1 00:00:02,324 --> 00:00:05,590 And I think you'll have to get rid of that. 2 00:00:05,590 --> 00:00:06,684 Just say continue there. 3 00:00:06,684 --> 00:00:09,310 Continue, right? 4 00:00:09,310 --> 00:00:12,740 >> That's because if someone knows something's being recorded, 5 00:00:12,740 --> 00:00:14,565 they may not wanna stay in the room. 6 00:00:14,565 --> 00:00:16,795 Sometimes privacy thing. 7 00:00:16,795 --> 00:00:20,081 >> So that's what I've noticed is to everybody's screen, 8 00:00:20,081 --> 00:00:23,730 when I click on record which actually is probably the right thing to do. 9 00:00:23,730 --> 00:00:25,040 >> Yeah. 10 00:00:25,040 --> 00:00:26,360 >> It's very interesting. 11 00:00:26,360 --> 00:00:27,530 >> Yeah. 12 00:00:27,530 --> 00:00:31,356 It's common when you call some plays on the phone, they say, 13 00:00:31,356 --> 00:00:32,910 this isn't a recorded. 14 00:00:32,910 --> 00:00:33,990 >> It's kind of the same thing. 15 00:00:33,990 --> 00:00:34,840 >> Yeah. 16 00:00:34,840 --> 00:00:35,733 >> All right. So anyway, 17 00:00:35,733 --> 00:00:38,630 you all know who Rajesh is pretty much but I'll just go ahead and 18 00:00:38,630 --> 00:00:41,330 say I'm really pleased that he agreed to give a talk and 19 00:00:41,330 --> 00:00:44,430 he's presenting something a little bit different than what we normally have, 20 00:00:44,430 --> 00:00:49,580 and I think it's a great idea to sort of invite us to consider a conference menu 21 00:00:49,580 --> 00:00:52,010 that comes particularly provocative and thoughtful. 22 00:00:52,010 --> 00:00:53,000 And I think that's great. 23 00:00:53,000 --> 00:00:55,850 I'm looking forward very much to hearing it, so Rajesh, take it away. 24 00:00:57,030 --> 00:01:03,590 >> Okay, thanks for coming and, do I know that? 25 00:01:06,150 --> 00:01:06,820 Yeah, okay. 26 00:01:06,820 --> 00:01:08,930 >> I'm not sure you might have to go back and forth. 27 00:01:08,930 --> 00:01:10,780 If you have PDFs that you're trying to show, 28 00:01:10,780 --> 00:01:14,080 okay, maybe that's something that I don't know. 29 00:01:15,460 --> 00:01:17,890 Okay, you can close it right now just so it doesn't interfere with it. 30 00:01:17,890 --> 00:01:18,440 >> Got it. 31 00:01:18,440 --> 00:01:19,640 Yeah. 32 00:01:19,640 --> 00:01:21,120 Okay. 33 00:01:21,120 --> 00:01:23,600 So, Doc, far. 34 00:01:23,600 --> 00:01:24,460 >> One more thing. 35 00:01:24,460 --> 00:01:25,130 Sorry. 36 00:01:25,130 --> 00:01:28,880 Maybe we can make that stayed on the last presentation the whole time. 37 00:01:28,880 --> 00:01:31,993 And then I think you can probably just hit the minus sign there and it'll go away. 38 00:01:33,884 --> 00:01:35,690 >> Maybe just keep it. 39 00:01:35,690 --> 00:01:36,550 >> Give away up there. 40 00:01:36,550 --> 00:01:38,515 >> Okay, thank you. 41 00:01:38,515 --> 00:01:39,860 [LAUGH] >> Yeah, so 42 00:01:39,860 --> 00:01:42,320 I also wanted to do something different. 43 00:01:42,320 --> 00:01:46,460 Usually, we can tap into or review or critique, that's my favorite thing to do. 44 00:01:46,460 --> 00:01:51,479 But this time, I wanted to make it more constructive. 45 00:01:51,479 --> 00:01:55,380 And especially for this talk, 46 00:01:55,380 --> 00:01:58,530 I would say because I'm going to express my views first. 47 00:01:58,530 --> 00:02:03,593 So, I want to make sure that, because it's the views 48 00:02:03,593 --> 00:02:09,780 that I have about things and not what we all is considered common or 49 00:02:09,780 --> 00:02:14,525 that so I wanna make sure that it's a disclaimer. 50 00:02:14,525 --> 00:02:20,229 Okay so, the two conferences 51 00:02:20,229 --> 00:02:25,478 that I went to upon Marty's 52 00:02:25,478 --> 00:02:30,499 insistence is right there 53 00:02:30,499 --> 00:02:35,072 in the pictures [LAUGH]. 54 00:02:35,072 --> 00:02:38,789 So, I went there last year and I went this year. 55 00:02:38,789 --> 00:02:44,119 Last year I just posted my dissertation chapters there and 56 00:02:44,119 --> 00:02:49,015 people found it interesting enough that they invited 57 00:02:49,015 --> 00:02:51,842 me to collaborate with them and 58 00:02:51,842 --> 00:02:57,410 this time I present the one that I did with my collaborators. 59 00:02:57,410 --> 00:02:58,940 I founded that company. 60 00:02:58,940 --> 00:03:01,120 So this is a conference I went to. 61 00:03:02,240 --> 00:03:06,600 And they also have it's hosted by kind of this 62 00:03:08,560 --> 00:03:12,720 computational sorts of Science Society of Americas group. 63 00:03:12,720 --> 00:03:20,210 And they also host this Complexity and Policy Studies conference in DC. 64 00:03:20,210 --> 00:03:27,170 So the way I see it, one of them is 65 00:03:27,170 --> 00:03:33,022 more kind of the science, the other one the CAPS, conferences, 66 00:03:33,022 --> 00:03:38,238 more implications of complex systems for policy. 67 00:03:38,238 --> 00:03:42,337 So they both, 68 00:03:44,812 --> 00:03:50,454 Okay the deadline for one of them is June and the other one is in February and 69 00:03:50,454 --> 00:03:54,280 the conference takes place in October and in May. 70 00:03:56,210 --> 00:04:02,110 Okay, so the reason I'm talking about this 71 00:04:02,110 --> 00:04:07,530 is that I went to the conference and then I found out that most of the work that 72 00:04:07,530 --> 00:04:13,470 resistance is to are more pertinent for the conference than my work was. 73 00:04:13,470 --> 00:04:19,106 So that's what kind of got me thinking that maybe we should have a presence 74 00:04:19,106 --> 00:04:27,510 in the conferences and people that I didn't expect to hear, 75 00:04:27,510 --> 00:04:32,320 like pastels and no one appeared in the presentation slides. 76 00:04:34,020 --> 00:04:39,530 And also the whole sequence of courses on simulation, 77 00:04:39,530 --> 00:04:45,340 the topics were all about agent based modeling discrete event simulation, 78 00:04:45,340 --> 00:04:46,620 and system dynamics. 79 00:04:46,620 --> 00:04:51,160 So it was kind of interesting that there is this conference and 80 00:04:51,160 --> 00:04:54,210 he do not have a significant present. 81 00:04:54,210 --> 00:04:59,770 So that's one of my motivations for presenting and kind 82 00:04:59,770 --> 00:05:06,858 of discussing whether we could make more of our presence in the future conferences. 83 00:05:06,858 --> 00:05:13,200 Okay, so the topical areas in the contrast that have 84 00:05:13,200 --> 00:05:17,973 seen in the last couple of years like as I said, 85 00:05:17,973 --> 00:05:22,987 system dynamics, discrete-event simulation, 86 00:05:22,987 --> 00:05:28,500 agent-based modeling, and when I talked to the organizers of the conference, 87 00:05:28,500 --> 00:05:30,210 they said when they started the conference, 88 00:05:30,210 --> 00:05:35,710 they wanted to make it explicitly agent based when they first started it. 89 00:05:35,710 --> 00:05:40,870 And then they could not stop themselves 90 00:05:40,870 --> 00:05:45,950 from also accommodating other approaches because all this is done with multiple 91 00:05:47,110 --> 00:05:52,490 modeling as a comparative, so, therefore they said they 92 00:05:52,490 --> 00:05:59,310 have to kind of expand the vision of the conference into other domains. 93 00:05:59,310 --> 00:06:04,250 And awfully the last few years supposedly a lot of people are working on 94 00:06:05,640 --> 00:06:10,030 network models and they like to present their work on network models. 95 00:06:10,030 --> 00:06:17,590 So, they saw a lot of network models, presentations and also, 96 00:06:18,740 --> 00:06:24,980 yeah, there's certain other things that I was not aware of. 97 00:06:24,980 --> 00:06:27,440 One of them was doing large scale. 98 00:06:29,180 --> 00:06:32,040 My micro simulation where 99 00:06:32,040 --> 00:06:35,670 they kind of pretty much do simulation of the entire population. 100 00:06:35,670 --> 00:06:41,190 That's something that I only loosely around but I saw it in practice 101 00:06:41,190 --> 00:06:45,140 in several of the talks, they were thinking about technology adoption and 102 00:06:45,140 --> 00:06:49,460 they take the model for counties and hold states using agent based modeling. 103 00:06:49,460 --> 00:06:56,710 And the same thing with kind of modeling of renewable energy markets. 104 00:06:57,900 --> 00:07:02,580 They use kind of competition economics and agent based modeling to do that. 105 00:07:02,580 --> 00:07:09,450 And I also saw a lot of Ted a few talks on the, again 106 00:07:09,450 --> 00:07:14,390 we made use of machine learning and data mining techniques to analyze social data. 107 00:07:14,390 --> 00:07:19,020 And then there was also a few talks on 108 00:07:19,020 --> 00:07:22,932 kind of generally data mining, health data and other things. 109 00:07:22,932 --> 00:07:28,720 And some theoretical papers, 110 00:07:28,720 --> 00:07:35,510 like for example, my work was kind of tagged as philosophical and 111 00:07:35,510 --> 00:07:41,195 theoretical by most of them and this year I saw some papers on 112 00:07:41,195 --> 00:07:45,665 evolutionary game theory and social biology and things like that. 113 00:07:45,665 --> 00:07:48,765 I think whether you present in your work on catastrophe theory, right? 