:09:14,880 in Arizona, but everybody's faced with sort of 220 00:09:14,880 --> 00:09:16,650 how will we use water more wisely? 221 00:09:16,650 --> 00:09:18,960 So amazing outcome, you know, 222 00:09:18,960 --> 00:09:22,290 for these small scale growers that are in the region. 223 00:09:22,290 --> 00:09:24,720 You mentioned already that you know, you're working 224 00:09:24,720 --> 00:09:26,820 with people in the region trying 225 00:09:26,820 --> 00:09:28,680 to help them use this information. 226 00:09:28,680 --> 00:09:30,180 Are there other examples you can share of 227 00:09:30,180 --> 00:09:32,100 how people are able to use your work sort 228 00:09:32,100 --> 00:09:34,140 of right now on the ground, 229 00:09:34,140 --> 00:09:36,630 kinda feeling the benefits that you're bringing? 230 00:09:36,630 --> 00:09:38,550 Yeah, so I'll describe one project 231 00:09:38,550 --> 00:09:40,380 that is currently at the beginning stages. 232 00:09:40,380 --> 00:09:43,200 So perfect. Yeah, not, not completely realized yet. 233 00:09:43,200 --> 00:09:45,810 I received a grant from a foundation called the 234 00:09:45,810 --> 00:09:48,150 S&R Evermay Foundation, 235 00:09:48,150 --> 00:09:52,800 and it was to create a satellite data product that could try 236 00:09:52,800 --> 00:09:55,950 to identify fields that were in stress 237 00:09:55,950 --> 00:09:57,900 and would eventually be low yielding at the 238

00:09:57,900 --> 00:09:59,040 end of the growing season. 239 00:09:59,040 --> 00:10:01,230 But to see whether we could identify those fields 240 00:10:01,230 --> 00:10:02,730 early in the season. 241 00:10:02,730 --> 00:10:05,430 And the hope is that we can then provide this information 242 00:10:05,430 --> 00:10:08,220 to agricultural extension agents who would then be able 243 00:10:08,220 --> 00:10:11,370 to target these fields in season to identify 244 00:10:11,370 --> 00:10:12,480 what the challenges are 245 00:10:12,480 --> 00:10:14,940 and hopefully help alleviate some of the problems. 246 00:10:14,940 --> 00:10:16,590 That's incredible. And I'm familiar 247 00:10:16,590 --> 00:10:20,160 with agriculture extension agents through the context of 248 00:10:20,160 --> 00:10:21,600 sort of US universities, 249 00:10:21,600 --> 00:10:25,410 land grant institutions have often extension offices, 250 00:10:25,410 --> 00:10:29,190 but those are presumably all over and able to be engaged 251 00:10:29,190 --> 00:10:32,340 and help people to put into practice some of these things 252 00:10:32,340 --> 00:10:33,660 that your team is solving. 253 00:10:33,660 --> 00:10:36,150 Exactly. Will you tell us a little bit more about your 254 00:10:36,150 --> 00:10:38,730 personal journey and sort of how you got here? 255 00:10:38,730 --> 00:10:41,760 What, what sort of undergraduate 256 00:10:41,760 --> 00:10:43,890 and graduate experiences did you have that sort 257 00:10:43,890 --> 00:10:45,150 of twists and turns you went through? 258 00:10:45,150 --> 00:10:47,130 How did you end up at this point in your career? 259 00:10:47,130 --> 00:10:48,510 So as far as I could remember, 260 00:10:48,510 --> 00:10:51,060 I've always been very interested in the environment. 261 00:10:51,060 --> 00:10:52,620 When I was a young kid, I think I wanted 262 00:10:52,620 --> 00:10:54,240 to be a marine biologist at first. 263 00:10:54,240 --> 00:10:55,905 Just knew that environment... Me too. 264 00:10:55,905 --> 00:10:58,050 Ah, the environment was just something 265 00:10:58,050 --> 00:10:59,940 that I wanted to work on. 266 00:10:59,940 --> 00:11:02,280 And in undergrad, I majored in a college 267 00:11:02,280 --> 00:11:04,260 and did a lot of amazing field work. 268 00:11:04,260 --> 00:11:06,030 I had really wonderful advisors. 269 00:11:06,030 --> 00:11:07,590 I got to go to Kenya and have an amazing 270 00:11:07,590 --> 00:11:09,030 field work experience. 271 00:11:09,030 --> 00:11:11,430 But as I started doing more work in this field, 272 00:11:11,430 --> 00:11:15,600 I realized humans are inextricably linked to the environment 273 00:11:15,600 --> 00:11:18,150 and I became much more passionate about 274 00:11:18,150 --> 00:11:19,440 how humans are interacting 275 00:11:19,440 --> 00:11:21,960 with the environment and our impacts. 276 00:11:21,960 --> 00:11:24,960 So I started to pursue that during my PhD. 277 00:11:24,960 --> 00:11:28,470 And during my PhD I saw the power

of remote sensing. 278 00:11:28,470 --> 00:11:31,560 So I, I never did any sort of satellite image analysis 279 00:11:31,560 --> 00:11:35,610 before going to grad school and starting my PhD. 280 00:11:35,610 --> 00:11:37,530 But after spending time in the field, 281 00:11:37,530 --> 00:11:39,630 even though household surveys are really wonderful, 282 00:11:39,630 --> 00:11:42,713 they give you this in-depth information about decision-making, 283 00:11:42,713 --> 00:11:43,954 I really wanted to understand 284 00:11:43,954 --> 00:11:45,720 what was really happening at scale. 285 00:11:45,720 --> 00:11:48,300 So I had had an advisor, a co-advisor 286 00:11:48,300 --> 00:11:51,120 who was a remote sensing expert and really learned a lot 287 00:11:51,120 --> 00:11:52,200 and was then able 288 00:11:52,200 --> 00:11:55,150 to marry those two tools together to do my work. 289 00:11:55,150 --> 00:11:57,160 That's a fantastic example of, you know, 290 00:11:57,160 --> 00:11:59,200 interdisciplinary research coming together 291 00:11:59,200 --> 00:12:01,960 to solve really pressing issues. 292 00:12:01,960 --> 00:12:04,480 And you mentioned last night when we were fortunate enough 293 00:12:04,480 --> 00:12:06,610 to fortunate enough to have dinner together. 294 00:12:06,610 --> 00:12:08,920 You described yourself as a geographer, 295 00:12:08,920 --> 00:12:11,530 but also have these other layers. 296 00:12:11,530 --> 00:12:12,790 Yeah, great question. Yeah. 297 00:12:12,790 --> 00:12:15,100 So people will ask me what my field is, 298 00:12:15,100 --> 00:12:16,570 and it's a little bit hard. 299 00:12:16,570 --> 00:12:19,660 I'd say that, yeah, I'm closest to being a geographer, 300 00:12:19,660 --> 00:12:21,160 but I think that's because geography is 301 00:12:21,160 --> 00:12:22,690 very interdisciplinary. 302 00:12:22,690 --> 00:12:25,330 So technically my degrees are in ecology 303 00:12:25,330 --> 00:12:27,100 and evolutionary biology, 304 00:12:27,100 --> 00:12:30,970 but especially in my PhD, I really broadened the types 305 00:12:30,970 --> 00:12:33,640 of classes that I took and the skill sets that I gained. 306 00:12:33,640 --> 00:12:36,280 So I actually did a certificate in environmental policy 307 00:12:36,280 --> 00:12:39,490 and through that I took classes in law, classes 308 00:12:39,490 --> 00:12:42,910 in human rights, classes in development economics, 309 00:12:42,910 --> 00:12:45,100 and all of those different classes, 310 00:12:45,100 --> 00:12:46,720 even though they seem very different and 311 00:12:46,720 --> 00:12:49,210 and distinct, really informed the way 312 00:12:49,210 --> 00:12:51,280 that I think about the problems that I look at 313 00:12:51,280 --> 00:12:53,380 and allowed me to take a much more holistic view. 314 00:12:53,380 --> 00:12:54,730 That's fantastic. I love it. 315 00:12:54,730 --> 00:12:57,010 And I hope that serves as some inspiration for folks 316 00:12:57,010 --> 00:12:59,140 who might be in grad school or undergrad right now. 317 00:12:59,140 --> 00:13:01,840 Trying to think about how do I create the career 318 00:13:01,840 --> 00:13:03,190 that I think I want to have, 319 00:13:03,190 --> 00:13:05,740 but also it's okay to explore. Yeah. 320 00:13:05,740 --> 00:13:08,920 Figure things out. Any advice you might have 321 00:13:08,920 --> 00:13:10,660 for early career researchers? 322 00:13:10,660 --> 00:13:14,140 I think for me, I just always followed 323 00:13:14,140 --> 00:13:15,790 what I was passionate about. 324 00:13:15,790 --> 00:13:17,710 Even if it meant that I had to pivot, 325 00:13:17,710 --> 00:13:19,210 maybe learn a new method 326 00:13:19,210 --> 00:13:23,260 or, you know, learn a new body of literature, I found that 327 00:13:23,260 --> 00:13:26,200 to be really energizing and it kept me really motivated 328 00:13:26,200 --> 00:13:29,260 and I feel like it allowed me to, you know, 329 00:13:29,260 --> 00:13:30,790 bring together disparate fields 330 00:13:30,790 --> 00:13:33,010 that maybe traditionally hadn't been brought together. 331 00:13:33,010 --> 00:13:35,590 So that's my advice, just to follow your passion, do 332 00:13:35,590 --> 00:13:37,210 what allows you to like, come 333 00:13:37,210 --> 00:13:39,615 to work every day and feel excited. I love it. 334 00:13:39,615 --> 00:13:41,200 One of the things we like to ask our guests is sort 335 00:13:41,200 --> 00:13:42,280 of a closing item. 336 00:13:42,280 --> 00:13:43,690 Is any recommended reading 337 00:13:43,690 --> 00:13:46,510 or any, any sources of information people might want 338 00:13:46,510 --> 00:13:48,940 to check out if they wanna learn more about 339 00:13:48,940 --> 00:13:50,380 your area of research. Yeah. 340 00:13:50,380 --> 00:13:52,630 Or related areas, things you think they should know. 341 00:13:52,630 --> 00:13:56,320 Sure. I think maybe what I'll talk about are some books 342 00:13:56,320 --> 00:13:58,660 that were really kind of formative for the way 343 00:13:58,660 --> 00:14:00,190 that I started thinking about my work. 344 00:14:00,190 --> 00:14:03,760 So it's not necessarily about small holder agriculture per 345 00:14:03,760 --> 00:14:07,360 per se, but more about development and rural communities. 