114 00:07:50,480 --> 00:07:55,712 So, [COUGH] it's a good mix of criteria. 115 00:07:55,712 --> 00:07:56,430 >> [INAUDIBLE] >> Yeah. 116 00:07:56,430 --> 00:08:00,560 >> I'm curious about these large scale simulations. 117 00:08:00,560 --> 00:08:06,354 I want to start off with somebody simulating a forest And 118 00:08:06,354 --> 00:08:11,310 modeling every tree in the forest, it struck me as absurd. 119 00:08:11,310 --> 00:08:15,490 So what do you think, 120 00:08:15,490 --> 00:08:20,600 when they have an agent for 121 00:08:20,600 --> 00:08:23,397 each person? 122 00:08:23,397 --> 00:08:25,089 What was your reaction to those papers? 123 00:08:25,089 --> 00:08:30,538 Did they convince you that it was valuable to model at that level of detail or 124 00:08:30,538 --> 00:08:32,537 did they not convince you? 125 00:08:32,537 --> 00:08:35,824 >> Basically, they had this mixture of, 126 00:08:35,824 --> 00:08:41,015 when I started out with skeptic, as I start with most things. 127 00:08:41,015 --> 00:08:46,089 So me being skeptic is not anything abnormal, 128 00:08:46,089 --> 00:08:51,942 but I did kind of realize that they're kind of working 129 00:08:51,942 --> 00:08:58,334 on things where they think working with a lot of unknowns. 130 00:08:58,334 --> 00:09:03,697 They have no idea how the real system even behave. 131 00:09:03,697 --> 00:09:08,399 So there's this kind of a knowable unknowns, so 132 00:09:08,399 --> 00:09:13,215 they wanted to kind of calibrate their models with 133 00:09:13,215 --> 00:09:17,353 some real world census level simulation. 134 00:09:17,353 --> 00:09:22,161 So their argument was kind of we have no 135 00:09:22,161 --> 00:09:26,970 idea how to understand the system, so 136 00:09:26,970 --> 00:09:32,099 we have to go all in with all approaches. 137 00:09:32,099 --> 00:09:35,401 >> Why do it at the ultimate micro level? 138 00:09:35,401 --> 00:09:42,424 Why not do it in some aggregated system dynamics general variables? 139 00:09:42,424 --> 00:09:47,861 >> So it's one of those things where it's kind of becoming very current 140 00:09:47,861 --> 00:09:52,935 in terms of acknowledging the role of space in social sciences. 141 00:09:52,935 --> 00:09:57,366 And the precise nature of geographical census tracks and 142 00:09:57,366 --> 00:10:02,071 geography has a definite role to play in kind of power grid and 143 00:10:02,071 --> 00:10:05,316 accessibility to renewable resources. 144 00:10:05,316 --> 00:10:11,256 So there's something unique about modeling the spatial context, 145 00:10:11,256 --> 00:10:16,164 and I agree that they could work on some simpler models. 146 00:10:16,164 --> 00:10:20,107 My hope is that they would eventually do it, but for 147 00:10:20,107 --> 00:10:25,345 now they want to kind of study it as a micro granularity that they do. 148 00:10:35,962 --> 00:10:40,891 Okay, I'll go with last year to this year. 149 00:10:40,891 --> 00:10:45,561 Last year- >> Most of the people are out there so- 150 00:10:45,561 --> 00:10:48,006 >> [CROSSTALK] 151 00:10:48,006 --> 00:10:50,982 >> So rather than talking to the screen, 152 00:10:50,982 --> 00:10:54,525 talk to people that are virtually out there. 153 00:10:54,525 --> 00:11:01,454 >> Last year I had Leigh Tesfatsion, she's [INAUDIBLE] I think. 154 00:11:01,454 --> 00:11:06,682 She talked about compilation economics,and she said she 155 00:11:06,682 --> 00:11:11,596 was an economist trying to find her place in the world. 156 00:11:11,596 --> 00:11:14,728 And then she went to [INAUDIBLE] conference from 95. 157 00:11:14,728 --> 00:11:20,151 As she kind of reoriented her mind completely, 158 00:11:20,151 --> 00:11:24,897 and she a joint appointment in electrical 159 00:11:24,897 --> 00:11:28,973 engineering as well as economics. 160 00:11:28,973 --> 00:11:34,868 So she's also kind of [INAUDIBLE] computational economics. 161 00:11:34,868 --> 00:11:40,045 So she was the person who was talking about microstimulation of markets. 162 00:11:40,045 --> 00:11:45,188 These days, even in Portland, they kind of have sensors now, 163 00:11:45,188 --> 00:11:51,110 so you don't actually collect any kind of power meter records anymore, 164 00:11:51,110 --> 00:11:52,969 it's all automated. 165 00:11:52,969 --> 00:11:58,934 And she's kind of trying to study how market responds to power grid changes, 166 00:11:58,934 --> 00:12:03,524 technology changes as they went from power to renewable and 167 00:12:03,524 --> 00:12:06,298 kind of making it into a smart grid. 168 00:12:06,298 --> 00:12:08,107 So that was our TED Talk. 169 00:12:08,107 --> 00:12:12,036 And with Mirta Galesic, she's a resident faculty at Santa Fe, 170 00:12:12,036 --> 00:12:15,763 and she talked about collective decision making analysis. 171 00:12:15,763 --> 00:12:21,340 But I found this more interesting and more pertinent for 172 00:12:21,340 --> 00:12:25,270 what we in system science are interested in. 173 00:12:25,270 --> 00:12:29,070 However, her work is more kinda academic and 174 00:12:29,070 --> 00:12:34,775 more technical in studying models of collective decision making. 175 00:12:34,775 --> 00:12:38,826 And the CS stock, one of them was from 176 00:12:38,826 --> 00:12:44,203 George Washington University, Matt Johnson. 177 00:12:44,203 --> 00:12:49,105 And he talked about building models of online 178 00:12:49,105 --> 00:12:53,749 extremist activity and other things, and 179 00:12:53,749 --> 00:13:01,109 I was not an applicant of jumping ship from physics a long time back. 180 00:13:01,109 --> 00:13:07,158 I was not convinced of the models that he was building, but it was interesting. 181 00:13:07,158 --> 00:13:13,505 He used a lot of statistical mechanic models for talking about online activity, 182 00:13:13,505 --> 00:13:17,533 and that was something that, it was interesting. 183 00:13:17,533 --> 00:13:22,296 But this connection with how it actually works in 184 00:13:22,296 --> 00:13:26,245 the real world without any sociology and 185 00:13:26,245 --> 00:13:31,822 other kind of contact with social science disciplines, 186 00:13:31,822 --> 00:13:35,091 I was kind of not very convinced. 187 00:13:35,091 --> 00:13:38,545 And then I made sure that my opinion was right, so 188 00:13:38,545 --> 00:13:43,182 I talked to other people who are working in terrorist activity and 189 00:13:43,182 --> 00:13:48,258 they also said that it's kind of very restricted to physics journals. 190 00:13:48,258 --> 00:13:53,061 The work is mostly about terrorist activity, but apply return for 191 00:13:53,061 --> 00:13:54,871 physicists in this job. 192 00:13:54,871 --> 00:14:00,313 So there's no engagement with actual researchers of terrorism and 193 00:14:00,313 --> 00:14:01,941 online extremism. 194 00:14:01,941 --> 00:14:05,589 The other one was interesting, 195 00:14:05,589 --> 00:14:12,085 she is now director of data science program in Colombia. 196 00:14:12,085 --> 00:14:17,903 But she's also kind of full time for science popularizer and 197 00:14:17,903 --> 00:14:21,538 she does stand-up comedy- >> [LAUGH] 198 00:14:21,538 --> 00:14:25,220 >> And a whole bunch of things. 199 00:14:25,220 --> 00:14:31,561 And she's very much into making science more cool. 200 00:14:31,561 --> 00:14:36,377 And actually, there was a [INAUDIBLE] start 201 00:14:36,377 --> 00:14:40,403 page where [INAUDIBLE] in Michigan. 202 00:14:40,403 --> 00:14:45,614 So she has some interesting things to say 203 00:14:45,614 --> 00:14:51,897 about what I can help to popularize in science and 204 00:14:51,897 --> 00:14:56,812 how to make science more attractive. 205 00:14:56,812 --> 00:15:01,431 I did not agree with most of what she said, but 206 00:15:01,431 --> 00:15:07,158 it was kind of attracted to certain demographic group. 207 00:15:07,158 --> 00:15:12,258 It was not like the notion of popularizing science and 208 00:15:12,258 --> 00:15:16,351 making it interesting and entertaining. 209 00:15:16,351 --> 00:15:20,784 That as a view can kind of trying to attract 210 00:15:20,784 --> 00:15:25,219 students who are not normally affected, 211 00:15:25,219 --> 00:15:30,675 it's appealing to certain socioeconomic class. 212 00:15:30,675 --> 00:15:33,536 But for the people who are kind of poor, 213 00:15:33,536 --> 00:15:38,655 the people who are powerless and feel like based on my own experience, 214 00:15:38,655 --> 00:15:42,668 I feel like further knowledge is always more radical. 215 00:15:42,668 --> 00:15:48,525 It's like they want emancipation and political power, 216 00:15:48,525 --> 00:15:52,551 like a part of them, no amount of X, Y, 217 00:15:52,551 --> 00:15:59,762 Z is cool is going to help them say I want to do complex systems. 218 00:15:59,762 --> 00:16:04,535 The target audience for that kind of intervention in education is Very limited. 219 00:16:04,535 --> 00:16:10,163 [INAUDIBLE] But yeah, and the most interesting talk that 220 00:16:10,163 --> 00:16:15,804 I found was this person from Brandeis who talked about. 