346 00:14:07,360 --> 00:14:09,850 So one book that was really informative to the way 347 00:14:09,850 --> 00:14:13,000 that I think is a book called Environmentality written 348 00:14:13,000 --> 00:14:14,230 by Arun Agrawal, 349 00:14:14,230 --> 00:14:16,780 who subsequently has actually become one of my mentors. 350 00:14:16,780 --> 00:14:19,990 And it's been a really wonderful relationship that really 351 00:14:21,130 --> 00:14:23,110 solidified my kind of passion 352 00:14:23,110 --> 00:14:26,230 and understanding of small-scale producers 353 00:14:26,230 --> 00:14:29,740 and people, rural communities working in the global south 354 00:14:29,740 --> 00:14:31,780 and the challenges that they face. 355 00:14:31,780 --> 00:14:32,860 And then another

book 356 00:14:32,860 --> 00:14:36,580 that really inspired me was Amartya Sen's book, 357 00:14:36,580 --> 00:14:38,140 I forget the exact title right now, 358 00:14:38,140 --> 00:14:40,210 but discussing capability sets. 359 00:14:40,210 --> 00:14:45,210 And once again, really thinking about how do you define 360 00:14:45,490 --> 00:14:47,470 kind of possibilities in these systems 361 00:14:47,470 --> 00:14:49,690 and what are the constraints that people are facing. 362 00:14:49,690 --> 00:14:51,080 Thank you so much for joining us. 363 00:14:51,080 --> 00:14:53,510 I really appreciate you taking the time out and for, 364 00:14:53,510 --> 00:14:56,030 and for traveling such a great distance to join us 365 00:14:56,030 --> 00:14:59,390 for this great celebration of your essay and your award 366 00:14:59,390 --> 00:15:01,485 and also for joining us in the podcast. Thank 367 00:15:01,485 --> 00:15:02,318 You. Thank 368 00:15:02,318 --> 00:15:02,700 You so much. 369 00:15:05,060 --> 00:15:06,980 Welcome, Mayank. Thank you for joining us. 370 00:15:06,980 --> 00:15:09,110 Thank you for having me. I'm really excited. 371 00:15:09,110 --> 00:15:10,220 Great. I am too. 372 00:15:10,220 --> 00:15:11,930 I can't wait to talk about your work, 373 00:15:11,930 --> 00:15:15,380 which is using the power of technology 374 00:15:15,380 --> 00:15:19,160 to reduce human trafficking, which is obviously 375 00:15:19,160 --> 00:15:21,800 a horrible problem worldwide. 376 00:15:21,800 --> 00:15:23,390 I'm getting on a plane today. 377 00:15:23,390 --> 00:15:25,730 I know that when I go to the restroom in the airports, 378 00:15:25,730 --> 00:15:27,470 there's a sign on the back of the stall door. 379 00:15:27,470 --> 00:15:29,900 This is, are you here of your free will, 380 00:15:29,900 --> 00:15:31,400 basically, are you being coerced? 381 00:15:31,400 --> 00:15:34,340 This is, it's a huge problem. 382 00:15:34,340 --> 00:15:38,600 Can you tell us why this is such a challenging problem 383 00:15:38,600 --> 00:15:40,160 to address right now? 384 00:15:40,160 --> 00:15:43,220 Yes. So it turns out that 385 00:15:43,220 --> 00:15:46,400 the web has made human trafficking a lot worse. 386 00:15:46,400 --> 00:15:48,020 So it's always been there, 387 00:15:48,020 --> 00:15:52,520 but it used to be a very physical business, if you will, 388 00:15:52,520 --> 00:15:56,780 you know, taking place in back alleys and, and the likes. 389 00:15:56,780 --> 00:16:01,100 But since 1998, the statistics have been truly shocking. 390 00:16:01,100 --> 00:16:03,830 So between '98 to 2018, 391 00:16:03,830 --> 00:16:08,600 there has been a 6,000 x 6,000 times increase in 6,000, 392 00:16:08,600 --> 00:16:12,410 6,000 times increase in child pornography on the web. 393 00:16:12,410 --> 00:16:14,030 And that is just what we know. 394 00:16:14,030 --> 00:16:15,920 So you know that one statistic 395 00:16:15,920 --> 00:16:18,890 and really shows you the dark side of the web. 396 00:16:18,890 --> 00:16:21,590 So one of the questions that we try to address is, 397 00:16:21,590 --> 00:16:25,040 can we turn the technology, can we use the technology 398 00:16:25,040 --> 00:16:26,420 to turn the tables on the 399 00:16:26,420 --> 00:16:27,420 traffickers? 