221 00:16:18,692 --> 00:16:19,192 Never. 222 00:16:27,928 --> 00:16:31,489 Yeah,I can totally picture target. 223 00:16:31,489 --> 00:16:36,213 So, this was a in a kind of very theoretical 224 00:16:36,213 --> 00:16:40,810 paper that appeared at the conference and 225 00:16:40,810 --> 00:16:45,418 it was presented by Paul Davis from Rand. 226 00:16:45,418 --> 00:16:51,162 And his argument was that we have to kind of reorient 227 00:16:51,162 --> 00:16:56,373 the entirety of how policy analysis is done and 228 00:16:56,373 --> 00:17:00,248 taught around systems science and 229 00:17:00,248 --> 00:17:06,544 complex systems that you say you wanted to kind of this. 230 00:17:06,544 --> 00:17:11,142 Yeah, they position paper was about how 231 00:17:11,142 --> 00:17:16,282 limitations of existing policy analysis and 232 00:17:16,282 --> 00:17:22,110 how system types and complex systems studies can. 233 00:17:22,110 --> 00:17:27,213 So I found it very interesting, but 234 00:17:27,213 --> 00:17:32,315 I felt that it was too optimistic and 235 00:17:32,315 --> 00:17:36,580 he was kind of in management. 236 00:17:36,580 --> 00:17:37,651 He kind of knew, 237 00:17:37,651 --> 00:17:42,545 kind of felt like he was trying to fulfill the unfinished promise of original 238 00:17:42,545 --> 00:17:47,745 administrative behavior industrial dynamics program of like Herbert Simon and 239 00:17:47,745 --> 00:17:53,033 then start making mention everybody as they say like the program is not finished. 240 00:17:53,033 --> 00:17:57,696 I feel like you're starting to see through but 241 00:17:57,696 --> 00:18:02,719 my attitude over the years has kinda of become more 242 00:18:02,719 --> 00:18:07,274 about kind of academics and theoreticians. 243 00:18:07,274 --> 00:18:12,998 I can share my idea with this kind of policy researcher and 244 00:18:12,998 --> 00:18:19,810 social scientist, Charles Lindbergh, he passed away last year. 245 00:18:19,810 --> 00:18:26,681 So he has this kind of interesting paper in the 1960s, 246 00:18:26,681 --> 00:18:30,757 called muddling through policy. 247 00:18:30,757 --> 00:18:35,210 And that was about policy as like incremental policy should be 248 00:18:35,210 --> 00:18:37,016 viewed incrementally. 249 00:18:37,016 --> 00:18:42,247 And then 20 years later, he wrote another rejoinder called 250 00:18:42,247 --> 00:18:47,990 still muddling through, and I think a few months back either this. 251 00:18:47,990 --> 00:18:52,479 What will estate book it's called usable knowledge and 252 00:18:52,479 --> 00:18:55,447 then he had a very interesting take. 253 00:18:55,447 --> 00:18:59,057 Basically he claimed that social scientists and 254 00:18:59,057 --> 00:19:03,196 researchers should not be involved in policymaking and 255 00:19:03,196 --> 00:19:08,666 it's only the net, the stakeholders be responsible for policymaking. 256 00:19:08,666 --> 00:19:14,909 And this is just to just enable stakeholders no matter what their position 257 00:19:14,909 --> 00:19:22,071 to make sure that through the confluence of integration of stakeholder ideas that. 258 00:19:22,071 --> 00:19:23,328 >> Were they lobbyists as well? 259 00:19:23,328 --> 00:19:28,175 [LAUGH] >> His argument is basically when we 260 00:19:28,175 --> 00:19:32,668 should not keep sources, what he calls professional social science 261 00:19:32,668 --> 00:19:37,561 inquirers like people in academia and like they say here's how to do this. 262 00:19:37,561 --> 00:19:40,166 Like that's not useful. 263 00:19:40,166 --> 00:19:44,621 And his argument is that the academics were currently 264 00:19:44,621 --> 00:19:49,472 professional social science inquiries have to be working 265 00:19:49,472 --> 00:19:54,134 with stakeholders to push their agenda no matter what. 266 00:19:54,134 --> 00:19:59,441 And let the market place of like who's kind of who's seen this 267 00:19:59,441 --> 00:20:04,444 kind of interaction between different stakeholders that 268 00:20:04,444 --> 00:20:10,072 are you might add a realistic solution to the real world problem. 269 00:20:10,072 --> 00:20:10,909 So that's. 270 00:20:10,909 --> 00:20:16,771 >> I think, stakeholders are not interested in the truth. 271 00:20:16,771 --> 00:20:23,225 And you need some voice that's deliberately committed to truth and not. 272 00:20:25,153 --> 00:20:27,508 The true is different from the good and 273 00:20:27,508 --> 00:20:32,669 the stakeholders are interested in what's good for them, which is already good. 274 00:20:32,669 --> 00:20:36,818 But somebody's got to be interested in what's true regardless of whether it's 275 00:20:36,818 --> 00:20:38,595 good for this group or that group. 276 00:20:38,595 --> 00:20:41,441 >> We could get a lot of debate on that position. 277 00:20:41,441 --> 00:20:46,720 [LAUGH] How many people, I mean, we sit in a world that's interested in the truth, 278 00:20:46,720 --> 00:20:50,254 but this is not the majority world we're in right now. 279 00:20:50,254 --> 00:20:55,977 >> Right, but I can see why you should argue that you're interested 280 00:20:55,977 --> 00:21:02,026 in what's true, or to recuse themselves from policy discussions. 281 00:21:02,026 --> 00:21:04,782 No, they should be even more involved. 282 00:21:04,782 --> 00:21:06,565 Their voice should be even louder. 283 00:21:06,565 --> 00:21:11,593 >> So yeah, [COUGH] I 284 00:21:11,593 --> 00:21:16,628 feel fail over. 285 00:21:16,628 --> 00:21:22,014 He basically say like the people who are can obsess about the when 286 00:21:22,014 --> 00:21:27,102 I'm paraphrasing, people who are obsessed about the truth 287 00:21:27,102 --> 00:21:33,286 have no role in kind of designing except to kind of associating themselves 288 00:21:33,286 --> 00:21:38,500 with a stake stakeholder that says, like there are gonna be. 289 00:21:38,500 --> 00:21:39,622 >> I think it's unreasonable. 290 00:21:39,622 --> 00:21:44,419 I mean, you want people who make policy to be not ignorant about what 291 00:21:44,419 --> 00:21:49,732 they're making about the actual area, we want them to be knowledgeable. 292 00:21:49,732 --> 00:21:53,423 So we want people who have knowledge to be present. 293 00:21:53,423 --> 00:21:56,672 They shouldn't necessarily make the decisions but 294 00:21:56,672 --> 00:21:59,336 they should be offering their expertise. 295 00:21:59,336 --> 00:22:06,689 >> Yeah, so he's all for offering via >> Yeah, 296 00:22:06,689 --> 00:22:07,657 offering it to the interest group. 297 00:22:07,657 --> 00:22:08,521 >> Interest groups. 298 00:22:08,521 --> 00:22:14,060 >> But the interest groups, by definition will, 299 00:22:14,060 --> 00:22:20,428 let's say, will use those facts that support them and 300 00:22:20,428 --> 00:22:26,943 will ignore and hide the fact that don't support them. 301 00:22:26,943 --> 00:22:28,085 [LAUGH] >> It is like a. 302 00:22:28,085 --> 00:22:31,053 >> I'm arguing with him, not with you. 303 00:22:31,053 --> 00:22:33,282 >> No, no, no, yeah, yeah. 304 00:22:33,282 --> 00:22:37,362 I tend to side with him a little bit, but it's fine. 305 00:22:37,362 --> 00:22:41,764 But it's kind of an interesting question of, 306 00:22:44,417 --> 00:22:47,331 Yeah, yeah, it's of interesting. 307 00:22:47,331 --> 00:22:54,063 But in general Lindblom that was kind of faster to incrementalist. 308 00:22:54,063 --> 00:22:58,023 And many folks think that like something, 309 00:22:58,023 --> 00:23:05,393 certain kinds of changes have to come like drastic, probably going to change, 310 00:23:05,393 --> 00:23:10,243 make incremental change with every kind of social. 311 00:23:10,243 --> 00:23:15,255 Funny he is the general consensus of like Al Qaeda trying to frame the kind 312 00:23:15,255 --> 00:23:20,202 of the general opinion that I've heard from interacting with people. 313 00:23:20,202 --> 00:23:26,262 There the content seems to be absorb tools from computational social science, 314 00:23:26,262 --> 00:23:30,605 systematic kind of more and more into policy analysis and 315 00:23:30,605 --> 00:23:32,783 make it part of operations. 316 00:23:32,783 --> 00:23:38,250 Such as interested in dating, and also like in policy analysis more gender. 317 00:23:38,250 --> 00:23:46,097 That seems to be like, the direction where like everybody's doing so 318 00:23:46,097 --> 00:23:52,481 that's like my kinda that's gonna whatever techniques 319 00:23:52,481 --> 00:23:58,075 we have like it should be kinda become part of the. 320 00:23:58,075 --> 00:24:00,025 >> Some stand up comedian stuff was out a little bit. 321 00:24:00,025 --> 00:24:06,640 [LAUGH] >> So, yeah, but not everything is. 322 00:24:08,020 --> 00:24:13,230 Not everything is kind of mean I agree with like, 323 00:24:13,230 --> 00:24:18,340 many of them I feel like too much off they 324 00:24:18,340 --> 00:24:22,830 were doing even like system dynamics modeling but with NetFlow. 325 00:24:22,830 --> 00:24:26,870 But they were kinda talking that we are doing agent based modeling. 326 00:24:26,870 --> 00:24:31,650 So I felt like this like that's where I feel 327 00:24:31,650 --> 00:24:36,498 like a more rigorous presentation of what agent based modeling is. 328 00:24:36,498 --> 00:24:40,520 And what it's not and kind of clearly delineating from system dynamics and 329 00:24:40,520 --> 00:24:46,110 this state even says, modeling all those things is kind of required but 330 00:24:46,110 --> 00:24:51,590 there was like one kind of criticism that I had and, 331 00:24:51,590 --> 00:24:56,590 and also, some of others also said, and I also felt like there's like, 332 00:24:56,590 --> 00:25:01,150 too much of meaning and very little of the modeling. 333 00:25:01,150 --> 00:25:05,970 So they come up with a simple predator prey model and 334 00:25:05,970 --> 00:25:10,980 the lot cover Kara and then they kind of add layers of meaning to it and 335 00:25:10,980 --> 00:25:15,230 then it becomes implications for some other topic. 336 00:25:15,230 --> 00:25:20,620 I felt like it's like too much of a storyteller that was like one. 337 00:25:22,530 --> 00:25:26,540 And based on my experience and my quest is they can actively 338 00:25:27,940 --> 00:25:31,720 discourage kind of working with us or 339 00:25:31,720 --> 00:25:36,150 social scientists like they want to do things on their own. 340 00:25:36,150 --> 00:25:38,480 And that's like my criticism. 341 00:25:39,700 --> 00:25:42,572 >> Well social scientists are going to bring in all kinds of fancy 342 00:25:42,572 --> 00:25:44,142 stuff that'll be hard to model. 343 00:25:44,142 --> 00:25:44,910 >> Yeah. >> [LAUGH] 344 00:25:44,910 --> 00:25:45,580 >> That's 345 00:25:45,580 --> 00:25:47,680 like the thing that needs to be there. 346 00:25:47,680 --> 00:25:49,920 I talked with them after the dog Matt Johnson. 347 00:25:49,920 --> 00:25:56,545 He said that you do your work for 10 years or society says we'll never be happy. 348 00:25:56,545 --> 00:26:00,895 But he said like if social scientists say like, you have something important 349 00:26:00,895 --> 00:26:03,905 physics, I have something important Physics Department. 350 00:26:03,905 --> 00:26:06,705 I will take it, but I know he's kind of. 351 00:26:06,705 --> 00:26:09,895 That's not the way it goes. 352 00:26:09,895 --> 00:26:13,200 But in their defense, I also say that. 353 00:26:13,200 --> 00:26:17,280 That's like, just based on my various by the way, 354 00:26:17,280 --> 00:26:22,630 like I'm using caste system as a metaphor don't please don't take it. 355 00:26:22,630 --> 00:26:23,631 >> Metaphor for what? 356 00:26:23,631 --> 00:26:28,406 >> Metaphor for higher education in the academic setting like that's like 357 00:26:28,406 --> 00:26:32,450 that is the real computational social scientists. 358 00:26:32,450 --> 00:26:36,250 And that is [INAUDIBLE] the second [INAUDIBLE] 359 00:26:36,250 --> 00:26:37,795 >> A pecking order. 360 00:26:37,795 --> 00:26:43,011 >> There's the pecking order [INAUDIBLE] difference level, 361 00:26:43,011 --> 00:26:48,546 the kind of stuff that they do, and if I were to bring [INAUDIBLE] 362 00:26:48,546 --> 00:26:53,337 argument about interdisciplinary they say like, 363 00:26:53,337 --> 00:26:57,914 that's not the way we do it doesn't certainly that 364 00:26:57,914 --> 00:27:03,180 thing is there and it's like you go to that conference. 365 00:27:03,180 --> 00:27:06,550 It's kind of very visible but 366 00:27:06,550 --> 00:27:11,470 what I really liked about the conference, is that unlike other conference where 367 00:27:11,470 --> 00:27:15,920 I been recently where after the break everybody's like 368 00:27:15,920 --> 00:27:20,730 a marketplace where everybody's kind of buying and selling post docs and students. 369 00:27:20,730 --> 00:27:23,240 And like everybody, 370 00:27:23,240 --> 00:27:28,910 all the participants are like trying to get professional enhancement. 371 00:27:28,910 --> 00:27:30,850 Here I saw and 372 00:27:30,850 --> 00:27:36,000 I also participate a lot of open range discuss, is very casual single track. 373 00:27:36,000 --> 00:27:38,450 And I like the fact about this conference. 374 00:27:38,450 --> 00:27:42,980 The competition was it was more of a genuine in terms of 375 00:27:42,980 --> 00:27:47,830 like they had they brought their unpolished ideas and they're looking for 376 00:27:47,830 --> 00:27:52,230 collaborators to collaborate with a lot of collaborations have emerged over the years 377 00:27:52,230 --> 00:27:56,550 that I can sell them and I also become part of it. 378 00:27:56,550 --> 00:28:00,040 And one thing to notice is that I've talked to 379 00:28:00,040 --> 00:28:02,599 some older generation of companies and social scientists. 380 00:28:03,740 --> 00:28:07,070 And they ever had bitter experiences being rejected in 381 00:28:08,310 --> 00:28:12,340 either kind of the kind of the pure pure journals. 382 00:28:12,340 --> 00:28:17,690 Like I met one person here I was talking about Robert Merton, the socialists. 383 00:28:17,690 --> 00:28:24,710 In the workshop that we organized the collaborators, and I listened 384 00:28:24,710 --> 00:28:30,410 to Robert Merton speak in the 1950s when I was a grad student so don't take. 385 00:28:32,216 --> 00:28:37,510 But he turns out, he kind of published some important work 386 00:28:37,510 --> 00:28:42,640 in econophysics in the 1980s which is highly cited and 387 00:28:42,640 --> 00:28:46,340 he told me that you just do your own thing. 388 00:28:46,340 --> 00:28:49,510 Don't listen to sociologists and don't listen to any, 389 00:28:49,510 --> 00:28:53,775 just do your own thing in computational sources. 390 00:28:53,775 --> 00:28:56,130 That's enough. 391 00:28:56,130 --> 00:28:59,010 He sounded very bitter because he 392 00:29:00,160 --> 00:29:04,280 interacted with people back in the day and he had some difficulty. 393 00:29:04,280 --> 00:29:09,530 And another person who's also the organizer of the conference? 394 00:29:09,530 --> 00:29:13,480 He was talking about, he was trying to, he presented some work on using 395 00:29:14,930 --> 00:29:20,350 nonlinear dynamic time series analysis to study health outcomes. 396 00:29:20,350 --> 00:29:25,400 And he is one of the kind of informal sessions that 397 00:29:25,400 --> 00:29:30,000 he had this paper on trying to do non linear time series analysis of 398 00:29:30,000 --> 00:29:34,530 macro economic data in Canada in various world economics. 399 00:29:34,530 --> 00:29:37,130 And for like 10 years there was no 400 00:29:37,130 --> 00:29:40,470 economic journals that was willing to publish that work. 401 00:29:40,470 --> 00:29:45,740 So like it sounded like although I can have this criticism that might be 402 00:29:45,740 --> 00:29:50,476 in response to that experience in the past. 403 00:29:50,476 --> 00:29:53,350 And also another day. 404 00:29:53,350 --> 00:29:58,960 That's like a very fixed way to kind of go through your kind of 405 00:29:58,960 --> 00:30:02,350 graduate program and how to get tenure. 406 00:30:02,350 --> 00:30:06,810 Like, for example, many of the students there, they're systematically over 407 00:30:06,810 --> 00:30:12,080 there kind of grad student life published paper every year for 408 00:30:12,080 --> 00:30:17,390 like three years and then that's considered kind of 409 00:30:17,390 --> 00:30:21,410 good enough accomplishment to take them to the next stage. 410 00:30:21,410 --> 00:30:28,500 And same thing with the none of the presenters of their work is always like, 411 00:30:28,500 --> 00:30:33,260 should I present at this conference is this conference recognized for 412 00:30:33,260 --> 00:30:36,260 my tenure committee? 413 00:30:36,260 --> 00:30:41,150 And for my professional society and so on, like a lot of them 414 00:30:42,160 --> 00:30:45,680 listen to kind of discussions and panel sessions about that. 415 00:30:45,680 --> 00:30:49,590 So that that seems to be like something to do with that to like, 416 00:30:49,590 --> 00:30:55,220 what is kind of the right paper to submit, which should be recognized by your peers. 417 00:30:55,220 --> 00:30:56,568 That's all like obviously. 418 00:30:56,568 --> 00:31:01,170 So my experience as we like I, as I say, 419 00:31:01,170 --> 00:31:08,290 like I've had some fun collaborators, and these are the people that I met. 420 00:31:08,290 --> 00:31:15,357 So yeah, so he's from St. Louis is like our director of Analytics program, 421 00:31:15,357 --> 00:31:21,350 [INAUDIBLE] and then he's in a university in Virginia. 422 00:31:21,350 --> 00:31:23,880 Steven Davis is a computer science person. 423 00:31:23,880 --> 00:31:28,080 And he's from Canada, working on urban city designer, 424 00:31:28,080 --> 00:31:33,190 and this year, I met Santiago who's 425 00:31:33,190 --> 00:31:37,370 working in the national supercomputing facility in Urbana Champaign. 426 00:31:37,370 --> 00:31:42,130 And as we speak, again upon planning to 427 00:31:42,130 --> 00:31:47,550 submit a paper to a life with some kind of I had some interesting, 428 00:31:47,550 --> 00:31:53,370 positive collaborators and we can meet every week to discuss those later. 429 00:31:53,370 --> 00:31:55,860 He's been like really good for me. 430 00:31:56,870 --> 00:32:01,500 And, like I said, a potential lot of potential funding opportunities. 