400 00:16:27,420 --> 00:16:30,020 Can you define human trafficking for us? 401 00:16:30,020 --> 00:16:31,040 So human trafficking is, 402 00:16:31,040 --> 00:16:33,800 is largely divided into two camps, labor trafficking 403 00:16:33,800 --> 00:16:35,330 and sex trafficking. 404 00:16:35,330 --> 00:16:37,850 I tend to study sex trafficking, 405 00:16:37,850 --> 00:16:40,430 which should be distinguished from sex work. 406 00:16:40,430 --> 00:16:43,010 Sex trafficking is a subset of sex work that tends 407 00:16:43,010 --> 00:16:44,540 to involve physical coercion, 408 00:16:45,860 --> 00:16:49,490 especially if it's not at the volition at all of the victim. 409 00:16:49,490 --> 00:16:50,720 And so the vast majority 410 00:16:50,720 --> 00:16:52,700 of human trafficking is still labor trafficking, 411 00:16:53,630 --> 00:16:55,580 but over 20% is sex trafficking 412 00:16:55,580 --> 00:16:58,970 and it tends to involve women and children by and large. 413 00:16:58,970 --> 00:17:02,300 You mentioned in your essay that the worldwide web, 414 00:17:02,300 --> 00:17:04,460 it's messy, the information is messy 415 00:17:04,460 --> 00:17:06,170 and it's hard to sort through. 416 00:17:06,170 --> 00:17:08,690 Can you tell us more about why this is such a challenge 417 00:17:08,690 --> 00:17:11,000 to investigate for investigators? 418 00:17:11,000 --> 00:17:14,060 So let's say if you search for my email address 419 00:17:14,060 --> 00:17:16,520 or my phone number, my work phone number, you know, 420 00:17:16,520 --> 00:17:19,820 you'll find it very easily because I want it to be found. 421 00:17:19,820 --> 00:17:21,590 It's on my webpage, it's on LinkedIn, 422 00:17:22,430 --> 00:17:26,150 but you know, you would not be able to find a, a victim 423 00:17:26,150 --> 00:17:28,040 or a sex worker by the phone number 424 00:17:28,040 --> 00:17:31,730 because they have obfuscated it using symbols, using emojis. 425 00:17:31,730 --> 00:17:34,790 So typically the way that they advertise is that when you go 426 00:17:34,790 --> 00:17:38,810 to a website that advertises this kind of service, 427 00:17:38,810 --> 00:17:40,640 they make it very easy to find someone 428 00:17:40,640 --> 00:17:43,040 by a physical characteristic 429 00:17:43,040 --> 00:17:45,770 because that is what the client might be looking for, 430 00:17:45,770 --> 00:17:47,900 but not by identifiers. 431 00:17:47,900 --> 00:17:51,630 Mm. And so law enforcement is, is is trying to find someone 432 00:17:51,630 --> 00:17:53,400 by using a phone number 433 00:17:53,400 --> 00:17:56,070 or, you know, maybe the last phone number they had 434 00:17:56,070 --> 00:18:00,330 before they went missing in some cases, or an email address 435

00:18:00,330 --> 00:18:03,930 or a social media handle or something like that. 436 00:18:03,930 --> 00:18:06,000 But they can't actually find anyone that way 437 00:18:06,000 --> 00:18:07,590 because it's all obfuscated. 438 00:18:07,590 --> 00:18:10,650 So it's really difficult to search on 439 00:18:10,650 --> 00:18:12,960 The web and that is one 440 00:18:12,960 --> 00:18:16,950 of the big reasons why we need technology to link billions 441 00:18:16,950 --> 00:18:19,020 of data points so we can map these networks 442 00:18:19,020 --> 00:18:20,940 of human trafficking and find victims. 443 00:18:20,940 --> 00:18:23,700 And the tool you built is called DIG. 444 00:18:23,700 --> 00:18:25,200 Tell us more about DIG. 445 00:18:25,200 --> 00:18:28,500 So DIG stands for Domain-specific Insight Graphs. 446 00:18:28,500 --> 00:18:30,240 So let me break that down a little bit. 447 00:18:30,240 --> 00:18:33,570 So a domain specific search engine is something like Amazon, 448 00:18:33,570 --> 00:18:35,310 which we're all familiar with. 449 00:18:35,310 --> 00:18:38,310 It's for e-commerce, you know, so in the case of Amazon 450 00:18:38,310 --> 00:18:39,810 or Yelp for reviews 451 00:18:39,810 --> 00:18:42,840 or YouTube for videos, those specific domain. 452 00:18:43,920 --> 00:18:46,410 So DIG is a domain specific search 453 00:18:46,410 --> 00:18:47,670 engine for human trafficking. 454 00:18:47,670 --> 00:18:50,370 So it, it's given to law enforcement agencies to 455 00:18:51,570 --> 00:18:53,850 address the challenge that I just spoke about earlier, 456 00:18:53,850 --> 00:18:58,710 which is how do we quickly find a victim using the last 457 00:18:58,710 --> 00:19:01,170 known phone number, using physical characteristics, 458 00:19:01,170 --> 00:19:03,780 using any other information that we might have, 459 00:19:03,780 --> 00:19:05,730 and then how do we find other people 460 00:19:05,730 --> 00:19:07,440 who might be linked to the victim? 