431 00:32:01,500 --> 00:32:07,490 I was told by Milton that if you go to the Katz conference and present your work 432 00:32:07,490 --> 00:32:12,822 that's actually Funding agencies coming in to kind of listen to the talks. 433 00:32:12,822 --> 00:32:16,571 So that's where he says that we should present the work but 434 00:32:16,571 --> 00:32:21,174 because I'm pretty sure that my work has got nothing to do with policy. 435 00:32:21,174 --> 00:32:25,312 I don't think I can- >> What is there? 436 00:32:25,312 --> 00:32:29,235 >> It's the conference- >> Applies to designs? 437 00:32:29,235 --> 00:32:33,909 >> Yeah, complexity and policy, 438 00:32:33,909 --> 00:32:36,878 is this one.- >> So I- 439 00:32:36,878 --> 00:32:37,851 >> The other one that you- 440 00:32:37,851 --> 00:32:39,774 >> No that was a sister conference is- 441 00:32:39,774 --> 00:32:40,385 >> Okay thank you. 442 00:32:40,385 --> 00:32:44,804 >> Okay, yeah, >> And they kind of buy, 443 00:32:44,804 --> 00:32:50,321 after you submit the paper here, they will do a second round of review and 444 00:32:50,321 --> 00:32:55,206 get it published in the complexity, policy analysis journal, 445 00:32:55,206 --> 00:32:58,205 that's like what kind of advantages? 446 00:32:58,205 --> 00:32:59,354 Anyway like that. 447 00:32:59,354 --> 00:33:00,029 >> Okay thanks. 448 00:33:00,029 --> 00:33:01,664 >> Yeah. 449 00:33:01,664 --> 00:33:07,031 >> So, but personally, I felt very bad because whenever 450 00:33:07,031 --> 00:33:12,396 I had any question they cheer me a little bit [LAUGH] when I 451 00:33:12,396 --> 00:33:17,542 think mostly joking [LAUGH] but they are they scared. 452 00:33:17,542 --> 00:33:21,764 That's like my kind of small goodwill I had this like that they 453 00:33:21,764 --> 00:33:26,482 pointed out that paper should not contain differential equations. 454 00:33:26,482 --> 00:33:31,708 >> So you can write a complex algorithm no problem but one equation forget it. 455 00:33:31,708 --> 00:33:36,381 >> Yeah that's kind of a bit- >> [LAUGH] 456 00:33:36,381 --> 00:33:38,922 >> But example a different of 457 00:33:38,922 --> 00:33:44,005 the equation, but okay, we don't want citation 458 00:33:44,005 --> 00:33:49,910 from other disciplines like we just want to do our modeling. 459 00:33:49,910 --> 00:33:54,680 That is also when I point out like they say you don't and 460 00:33:54,680 --> 00:34:00,800 then also where people are, like I've been doing this long enough and 461 00:34:00,800 --> 00:34:05,774 I typically get a lot of this, but I got a little bit of that 462 00:34:05,774 --> 00:34:09,841 in my kind of back and forth with people there. 463 00:34:09,841 --> 00:34:16,923 [LAUGH] But anyway, and yeah my experience is mostly good. 464 00:34:16,923 --> 00:34:22,160 And I felt like there was like a whole generation of people 465 00:34:22,160 --> 00:34:27,286 to kind of have too much kind of associated with Nitro was 466 00:34:27,286 --> 00:34:32,977 that like they could not break off of their kind of candidate. 467 00:34:32,977 --> 00:34:37,859 They could not model conceptually like, how kind of the sermon talks about or 468 00:34:37,859 --> 00:34:40,789 like where you can build a conceptual model and 469 00:34:40,789 --> 00:34:44,019 then decide what's like the appropriate model but 470 00:34:44,019 --> 00:34:48,252 they always start off from, like the NetLogo was on to some model. 471 00:34:48,252 --> 00:34:51,132 So I feel like I kind of had to do it and 472 00:34:51,132 --> 00:34:57,180 my collaborators were like really kind of willing to kind of listen to me and 473 00:34:57,180 --> 00:35:04,097 they allowed me to present my stochastic differential equation models in the paper. 474 00:35:04,097 --> 00:35:09,057 And for most of them, they feel that the papers published in physical review and 475 00:35:09,057 --> 00:35:13,943 physical physics, that's like the gold standard of how to do it and I felt like 476 00:35:13,943 --> 00:35:18,848 there's like a lot of really good social science that other people have done. 477 00:35:18,848 --> 00:35:24,062 And I was able to kind of slowly initiate them into the various reading and 478 00:35:24,062 --> 00:35:27,481 I kind of do some book review every time I meet and 479 00:35:27,481 --> 00:35:31,263 I'm kind of happy that I was able to do my little bit. 480 00:35:31,263 --> 00:35:38,745 As I said, like I've been a lot of work that I've found with that and 481 00:35:38,745 --> 00:35:45,458 I can see kind of me being tech support for a lot of their work. 482 00:35:45,458 --> 00:35:49,219 So I feel happy with that road, and 483 00:35:49,219 --> 00:35:54,408 I feel like when first thing I was exposed to this 484 00:35:54,408 --> 00:35:59,209 was the system sign says architect metaphor 485 00:35:59,209 --> 00:36:04,269 that I have assignment but also quite often but 486 00:36:04,269 --> 00:36:08,681 I don't think that's like that stuff, 487 00:36:08,681 --> 00:36:14,158 place that everybody appreciates into this way. 488 00:36:14,158 --> 00:36:20,081 Folks, and this is like ten years back there's this researcher, 489 00:36:20,081 --> 00:36:23,994 Ivan Roth, he's an economist in Harvard, 490 00:36:23,994 --> 00:36:28,858 he won the Nobel Prize and one of his famous lectures and 491 00:36:28,858 --> 00:36:33,547 essays is on this topic of economist as an engineer. 492 00:36:33,547 --> 00:36:40,178 And recently, Duflo, she and along with others she won the Nobel Prize this year, 493 00:36:40,178 --> 00:36:45,972 and she says that actually our economy should not just be kind of a thing. 494 00:36:45,972 --> 00:36:49,419 In terms of engineers and designers, 495 00:36:49,419 --> 00:36:54,058 we should think in terms of kind of being a plumber. 496 00:36:54,058 --> 00:36:58,559 And like worry about like nitty gritty small 497 00:36:58,559 --> 00:37:03,183 mechanisms that enable the system to work and 498 00:37:03,183 --> 00:37:09,267 I feel like this is like the approach that I've taken like, 499 00:37:09,267 --> 00:37:14,617 handy based on like, we can be in a part of a team where 500 00:37:14,617 --> 00:37:19,615 we're like little bit of everything- >> I like handyman a lot better, 501 00:37:19,615 --> 00:37:21,380 a handyman is a generalist of one versus a specialist. 502 00:37:21,380 --> 00:37:22,464 >> Yeah. 503 00:37:22,464 --> 00:37:23,368 >> Much more often. 504 00:37:23,368 --> 00:37:27,003 >> Yeah so absolutely, and personally, 505 00:37:27,003 --> 00:37:31,630 my own personal take like this looking forward for 506 00:37:31,630 --> 00:37:36,808 that kind of whatever I'm gonna do, I am kind of feeling 507 00:37:36,808 --> 00:37:42,773 like this role where like, that's the place where I want to be. 508 00:37:42,773 --> 00:37:47,887 I'm not saying that I really like this role in general, but 509 00:37:47,887 --> 00:37:53,939 this is just my personal take that I feel that I can be kind of elevated. 510 00:37:53,939 --> 00:37:59,051 On my good days and bad days either I got vacillate between thinking 511 00:37:59,051 --> 00:38:04,072 about myself as a curator or collecting here myself as a custodian 512 00:38:04,072 --> 00:38:08,001 because I have wondered like a- >> A clean up person. 513 00:38:08,001 --> 00:38:13,257 >> Yeah therefore a lot of the models that I help build or 514 00:38:13,257 --> 00:38:17,438 I can help do the cleanup in terms of like, 515 00:38:17,438 --> 00:38:21,739 this is not a that's not a service next, so 516 00:38:21,739 --> 00:38:26,894 I guess I'm kind of calling myself the custodian. 517 00:38:26,894 --> 00:38:30,242 But on some good days, 518 00:38:30,242 --> 00:38:35,008 I feel like I can curate things. 519 00:38:35,008 --> 00:38:36,963 >> What I hear though is that, 520 00:38:36,963 --> 00:38:41,638 bringing more rigor is part of it, that that rigor is worrisome, 521 00:38:41,638 --> 00:38:46,996 we're not always present in the complex system work that you're seeing. 522 00:38:46,996 --> 00:38:50,630 And so that's that's an important thing that we- 523 00:38:50,630 --> 00:38:52,803 >> What do you mean by cure? 524 00:38:52,803 --> 00:38:56,987 >> Curator the museum like organize knowledge 525 00:38:56,987 --> 00:39:00,610 it's separate wheat from the chaff. 526 00:39:00,610 --> 00:39:05,804 It's important though for like a something 527 00:39:05,804 --> 00:39:12,410 can be in between like one step librarian or something. 528 00:39:12,410 --> 00:39:15,882 >> So you don't have among the different roles, 529 00:39:15,882 --> 00:39:20,343 this is scientists as an emergency doctor emergency room. 530 00:39:20,343 --> 00:39:24,191 >> Yeah that must make your reader gate septic- 531 00:39:24,191 --> 00:39:27,562 >> Or a system science as a metaphysical 532 00:39:27,562 --> 00:39:28,438 doctor. 533 00:39:28,438 --> 00:39:31,664 [LAUGH] That's my take. 534 00:39:31,664 --> 00:39:34,753 >> So yeah that's- >> Nick and 535 00:39:34,753 --> 00:39:39,816 one of the people put a chat it says something like what what do you think of 536 00:39:39,816 --> 00:39:44,809 system scientists work towards developing philosophies of science? 537 00:39:44,809 --> 00:39:47,964 >> But yeah, >> Just the comments want to share that it 538 00:39:47,964 --> 00:39:49,114 doesn't have a voice here. 539 00:39:49,114 --> 00:39:50,899 >> [LAUGH] >> Yeah. 540 00:39:50,899 --> 00:39:56,371 >> Yeah, that's a good way to think about it is also interesting 541 00:39:56,371 --> 00:40:01,106 based on my kind of dissertation writing I found out, 542 00:40:01,106 --> 00:40:07,542 this was Mario panda, not so well known and those are the nice like. 543 00:40:07,542 --> 00:40:11,929 The second or third highly cited philosopher, but 544 00:40:11,929 --> 00:40:17,235 if nobody kind of seem to know in the lit that's available some 545 00:40:17,235 --> 00:40:22,643 kinda obscurity, that kinda systems idea seems to have like, 546 00:40:22,643 --> 00:40:26,541 I don't know what it was kinda interesting. 547 00:40:26,541 --> 00:40:31,570 So some useful leads for us, I would say IRS supposedly like they have 548 00:40:31,570 --> 00:40:36,597 public data sets and they're willing to I kind of talked to two people 549 00:40:36,597 --> 00:40:41,714 from the IRS statistical analysis team they were interested in it, so 550 00:40:41,714 --> 00:40:47,640 I kind of sent feelers out and there's something, this is really interesting. 551 00:40:47,640 --> 00:40:48,507 Those in, 552 00:40:59,926 --> 00:41:01,277 So this is a project, 553 00:41:01,277 --> 00:41:06,766 NSF-funded project in company's department in Urbana Champaign, West San Diego. 554 00:41:06,766 --> 00:41:11,768 My collaborator there is and they built a whole platform for 555 00:41:11,768 --> 00:41:18,572 integrating reproducible research where like, you can upload your programs, 556 00:41:18,572 --> 00:41:23,590 Python programs, R programs and data sets in a secure way. 557 00:41:23,590 --> 00:41:26,110 They will kinda make sure that it's structured 558 00:41:26,110 --> 00:41:30,290 in the right way whether anybody can play visit and reproduce it. 559 00:41:30,290 --> 00:41:35,350 So I think next week well my travel to San Diego 560 00:41:35,350 --> 00:41:40,720 is initiated with the PA office project, 561 00:41:40,720 --> 00:41:46,035 and they were interested in kind of trying to 562 00:41:46,035 --> 00:41:52,460 integrate our projects that we are working on as part of this so 563 00:41:52,460 --> 00:41:57,180 it's something that I feel, well it's like simulation or 564 00:41:57,180 --> 00:42:05,146 data analysis something that this can be really useful for us. 565 00:42:05,146 --> 00:42:10,415 And there's also this person from Paul Brenner. 566 00:42:10,415 --> 00:42:13,975 He's kind of the director of data analytics and 567 00:42:13,975 --> 00:42:17,825 computational science program in the Notre Dame, and 568 00:42:17,825 --> 00:42:21,542 he had this kind of really heavy duty public service data 569 00:42:22,662 --> 00:42:26,852 platform that can scale, talking about micro simulation, 570 00:42:26,852 --> 00:42:32,272 he present his work was on analyzing Twitter data from 2 billion people surfing 571 00:42:32,272 --> 00:42:36,872 and news sources from 50 countries, all the news data they select, so they 572 00:42:38,720 --> 00:42:43,530 might be able to kind of provide some service for us. 573 00:42:43,530 --> 00:42:50,730 And there's also this program in South Florida 574 00:42:50,730 --> 00:42:55,450 called Deep Agent by this researcher, Ivan Garibay. 575 00:42:55,450 --> 00:43:00,290 And it was kind of a pretty interesting. 576 00:43:00,290 --> 00:43:05,780 And also I found some interesting micro simulation of 577 00:43:05,780 --> 00:43:11,530 disaster response in George Mason University Center for 578 00:43:11,530 --> 00:43:16,030 Social Complexity that they're trying to build response to natural 579 00:43:16,030 --> 00:43:20,930 disasters by doing agent based modeling, network models and everything. 580 00:43:20,930 --> 00:43:22,270 >> That makes a lot of sense. 581 00:43:22,270 --> 00:43:23,760 >> Yeah, very good. 582 00:43:23,760 --> 00:43:28,820 >> Okay, they were focused on kind of nuclear event response. 583 00:43:28,820 --> 00:43:34,630 And I remember Shawn was very interested in it. 584 00:43:34,630 --> 00:43:37,861 There's gonna be just. 585 00:43:37,861 --> 00:43:42,087 But I overall I felt like the place is pretty good for our students to do, 586 00:43:42,087 --> 00:43:46,174 like as I said, whatever we do in data mining and the theory courses, 587 00:43:46,174 --> 00:43:50,704 system theory, game theory, as well as the modeling simulation courses. 588 00:43:50,704 --> 00:43:54,389 They're all kind of conducive for 589 00:43:54,389 --> 00:43:59,217 publication are places in the conference and 590 00:43:59,217 --> 00:44:03,918 it's based on my experience teaching here I 591 00:44:03,918 --> 00:44:08,872 felt like if we can somehow kind of cater kind of 592 00:44:08,872 --> 00:44:13,572 instructors the term papers as like a public 593 00:44:13,572 --> 00:44:18,019 publication last way to get published and 594 00:44:18,019 --> 00:44:22,620 third these venues might be interesting. 595 00:44:22,620 --> 00:44:23,685 >> Lots of journals too. 596 00:44:23,685 --> 00:44:24,265 Yeah. >> Yeah. 597 00:44:24,265 --> 00:44:26,475 >> And something I learned, I didn't know this. 598 00:44:26,475 --> 00:44:30,165 I think computer science conferences have a higher regard than most journals. 599 00:44:30,165 --> 00:44:31,110 >> Yeah. 600 00:44:31,110 --> 00:44:34,750 >> Whereas most fields you have to move it on a journal before [CROSSTALK] are really 601 00:44:34,750 --> 00:44:35,612 good- >> Yeah, 602 00:44:35,612 --> 00:44:40,156 it's interesting because the organizers there were like two 603 00:44:40,156 --> 00:44:45,580 organizers whose tenure I think is up for kind of being evaluated. 604 00:44:45,580 --> 00:44:49,404 They say we don't want to publish here because it's not recognized. 605 00:44:49,404 --> 00:44:54,341 And there's these two other organizers where the computer site who we 606 00:44:54,341 --> 00:44:59,027 said no conferences are recognized we should kind of keep it public 607 00:44:59,027 --> 00:45:03,224 whether conferences count or not like that sort of things. 608 00:45:03,224 --> 00:45:04,770 >> That are complicated. 609 00:45:06,130 --> 00:45:10,782 >> I stopped putting conferences on my things, like if I'm supposed to 610 00:45:10,782 --> 00:45:15,846 create a exciting portfolio I don't feature conferences at all anymore. 611 00:45:15,846 --> 00:45:22,484 I used to think that it was secondary but definitely valued. 612 00:45:22,484 --> 00:45:23,271 But finally. 613 00:45:23,271 --> 00:45:28,200 >> Yeah, like another metric to the graduating students this year. 614 00:45:28,200 --> 00:45:33,170 They've got tenure line positions in kind of colleges and 615 00:45:33,170 --> 00:45:36,660 universities and postdocs based on their publication. 616 00:45:36,660 --> 00:45:39,750 This conference alone like they published for three years and 617 00:45:39,750 --> 00:45:43,850 they say like that was like good enough for them to get a position. 618 00:45:43,850 --> 00:45:47,601 It's like it seems like it's very much dependent on the field and everything. 619 00:45:47,601 --> 00:45:52,436 So but okay, but why bother with CSS this is like my 620 00:45:52,436 --> 00:45:56,245 agenda very personal kind of opinion. 621 00:45:56,245 --> 00:45:59,920 So this fall like I already kind of put it up there. 622 00:45:59,920 --> 00:46:04,670 I feel like there's like a lot of money and there's like, for 623 00:46:04,670 --> 00:46:09,290 example, all the people who had this, the de novo approach, 624 00:46:09,290 --> 00:46:13,036 they all had $12 million grants and everything. 625 00:46:13,036 --> 00:46:19,653 [LAUGH] And they were like arguing with me you don't know what you're asking. 626 00:46:19,653 --> 00:46:24,214 But I was happy because like our paper that we submitted, 627 00:46:24,214 --> 00:46:28,689 we fell second is the best paper list among 40 papers. 628 00:46:28,689 --> 00:46:34,472 That was with zero funding, so it's like we did it in our free time and 629 00:46:34,472 --> 00:46:41,081 we didn't have any DARPA funding or NSF funding but we are happy that we got it. 630 00:46:41,081 --> 00:46:46,036 But anyways, I feel like there's something associated with this indexing of 631 00:46:46,036 --> 00:46:48,860 computational sources translate. 632 00:46:48,860 --> 00:46:53,140 For example, even IEEE now has a journal in network science and 633 00:46:53,140 --> 00:46:57,987 they have a tag computation sources and why, because if engineers want 634 00:46:57,987 --> 00:47:02,590 to publish papers, they will not go anywhere without the tag IEEE so 635 00:47:02,590 --> 00:47:06,015 they have this IEEE and backed with competition. 636 00:47:06,015 --> 00:47:09,283 >> It's not the social science it's the computational. 637 00:47:09,283 --> 00:47:11,968 >> Yeah. [LAUGH] Unfortunately I feel like this 638 00:47:11,968 --> 00:47:16,644 club pretty much like there's a modern I feel like we are doing mostly of what 639 00:47:16,644 --> 00:47:19,300 computational source science people do. 640 00:47:19,300 --> 00:47:23,980 So and in terms of publications for students and 641 00:47:23,980 --> 00:47:28,450 everything that agencies recognize the term more than systems. 