461 00:19:07,440 --> 00:19:12,440 So this is the insight part of DIG, which is how do we 462 00:19:12,480 --> 00:19:14,130 very easily get law enforcement 463 00:19:14,130 --> 00:19:15,690 to have sophisticated insights 464 00:19:15,690 --> 00:19:18,840 that would take them years, if not months, to determine. 465 00:19:18,840 --> 00:19:20,520 So just to give you an example, 466 00:19:20,520 --> 00:19:22,140 let's say someone goes missing, 467 00:19:22,140 --> 00:19:23,730 we have very little information about them. 468 00:19:23,730 --> 00:19:27,030 Maybe we have a picture before they went missing, you know, 469 00:19:27,030 --> 00:19:28,260 but that's going to change 470 00:19:28,260 --> 00:19:31,740 because once they are trafficked, 471 00:19:31,740 --> 00:19:34,260 their physical appearance will change, the, 472 00:19:34,260 --> 00:19:35,520 their phone number will change. 473 00:19:35,520 --> 00:19:38,250 You know, ev they'll basically disappear, you know, off 474 00:19:38,250 --> 00:19:41,160 of the digital world and come back in a new guise. 475 00:19:41,160 --> 00:19:43,170 And so how do we link the two is a is a very 476 00:19:43,170 --> 00:19:44,430 hard AI problem. 477 00:19:45,480 --> 00:19:48,780 So what DIG does is it uses AI to gather a lot 478 00:19:48,780 --> 00:19:52,470 of data points from sex advertisements across the internet. 479 00:19:52,470 --> 00:19:53,760 It links them together 480 00:19:53,760 --> 00:19:56,010 and it puts it behind the interface 481 00:19:56,010 --> 00:19:58,980 that law enforcement can use to gain the kind of insights 482 00:19:58,980 --> 00:20:01,260 that would allow them to rescue victims. 483 00:20:01,260 --> 00:20:02,340 That's amazing. 484 00:20:02,340 --> 00:20:05,130 And I imagine really this tool wasn't, 485 00:20:05,130 --> 00:20:08,040 wouldn't be possible prior to this sort 486 00:20:08,040 --> 00:20:10,020 of widespread adoption of AI. 487 00:20:10,020 --> 00:20:11,730 Yeah, it actually preceded AI. 488 00:20:11,730 --> 00:20:15,690 So we were working on the tool, I joined USC in 2016, 489 00:20:15,690 --> 00:20:19,410 but the project was around from 2014 to 2018. 490 00:20:19,410 --> 00:20:21,870 So in the last two or three years of the project, 491 00:20:21,870 --> 00:20:23,100 which was funded by DARPA, 492 00:20:23,100 --> 00:20:25,440 the Defense Advanced Research Projects Agency, 493 00:20:25,440 --> 00:20:27,060 the AI was taking off, 494 00:20:27,060 --> 00:20:29,070 but we didn't have the large language models 495 00:20:29,070 --> 00:20:32,190 or any of the AI that we have today. 496 00:20:32,190 --> 00:20:34,140 So it was very challenging to build AI 497 00:20:34,140 --> 00:20:37,590 that had high enough quality that was scalable. 498 00:20:37,590 --> 00:20:38,700 It's never perfect. 499 00:20:38,700 --> 00:20:41,250 And so we were always talking to stakeholders 500 00:20:41,250 --> 00:20:43,620 that even if we're not perfect, 501 00:20:43,620 --> 00:20:48,220 how do we do a good enough job that we are helping you to 502 00:20:48,220 --> 00:20:51,430 truncate months and years of effort into days. 503 00:20:51,430 --> 00:20:53,620 We don't want to completely automate the process 504 00:20:53,620 --> 00:20:55,570 because we always want a human in the loop 505 00:20:55,570 --> 00:20:57,220 with something so sensitive. 506 00:20:57,220 --> 00:20:58,450 But our goal was, you know, 507 00:20:58,450 --> 00:21:00,850 how do we meet you where you are? 508 00:21:00,850 --> 00:21:05,350 And instead of taking a year, we turn it into a week. 509 00:21:05,350 --> 00:21:08,110 And so this way you can investigate many more cases. 510 00:21:09,310 --> 00:21:11,140 In the case of prosecutors 511 00:21:11,140 --> 00:21:13,690 and district attorney's offices, we wanted to make sure 512 00:21:13,690 --> 00:21:16,840 that the tool could be used to find evidence 513 00:21:16,840 --> 00:21:19,090 that then allows them to put traffickers behind bars. 514 00:21:19,090 --> 00:21:20,800 And there were a few success