642 00:47:28,450 --> 00:47:30,960 And for undergraduate outreach, 643 00:47:30,960 --> 00:47:33,650 everybody's talking computational social science. 644 00:47:34,840 --> 00:47:38,120 And. >> I have to have the word social, though. 645 00:47:38,120 --> 00:47:40,130 That's the whole world isn't social, right? 646 00:47:40,130 --> 00:47:41,010 I mean, yeah, but. 647 00:47:41,010 --> 00:47:43,160 >> Otherwise it would sound like computer science. 648 00:47:43,160 --> 00:47:45,989 >> Yeah. So the computational social science is 649 00:47:45,989 --> 00:47:50,834 like, and I've seen like so many job openings and posters, so yeah. 650 00:47:50,834 --> 00:47:55,370 In even a big business running million in Michigan. 651 00:47:55,370 --> 00:47:57,420 They're opening tenure lines with that. 652 00:47:57,420 --> 00:48:01,710 We are looking for tenure line faculty in computational social science. 653 00:48:01,710 --> 00:48:05,930 So that and for those people that can talk to some of them and they always say, 654 00:48:05,930 --> 00:48:10,120 hey, we published in this two, three flagship conferences and everybody knows. 655 00:48:10,120 --> 00:48:16,070 Since those societies and therefore we all have. 656 00:48:16,070 --> 00:48:20,823 So I yeah, like I feel that because of all that like when you 657 00:48:20,823 --> 00:48:25,581 can find a new home for systems science eventually I don't 658 00:48:25,581 --> 00:48:30,171 know what it's always been wanted for long time for us. 659 00:48:31,190 --> 00:48:35,650 I feel like somehow, maybe some 660 00:48:35,650 --> 00:48:40,611 certificates or like minor types. 661 00:48:40,611 --> 00:48:45,922 When minors, a bachelor's in kind of social science or something and like even 662 00:48:45,922 --> 00:48:51,154 more controversially, I would say that there's no harm in changing the title 663 00:48:51,154 --> 00:48:56,169 of the program you might accomplish with social sciences like [INAUDIBLE]. 664 00:48:56,169 --> 00:49:01,054 And if that's the way that's my kind of very, very controversial take on it, 665 00:49:01,054 --> 00:49:04,468 but it's okay to do system settings within a company. 666 00:49:04,468 --> 00:49:08,460 So [LAUGH] that's like my I'll end on that note. 667 00:49:08,460 --> 00:49:09,034 >> Okay. 668 00:49:09,034 --> 00:49:09,744 >> Okay. 669 00:49:09,744 --> 00:49:11,879 >> Well thank you for presenting [INAUDIBLE]. 670 00:49:11,879 --> 00:49:17,039 >> [APPLAUSE] >> Here's another comment 671 00:49:17,039 --> 00:49:21,684 from Shane when he says George Mason's 672 00:49:21,684 --> 00:49:27,040 system science program is a SS type program. 673 00:49:27,040 --> 00:49:31,668 Northeastern's is complex systems and networks. 674 00:49:31,668 --> 00:49:36,205 So there's a couple of suggestions. 675 00:49:36,205 --> 00:49:37,406 Yeah, please feel free to comment. 676 00:49:37,406 --> 00:49:39,851 You could actually take a microphone. 677 00:49:39,851 --> 00:49:40,900 We can hear you. 678 00:49:40,900 --> 00:49:42,346 If somebody has a mic and wants to make a comment or 679 00:49:42,346 --> 00:49:43,449 question go ahead and give it a shot. 680 00:49:43,449 --> 00:49:45,351 I think it will work fine. 681 00:49:45,351 --> 00:49:49,655 Just I haven't made it so you can't unmute and talk. 682 00:49:49,655 --> 00:49:53,455 So type it if you want and I can read it or whatever. 683 00:49:53,455 --> 00:49:55,308 We could actually put that on the screen right now. 684 00:49:55,308 --> 00:49:58,046 Yeah, it's not that I can see him. 685 00:49:58,046 --> 00:50:06,380 I got the thing pointed out that way but let's see. 686 00:50:06,380 --> 00:50:08,985 I guess this is off. 687 00:50:08,985 --> 00:50:10,907 It must have got bumped. 688 00:50:10,907 --> 00:50:15,834 So I was just gonna put the chat thing off if people want to chat then everybody will 689 00:50:15,834 --> 00:50:16,930 be able to see it. 690 00:50:16,930 --> 00:50:20,468 To do that I need I think I'll stop sharing this for now. 691 00:50:20,468 --> 00:50:23,449 And then I will go to the top. 692 00:50:23,449 --> 00:50:23,949 I need to go. 693 00:50:27,014 --> 00:50:33,807 Probably take us through that. 694 00:50:33,807 --> 00:50:35,478 >> No,I need to go too I guess. 695 00:50:38,801 --> 00:50:39,301 Very strange. 696 00:50:43,330 --> 00:50:44,110 So what, so? 697 00:50:44,110 --> 00:50:44,882 >> Are you ready? 698 00:50:44,882 --> 00:50:45,643 >> Yeah, I have a question. 699 00:50:45,643 --> 00:50:49,419 So why do so many people think that NetLogo is agent based model? 700 00:50:49,419 --> 00:50:52,715 I've also heard that term for that logo. 701 00:50:52,715 --> 00:50:54,081 >> It is. >> It is. 702 00:50:54,081 --> 00:50:54,847 Yeah, like men do. 703 00:50:54,847 --> 00:50:57,583 It was the kind of I think villains. 704 00:50:57,583 --> 00:51:02,583 He and others like when they started that was the platform 705 00:51:02,583 --> 00:51:06,749 that they started off with and one or even like so 706 00:51:06,749 --> 00:51:11,749 much that now what is it the idea of agent based modeling is 707 00:51:11,749 --> 00:51:16,870 like in the sharing sense, our sharing senses like that. 708 00:51:16,870 --> 00:51:22,292 We are expected to kind of come up with some macroscopic phenomena to explain. 709 00:51:22,292 --> 00:51:27,332 So it's like coming up with the my using the micro to explain the macro, 710 00:51:27,332 --> 00:51:30,955 so that's the way APM was originally envisioned. 711 00:51:30,955 --> 00:51:35,210 And then all the years as it became more and more handy. 712 00:51:35,210 --> 00:51:39,354 And it is basically your the models look like that. 713 00:51:39,354 --> 00:51:42,344 They put in the macro to get a macro. 714 00:51:42,344 --> 00:51:46,580 So there is nothing like you understood some generative 715 00:51:46,580 --> 00:51:51,546 mechanism of the underlying phenomena you're trying to model. 716 00:51:51,546 --> 00:51:55,329 So that's the it's become like. 717 00:51:55,329 --> 00:51:58,566 >> It's the only way that people are inputting. 718 00:51:58,566 --> 00:51:59,221 >> Yes. 719 00:51:59,221 --> 00:52:01,521 >> Therefore [INAUDIBLE]. 720 00:52:01,521 --> 00:52:02,417 >> Yeah. 721 00:52:02,417 --> 00:52:04,065 Yes, yes. 722 00:52:04,065 --> 00:52:08,918 And it's one of those things that some of my collaborators who are like 723 00:52:08,918 --> 00:52:12,628 pretty neutral about net logo, they want to kind of. 724 00:52:12,628 --> 00:52:16,980 They will argue they've argued that instead of messing with the actual 725 00:52:16,980 --> 00:52:19,340 mechanisms you're trying to model. 726 00:52:19,340 --> 00:52:24,301 So we are trying to provide some alternative things where like data will 727 00:52:24,301 --> 00:52:28,776 simulate it one way and you program it and you get it another way. 728 00:52:28,776 --> 00:52:34,061 And you want to kind of building on some case study to compare whether NetLogo, 729 00:52:34,061 --> 00:52:38,165 how it kinda schedules the interaction between the agents. 730 00:52:38,165 --> 00:52:39,861 Whether that's kinda messing up with the model. 731 00:52:39,861 --> 00:52:44,100 Like whether that's evolved, 732 00:52:44,100 --> 00:52:48,967 that logo's kind of coarse graining 733 00:52:48,967 --> 00:52:54,629 the details of the agent based mechanism. 734 00:52:54,629 --> 00:52:58,240 >> Okay, yeah, I don't have other questions coming in. 735 00:52:58,240 --> 00:53:00,786 Did anybody else try to take the microphone and ask a question? 736 00:53:00,786 --> 00:53:03,144 I think it should work but. 737 00:53:03,144 --> 00:53:04,124 >> Can you guys hear me? 738 00:53:04,124 --> 00:53:05,549 >> Yes, we hear you. 739 00:53:05,549 --> 00:53:11,290 >> Hi, Rajash and I talk about this stuff every day so I had other stuff to do. 740 00:53:11,290 --> 00:53:12,662 I wasn't able to make it in person. 741 00:53:12,662 --> 00:53:17,895 >> Enjoy little bit of fishing [INAUDIBLE], right? 742 00:53:17,895 --> 00:53:20,031 [LAUGH] >> Thank you. 743 00:53:20,031 --> 00:53:22,328 But yeah, I found it. 744 00:53:22,328 --> 00:53:28,188 There's a list of comparing agent based modeling software just on Wikipedia. 745 00:53:28,188 --> 00:53:31,010 NetLogo is easily the most popular. 746 00:53:31,010 --> 00:53:33,390 It's probably the only one you'll hear 747 00:53:33,390 --> 00:53:38,400 if you're not deeply invested in finally finding agent based modeling platforms, 748 00:53:38,400 --> 00:53:42,460 but there are a few others that are out there. 749 00:53:42,460 --> 00:53:45,670 I just think that NetLogo might be the most accessible and 750 00:53:45,670 --> 00:53:49,490 so it's kind of synonymous with the practice at this point. 751 00:53:51,230 --> 00:53:56,290 There is a Python platform called Mesa that I've been trying to use and 752 00:53:56,290 --> 00:54:02,311 it's kind of nice, but it's still just it's still really early in development. 