stories, 515 00:21:20,800 --> 00:21:22,990 but we were able to gather evidence through the tool 516 00:21:22,990 --> 00:21:25,720 that actually did allow prosecutors to prove 517 00:21:25,720 --> 00:21:27,010 beyond a reasonable doubt 518 00:21:27,010 --> 00:21:28,960 that the victims were getting trafficked. 519 00:21:28,960 --> 00:21:30,760 What an amazing outcome. 520 00:21:30,760 --> 00:21:33,490 You mentioned, you know, this is sensitive data enforcement 521 00:21:33,490 --> 00:21:35,740 officials are using this for legitimate purposes, 522 00:21:35,740 --> 00:21:38,320 but is there, is there any risk to DIG, 523 00:21:38,320 --> 00:21:40,630 this is a powerful surveillance tool, 524 00:21:40,630 --> 00:21:44,260 are there privacy concerns for individuals out there? 525 00:21:44,260 --> 00:21:47,140 How, how do we mitigate those two competing worries 526 00:21:47,140 --> 00:21:48,700 that people have sometimes? 527 00:21:48,700 --> 00:21:50,530 Yeah, it's a great question. 528 00:21:50,530 --> 00:21:54,820 We explicitly designed the tool so that it only goes 529 00:21:54,820 --> 00:21:57,070 after the domain-specific websites 530 00:21:57,070 --> 00:21:58,240 that have interest over here. 531 00:21:58,240 --> 00:22:01,810 And so these were sex work websites where we do know a lot 532 00:22:01,810 --> 00:22:03,130 of trafficking and especially child 533 00:22:03,130 --> 00:22:05,140 trafficking does happen a lot. 534 00:22:05,140 --> 00:22:08,500 There is a privacy concern with respect to sex workers 535 00:22:08,500 --> 00:22:10,930 because they are victims in this 536 00:22:10,930 --> 00:22:14,110 and we don't want them to be victimized even further. 537 00:22:15,310 --> 00:22:17,530 You know, there's no way the technology can prevent that. 538 00:22:17,530 --> 00:22:19,180 So it was really important for us to work 539 00:22:19,180 --> 00:22:21,250 with our partners in law enforcement to ensure 540 00:22:21,250 --> 00:22:23,290 that it wasn't getting used by that. 541 00:22:23,290 --> 00:22:25,750 One of the pieces of evidence that gave me a lot of hope 542 00:22:25,750 --> 00:22:28,480 and that it was being used for the right purpose is 543 00:22:28,480 --> 00:22:32,350 that the Manhattan Human Trafficking Resource Unit gave us 544 00:22:32,350 --> 00:22:35,830 some data very early on in the project where it turned out 545 00:22:35,830 --> 00:22:38,950 that the number of prostitution arrests 546 00:22:38,950 --> 00:22:41,200 that they were making in Manhattan went down 547 00:22:41,200 --> 00:22:42,850 very significantly. 548 00:22:42,850 --> 00:22:45,850 But the fraction of arrests that they were investigating 549 00:22:45,850 --> 00:22:48,460 for human trafficking went from less than 1% 550 00:22:48,460 --> 00:22:51,910 to more than 62% after the tool was rolled out. 551 00:22:51,910 --> 00:22:54,460 The reason why this is so significant is 552 00:22:54,460 --> 00:22:57,640 because even if you are an undocumented immigrant 553 00:22:57,640 --> 00:23:00,580 or you don't have any paperwork, if under the law, 554 00:23:00,580 --> 00:23:03,460 if there is reason to believe that you are being trafficked, 555 00:23:03,460 --> 00:23:07,270 then there are certain protections, refugee protections, 556 00:23:07,270 --> 00:23:11,230 and other protections that kick in where you're safe. 557 00:23:11,230 --> 00:23:13,660 And you also get a lot of resources, you know, 558 00:23:13,660 --> 00:23:15,190 to rehabilitate you 559 00:23:15,190 --> 00:23:18,430 and to give you the resources that are necessary for you 560 00:23:18,430 --> 00:23:20,380 to recover from the trauma. 561 00:23:20,380 --> 00:23:24,580 So this is really an an amazing example of, of the power 562 00:23:24,580 --> 00:23:28,960 of AI coming in, really facilitating this sort of rapid kind 563 00:23:28,960 --> 00:23:31,120 of acceleration of this tool 564 00:23:31,120 --> 00:23:34,300 and AI being used for good in this, in this context. 565 00:23:34,300 --> 00:23:36,550 Can you tell us more about your personal journey 566 00:23:36,550 --> 00:23:38,410 and sort of how you got here to this point 567 00:23:38,410 --> 00:23:42,010 where you were doing this really inspiring research? 568 00:23:42,010 --> 00:23:47,010 So I, I grew up in India, you know, I went to 569 00:23:47,120 --> 00:23:48,800 a very academically rigorous school 570 00:23:48,800 --> 00:23:51,440 that was very deep into math and science. 571 00:23:51,440 --> 00:23:53,930 So I was always very interested in computer science and math 572 00:23:53,930 --> 00:23:56,480 and physics, but I thought I wanted to be a physicist. 573 00:23:57,380 --> 00:24:00,710 So when I came to University of Illinois at the age of 19, 574 00:24:01,670 --> 00:24:04,400 I double majored in in physics and electrical 575 00:24:04,400 --> 00:24:05,540 and computer engineering. 576 00:24:06,620 --> 00:24:10,670 And I had the, the, the good fortune where I was introduced 577 00:24:10,670 --> 00:24:12,050 to someone in the National Center 578 00:24:12,050 --> 00:24:14,840 for Supercomputing Applications and UIUC 579 00:24:14,840 --> 00:24:17,330 and I got my first research experience during the summer 580 00:24:17,330 --> 00:24:20,210 and I just fell in love with it, just reading papers 581 00:24:20,210 --> 00:24:21,380 and going into depth. 