753 00:54:02,311 --> 00:54:04,941 I don't even know if they would call it at a 1.0 level yet. 754 00:54:04,941 --> 00:54:07,245 So it's missing some things but 755 00:54:07,245 --> 00:54:12,638 it's certainly more flexible than NetLogo, because it's just Python. 756 00:54:12,638 --> 00:54:16,868 So you can plug in any other Python libraries you want and 757 00:54:16,868 --> 00:54:21,286 do you've got more flexibility in what you can do with it. 758 00:54:21,286 --> 00:54:23,501 But it's harder to use, right? 759 00:54:23,501 --> 00:54:24,845 So that's the trade off. 760 00:54:24,845 --> 00:54:30,033 >> Yeah, I think the collaborator from San Diego. 761 00:54:30,033 --> 00:54:33,498 His dissertation work is on these. 762 00:54:33,498 --> 00:54:38,240 What's the right way to do agent based modeling? 763 00:54:38,240 --> 00:54:40,093 So he's a computer science person. 764 00:54:40,093 --> 00:54:44,674 He thinks there's this new language called Elixir, 765 00:54:44,674 --> 00:54:51,718 which is like the compiler state itself does like a lot of ingrained concurrency. 766 00:54:51,718 --> 00:54:57,858 And he's kind of arguing this dissertation. 767 00:54:57,858 --> 00:55:01,159 We need kind of threads and 768 00:55:01,159 --> 00:55:05,821 processes running to simulate a VM. 769 00:55:05,821 --> 00:55:07,476 And if you don't do that, 770 00:55:07,476 --> 00:55:12,841 you are inherently serializing what's like a genuinely parallel social process. 771 00:55:12,841 --> 00:55:16,379 So he's kind of working on some case studies to show that if you do it in 772 00:55:16,379 --> 00:55:20,603 NetLogo, you get one result but if you do it concurrently you get another result. 773 00:55:20,603 --> 00:55:27,055 >> [INAUDIBLE] >> Yes, 774 00:55:27,055 --> 00:55:32,065 yeah, unless you very carefully take it out on 775 00:55:32,065 --> 00:55:37,211 multiple processors where these processors. 776 00:55:37,211 --> 00:55:38,529 >> Yeah, so again render. 777 00:55:38,529 --> 00:55:40,144 >> There's no way to avoid that. 778 00:55:40,144 --> 00:55:45,017 >> Yeah, it's a physical impossibility to use a single processor. 779 00:55:45,017 --> 00:55:47,768 It's really parallel period. 780 00:55:47,768 --> 00:55:51,960 >> Well, you could do it sequentially. 781 00:55:51,960 --> 00:55:57,916 >> I think it's pretty much been fully demonstrated. 782 00:55:57,916 --> 00:56:01,377 That in order to get genuinely parallel processes, 783 00:56:01,377 --> 00:56:05,319 you need processors that are independent from each other. 784 00:56:05,319 --> 00:56:09,402 So I think that's pretty well known that you'd 785 00:56:09,402 --> 00:56:13,290 be skeptical because it doesn't seem like. 786 00:56:14,773 --> 00:56:20,519 If you don't update things, you can do what you do in parallel and 787 00:56:20,519 --> 00:56:23,982 sequence and update it at the very end. 788 00:56:23,982 --> 00:56:25,168 >> Yeah. 789 00:56:25,168 --> 00:56:30,127 >> Yeah, so San Diego [INAUDIBLE]. 790 00:56:30,127 --> 00:56:35,121 He's arguing that you have to be integrated with the OS level, 791 00:56:35,121 --> 00:56:41,343 that you make each agent into like a separate processor thread with his 792 00:56:41,343 --> 00:56:47,392 with its own clock and everything so that you can kind of have micro control. 793 00:56:47,392 --> 00:56:51,725 So that his decision that's what he kind of saying like 794 00:56:51,725 --> 00:56:53,992 I want to pass at that level. 795 00:56:53,992 --> 00:56:59,152 Go all the way to have each agent have a separate thread at the OS level and 796 00:56:59,152 --> 00:57:02,352 then do- >> 99% of them [CROSSTALK]. 797 00:57:02,352 --> 00:57:04,483 If they do, you're stuck. 798 00:57:04,483 --> 00:57:05,802 >> Yeah, so that's my guess. 799 00:57:08,528 --> 00:57:10,866 Yeah. 800 00:57:10,866 --> 00:57:12,703 >> Yeah. 801 00:57:12,703 --> 00:57:17,993 >> I found repast HPC, which I went to a one day workshop on. 802 00:57:17,993 --> 00:57:20,954 I caught absolutely nothing [INAUDIBLE] eight hours of work. 803 00:57:20,954 --> 00:57:28,214 So it's a pretty steep learning curve to get into some of these things. 804 00:57:28,214 --> 00:57:31,368 >> [INAUDIBLE] >> It still exists but 805 00:57:31,368 --> 00:57:34,145 nobody uses it that I know of. 806 00:57:34,145 --> 00:57:36,771 [INAUDIBLE] I don't know. 807 00:57:36,771 --> 00:57:40,095 Probably some legacy stuff that it still runs. 808 00:57:40,095 --> 00:57:42,177 Yeah. 809 00:57:42,177 --> 00:57:45,995 So the reason that that love is alive is the government invested $20 810 00:57:45,995 --> 00:57:49,760 million in getting it to where it is, so that is a really useful tool. 811 00:57:49,760 --> 00:57:52,856 But it also includes system dynamics functionality now so 812 00:57:52,856 --> 00:57:56,395 that you can actually build some differential equations stuff and 813 00:57:56,395 --> 00:57:58,932 parallel and end it with sort of a hybrid model. 814 00:57:58,932 --> 00:58:02,931 And if they were using that module, they weren't doing state of the art system 815 00:58:02,931 --> 00:58:06,462 dynamics work but they were doing system dynamics, might be wrong. 816 00:58:06,462 --> 00:58:12,210 And of course there's Any Logic which came out of Russia which is 817 00:58:12,210 --> 00:58:17,319 truly able to do multi hybrids with good solid support for 818 00:58:17,319 --> 00:58:23,402 Monte Carlo work or agent based thinking and persistent dynamics. 819 00:58:23,402 --> 00:58:26,321 However, it's not in the public domain. 820 00:58:26,321 --> 00:58:34,871 >> So this project, this net worth $12 million project and 821 00:58:34,871 --> 00:58:39,663 they're claiming the PA claim. 822 00:58:39,663 --> 00:58:41,852 It's got all the big names. 823 00:58:41,852 --> 00:58:45,803 Epstein is there and I think this is there, but 824 00:58:45,803 --> 00:58:52,503 the claim is that they're going to do social simulation like fluid dynamics. 825 00:58:52,503 --> 00:58:57,643 And I kind of suspect that social fluids are, 826 00:58:57,643 --> 00:59:05,069 there's nothing like missing a lot of- >> Where's the agency on fluid? 827 00:59:05,069 --> 00:59:08,098 >> [LAUGH] >> The basic response was like what the so 828 00:59:08,098 --> 00:59:11,138 we can all go home so we do not have to do this. 829 00:59:11,138 --> 00:59:13,802 So therefore, I did not. 830 00:59:13,802 --> 00:59:17,383 >> But I would not discourage us from trying everything. 831 00:59:17,383 --> 00:59:18,074 >> Yeah, absolutely. 832 00:59:18,074 --> 00:59:22,068 We have a in silico simulation and 833 00:59:22,068 --> 00:59:27,062 their Twitter and we don't need anything 834 00:59:27,062 --> 00:59:31,410 else because they too much >> And 835 00:59:31,410 --> 00:59:33,742 they can't predict the next week at all. 836 00:59:33,742 --> 00:59:38,395 >> [LAUGH] >> Yeah, I was asking about like whether 837 00:59:38,395 --> 00:59:44,276 to call it known unknowns and unknown unknowns. 838 00:59:44,276 --> 00:59:48,915 So we kind of decided to get rid of it, but 839 00:59:48,915 --> 00:59:54,653 these are things where like new platforms where? 840 00:59:54,653 --> 00:59:59,778 Yeah, these two I met in CSS with I think that he 841 00:59:59,778 --> 01:00:04,774 was actually writing for Bosnia in this day, 842 01:00:04,774 --> 01:00:08,732 you are a very interesting person. 843 01:00:08,732 --> 01:00:15,292 And of course Josh has been there so I think I actually remember these names. 844 01:00:15,292 --> 01:00:16,123 >> Yeah, for sure. 845 01:00:16,123 --> 01:00:17,908 >> Yeah, a lot of people. 846 01:00:17,908 --> 01:00:19,212 >> Well, let's wrap up. 847 01:00:19,212 --> 01:00:22,260 >> Yeah, I think we're a little past one and 848 01:00:22,260 --> 01:00:25,019 one more thank you for >> [APPLAUSE] 849 01:00:25,019 --> 01:00:27,838 >> [INAUDIBLE] with personal opinions and 850 01:00:27,838 --> 01:00:31,193 an opportunity for some lively discussion. 851 01:00:31,193 --> 01:00:32,734 Thanks everybody. 852 01:00:32,734 --> 01:00:35,770 I think I got everybody down who attended in the ether, 853 01:00:35,770 --> 01:00:39,007 which was easily four times as many as attended in person. 854 01:00:39,007 --> 01:00:42,434 [LAUGH] >> So that's good. 855 01:00:42,434 --> 01:00:45,507 >> Happy holidays to everybody. 856 01:00:45,507 --> 01:00:48,388 I won't be seeing many of you again now because it's getting pretty much to 857 01:00:48,388 --> 01:00:49,175 the end of the term. 858 01:00:49,175 --> 01:00:50,328 And we have a new series. 859 01:00:50,328 --> 01:00:54,420 I've got most of the speakers already lined up for next term amazingly and 860 01:00:54,420 --> 01:00:55,946 a few more slots to fill in. 861 01:00:55,946 --> 01:00:58,212 Thanks for your suggestions by the way. 862 01:00:58,212 --> 01:00:59,572 >> Is there a class next week? 863 01:00:59,572 --> 01:01:00,334 >> No, no. 864 01:01:00,334 --> 01:01:02,189 >> Okay. 865 01:01:02,189 --> 01:01:04,185 >> Did they stopped there? 866 01:01:04,185 --> 01:01:07,125 >> I think I have that I've got the controls. 867 01:01:07,125 --> 01:01:08,834 I'll stop the recording now. 868 01:01:14,035 --> 01:01:15,394 Stop recording [INAUDIBLE]