582 00:24:21,380 --> 00:24:25,460 And it was just something that, that just clicked. 583 00:24:26,480 --> 00:24:28,850 And I decided that I could not give that up 584 00:24:28,850 --> 00:24:30,500 and I should go and pursue a PhD. 585 00:24:30,500 --> 00:24:34,340 And so went onto University of Texas at Austin. 586 00:24:34,340 --> 00:24:36,260 My research area was fairly boring, you know, 587 00:24:36,260 --> 00:24:37,670 in a way I'm still involved in that. 588 00:24:37,670 --> 00:24:39,200 But it was an AI area 589 00:24:39,200 --> 00:24:41,450 and it, it had nothing to do with crime 590 00:24:41,450 --> 00:24:43,910 or human trafficking or anything

like that. 591 00:24:43,910 --> 00:24:46,760 And I wasn't planning to go into academia after that. 592 00:24:46,760 --> 00:24:48,860 I had an industry job lined up 593 00:24:48,860 --> 00:24:53,480 and I was invited by some researchers who I knew very well. 594 00:24:53,480 --> 00:24:55,880 I mean, I'd read all their papers and so I looked up to them 595 00:24:55,880 --> 00:24:59,780 and I ran into them in conferences and they were at USA 596 00:24:59,780 --> 00:25:01,280 and they were like, well, you know, 597 00:25:01,280 --> 00:25:02,240 there's this really interesting 598 00:25:02,240 --> 00:25:03,620 project on human trafficking. 599 00:25:03,620 --> 00:25:06,170 It fits very perfectly with your research. 600 00:25:06,170 --> 00:25:09,260 We need to apply the AI that you're building now 601 00:25:09,260 --> 00:25:10,670 to this problem. 602 00:25:10,670 --> 00:25:14,300 And so, you know, if you are open at all, you know, 603 00:25:14,300 --> 00:25:16,160 just come down, give a talk 604 00:25:16,160 --> 00:25:17,900 and you know, it's the worst 605 00:25:17,900 --> 00:25:20,780 that will happen is you will have come down to Los Angeles 606 00:25:20,780 --> 00:25:22,790 for a couple of days, that's not too bad. 607 00:25:22,790 --> 00:25:24,890 And then, you know, if you, if you decide it's not for you, 608 00:25:24,890 --> 00:25:26,330 then you can go on to 609 00:25:26,330 --> 00:25:28,220 what you were thinking of doing before. 610 00:25:28,220 --> 00:25:29,960 So I went to the institute 611 00:25:29,960 --> 00:25:32,000 and I, I just fell in love with it, 612 00:25:33,230 --> 00:25:35,210 the Information Sciences Institute at USC, 613 00:25:35,210 --> 00:25:36,950 which is my home institution. 614 00:25:36,950 --> 00:25:40,190 Any advice you'd offer for early career researchers? 615 00:25:40,190 --> 00:25:41,960 So this is going to sound a little bit strange, 616 00:25:41,960 --> 00:25:46,640 but I would, my advice would be trust serendipity, 617 00:25:47,750 --> 00:25:49,220 don't overplan. 618 00:25:49,220 --> 00:25:53,810 I think sometimes I see, you know, especially the, you know, 619 00:25:53,810 --> 00:25:56,420 young people, I think for good reasons, you know, they, 620 00:25:56,420 --> 00:25:58,250 they see that the world is very competitive 621 00:25:58,250 --> 00:25:59,870 and they have this notion that, okay, I need 622 00:25:59,870 --> 00:26:01,850 to check all these boxes. 623 00:26:01,850 --> 00:26:04,190 I need to publish here, or I need to do this, 624 00:26:04,190 --> 00:26:06,380 or I need to have accomplished this by my second year 625 00:26:06,380 --> 00:26:08,150 and by my fourth year 626 00:26:08,150 --> 00:26:10,610 and I need to have these line items on my resume. 627 00:26:10,610 --> 00:26:12,620 And I'm not saying that's not important, you know, 628 00:26:12,620 --> 00:26:14,420 of course you should have some kind of plan 629 00:26:14,420 --> 00:26:15,620 and, and all of that. 630 00:26:15,620 --> 00:26:17,090 But

I would say don't overplan. 631 00:26:17,090 --> 00:26:20,060 Like think of this as 70% is planning 632 00:26:20,060 --> 00:26:22,610 and 30% you have to trust the world. 633 00:26:22,610 --> 00:26:24,800 That's my one piece of advice, which is say yes 634 00:26:24,800 --> 00:26:28,700 to many things, you know, take an exploratory view, 635 00:26:28,700 --> 00:26:30,770 trust serendipity, don't overplan. 636 00:26:32,390 --> 00:26:37,010 And you know, you will be surprised by the wonderful things 637 00:26:37,010 --> 00:26:38,600 that are waiting for you out there. 638 00:26:38,600 --> 00:26:41,030 I like the word serendipity. It feels perfect. 639 00:26:42,020 --> 00:26:45,600 One of the last things we ask folks is, you know, 640 00:26:45,600 --> 00:26:47,520 if somebody hears the podcast 641 00:26:47,520 --> 00:26:49,530 and wants to learn more about this 642 00:26:49,530 --> 00:26:52,200 or related areas, are there books 643 00:26:52,200 --> 00:26:54,570 or other places you would steer them to, 644 00:26:54,570 --> 00:26:56,490 to check out if they wanna keep learning 645 00:26:56,490 --> 00:26:57,690 or books that inspired you 646 00:26:57,690 --> 00:26:59,670 that got you started in this process? 647 00:26:59,670 --> 00:27:02,550 If anyone is interested in very specific books, you know, 648 00:27:02,550 --> 00:27:05,040 there was a great book written about Backpage, 649 00:27:05,040 --> 00:27:08,100 which was a huge website that had a lot of sex work ads 650 00:27:08,100 --> 00:27:09,510 and it got taken down. 651 00:27:09,510 --> 00:27:13,770 But what I, to me, I really love reading old books. 652 00:27:13,770 --> 00:27:16,140 So I was just thinking about Max Weber, 653 00:27:17,460 --> 00:27:19,710 who was a great German sociologist, 654 00:27:19,710 --> 00:27:23,820 and he had, he did these two lectures. 655 00:27:23,820 --> 00:27:26,010 One of them, they called the vocation lecture. 656 00:27:26,010 --> 00:27:28,500 I believe he gave them at a university 657 00:27:28,500 --> 00:27:29,760 or some kind of audience. 658 00:27:29,760 --> 00:27:31,530 It was not intended as a book, 659 00:27:31,530 --> 00:27:35,340 but one of the lectures was called Science as a Vocation. 660 00:27:35,340 --> 00:27:38,010 Usually they publish as two lectures, "Science as a Vocation," 661 00:27:38,010 --> 00:27:39,900 and "Politics as a Vocation." 662 00:27:39,900 --> 00:27:41,580 I don't recommend reading the second one 663 00:27:41,580 --> 00:27:43,410 unless you're very interested in politics, 664 00:27:43,410 --> 00:27:45,540 but I think for us, scientists, "Science 665 00:27:45,540 --> 00:27:47,460 as a Vocation" was an incredible lecture. 666 00:27:47,460 --> 00:27:50,220 I mean, I read it and it still resonates today 667 00:27:50,220 --> 00:27:54,150 and it really reminds me why we choose to become scientists. 668 00:27:54,150 --> 00:27:56,640 A much more recent book called, 669 00:27:56,640 --> 00:28:00,445 I believe it's called "The Information," or, or... Oh, sure. 670 00:28:00,445 --> 00:28:02,010 I I am also familiar that yeah, 671 00:28:02,010 --> 00:28:03,450 "The Information" of like, Yep. 672 00:28:03,450 --> 00:28:04,590 I think it's by James Gleick. That's right. 673 00:28:04,590 --> 00:28:08,010 I believe is a science, a historian of science at Harvard. 674 00:28:08,010 --> 00:28:11,370 I think that, that it really gives you a very deep 400-year 675 00:28:11,370 --> 00:28:15,150 perspective on a concept that seems 676 00:28:15,150 --> 00:28:19,710 so simple on the surface that involves so much intricacy. 677 00:28:19,710 --> 00:28:22,770 And to me, that's the beauty of scientific discovery. 678 00:28:22,770 --> 00:28:25,140 I also agree that's a terrific book. 679 00:28:25,140 --> 00:28:26,490 Thank you so much for joining us. 680 00:28:26,490 --> 00:28:28,530 This has been a lot of fun. 681 00:28:28,530 --> 00:28:29,500 Oh, thank you. 682 00:28:33,100 --> 00:28:35,760 Thank you to Meha and Mayank for joining us today. 683 00:28:35,760 --> 00:28:39,420 And thank you to Lara for hosting. Thank you for listening. 684 00:28:39,420 --> 00:28:41,460 If you enjoyed the show, please leave us a five-star review 685 00:28:41,460 --> 00:28:43,590 on the podcast app of your choice. 686 00:28:43,590 --> 00:28:45,600 If you wanna learn more about AAAS 687 00:28:45,600 --> 00:28:46,890 and support their work, 688 00:28:46,890 --> 00:28:49,890 consider joining the AAAS member community. 689 00:28:49,890 --> 00:28:51,990 More information in the show notes. 690 00:28:51,990 --> 00:28:55,530 Hey, remember, lab coats are optional on this show, 691 00:28:55,530 --> 00:28:57,690 but we wanna keep our researchers safe in the lab. 692 00:28:57,690 --> 00:28:59,423 So wear your PPE. 693 00:28:59,423 --> 00:29:01,050 Lab Coat Optional is a production 694 00:29:01,050 --> 00:29:03,060 of the talented people at the Knowledge Enterprise, 695 00:29:03,060 --> 00:29:05,520 Strategic Marketing and Communications team. 696 00:29:05,520 --> 00:29:06,780 Knowledge Enterprise is led 697 00:29:06,780 --> 00:29:09,510 by Executive Vice President Sally C. Morton, 698 00:29:09,510 --> 00:29:10,770 who's right over here. 699 00:29:11,940 --> 00:29:12,940 Hi Sally. 700 00:29:13,920 --> 00:29:16,980 Until next time, stay curious.