1 00:00:00,090 --> 00:00:02,430 The following content is provided under a Creative 2 00:00:02,430 --> 00:00:03,820 Commons license. 3 00:00:03,820 --> 00:00:06,030 Your support will help MIT OpenCourseWare 4 00:00:06,030 --> 00:00:10,120 continue to offer high quality educational resources for free. 5 00:00:10,120 --> 00:00:12,660 To make a donation or to view additional materials 6 00:00:12,660 --> 00:00:16,620 from hundreds of MIT courses, visit MIT OpenCourseWare 7 00:00:16,620 --> 00:00:17,992 at ocw.mit.edu. 8 00:00:23,070 --> 00:00:25,570 WILLIAM BONVILLIAN: Tonight is a deep dive into DARPA, which 9 00:00:25,570 --> 00:00:29,880 is a very different model. 10 00:00:29,880 --> 00:00:34,190 And we've talked around DARPA, but-- 11 00:00:34,190 --> 00:00:36,130 we talked about snippets of some of the stuff 12 00:00:36,130 --> 00:00:38,350 that we're going to be going through, 13 00:00:38,350 --> 00:00:40,480 but tonight will be kind of a deep dive 14 00:00:40,480 --> 00:00:43,600 into looking this model. 15 00:00:43,600 --> 00:00:45,250 And the model is an important one 16 00:00:45,250 --> 00:00:50,020 because we've gotten so much out of DARPA. 17 00:00:50,020 --> 00:00:52,900 It's been a remarkable run. 18 00:00:52,900 --> 00:00:58,620 And so we'll look at the foundational period. 19 00:00:58,620 --> 00:01:00,620 And there are a lot of good MIT stories in here, 20 00:01:00,620 --> 00:01:03,820 which we'll get on the table. 21 00:01:03,820 --> 00:01:08,020 But we'll also try to frame really 22 00:01:08,020 --> 00:01:13,580 how this organization, institution functions. 23 00:01:13,580 --> 00:01:18,440 So this is a JCR Licklider story-- 24 00:01:18,440 --> 00:01:21,610 one of MIT's all-time greats. 25 00:01:21,610 --> 00:01:26,530 And he doesn't get talked about that much at MIT, 26 00:01:26,530 --> 00:01:29,020 but he really is foundational. 27 00:01:29,020 --> 00:01:40,870 And he comes in at a time when computing is really completely 28 00:01:40,870 --> 00:01:42,670 different than it is today, and is 29 00:01:42,670 --> 00:01:49,240 able, with his team at DARPA, to really work through-- 30 00:01:49,240 --> 00:01:52,840 really in an intergenerational between DARPA program manager 31 00:01:52,840 --> 00:01:55,390 and office director generations-- 32 00:01:55,390 --> 00:01:59,070 really rework the entire sector. 33 00:01:59,070 --> 00:02:03,100 So we've talked a little about Whirlwind and SAGE earlier. 34 00:02:03,100 --> 00:02:07,000 And just to remind you, this was one 35 00:02:07,000 --> 00:02:09,220 of the foundational early computer systems. 36 00:02:09,220 --> 00:02:12,790 This was MIT's at Whirlwind and SAGE. 37 00:02:12,790 --> 00:02:16,960 And this is Jay Forrester, and he's holding the magnetic core 38 00:02:16,960 --> 00:02:20,290 memory that he and one of his graduate students 39 00:02:20,290 --> 00:02:22,840 came up with, which enabled real-time computing, 40 00:02:22,840 --> 00:02:25,210 a really critical advance. 41 00:02:25,210 --> 00:02:30,970 And you start to see at Whirlwind, in the upper-left-- 42 00:02:30,970 --> 00:02:33,700 which merges into the SAGE system-- 43 00:02:33,700 --> 00:02:36,400 you start to see where computing is going. 44 00:02:36,400 --> 00:02:45,230 So that picture of this computer operator, 45 00:02:45,230 --> 00:02:47,960 this lady sitting in front of a keyboard in front of a cathode 46 00:02:47,960 --> 00:02:50,480 ray screen. 47 00:02:50,480 --> 00:02:53,600 That's this, and there it is. 48 00:02:53,600 --> 00:02:59,480 There it is at late '40s, early '50s when this is going on. 49 00:02:59,480 --> 00:03:03,530 And how did this all evolve? 50 00:03:03,530 --> 00:03:06,080 So on the lower left is George Valley, 51 00:03:06,080 --> 00:03:07,870 who was an MIT faculty member. 52 00:03:07,870 --> 00:03:10,520 He was part of the Rad Lab. 53 00:03:10,520 --> 00:03:14,120 And Valley is on a scientific advisory board to the Air 54 00:03:14,120 --> 00:03:18,620 Force, and he sounds the alarm, essentially alerting 55 00:03:18,620 --> 00:03:22,820 the Air Force and frankly, the military in general, 56 00:03:22,820 --> 00:03:29,280 that the Soviet Union has developed a long-range bombing 57 00:03:29,280 --> 00:03:29,820 aircraft. 58 00:03:29,820 --> 00:03:32,340 They have nuclear weapons, and they have, therefore, 59 00:03:32,340 --> 00:03:34,140 first strike capability. 60 00:03:34,140 --> 00:03:37,440 There is absolutely nothing in their way 61 00:03:37,440 --> 00:03:40,140 of reaching the northern part of the United States. 62 00:03:40,140 --> 00:03:43,620 And that's a terrifying realization moment. 63 00:03:43,620 --> 00:03:48,100 So Valley gets the Air Force mobilized around an air defense 64 00:03:48,100 --> 00:03:49,140 mission. 65 00:03:49,140 --> 00:03:52,440 And computers are going to be at the heart of it. 66 00:03:52,440 --> 00:03:54,060 And he goes to Jay Forrester, who 67 00:03:54,060 --> 00:03:58,560 was developing this Whirlwind system for the Navy. 68 00:03:58,560 --> 00:04:01,590 The Navy's is canceling Whirlwind, 69 00:04:01,590 --> 00:04:04,500 so Valley just moves Forrester and the team, 70 00:04:04,500 --> 00:04:10,980 and the Whirlwind system over to the SAGE project of developing 71 00:04:10,980 --> 00:04:13,410 an air defense and controllers. 72 00:04:13,410 --> 00:04:16,380 And the picture here is what-- 73 00:04:16,380 --> 00:04:17,940 essentially what the picture looks 74 00:04:17,940 --> 00:04:20,010 like of all these different defense 75 00:04:20,010 --> 00:04:25,350 systems, from early missiles, and artillery, and fighter 76 00:04:25,350 --> 00:04:31,120 aircraft, all being linked into a communications and computer 77 00:04:31,120 --> 00:04:31,620 system. 78 00:04:31,620 --> 00:04:35,490 And then fascinatingly, the communications lines 79 00:04:35,490 --> 00:04:38,310 are across phone lines. 80 00:04:38,310 --> 00:04:41,430 So the computers are running a communication system that comes 81 00:04:41,430 --> 00:04:44,040 into them across phone lines. 82 00:04:44,040 --> 00:04:49,930 The picture of those Air Force airmen 83 00:04:49,930 --> 00:04:51,910 sitting in front of cathode ray screens, 84 00:04:51,910 --> 00:04:55,030 they're holding this peculiar thing. 85 00:04:55,030 --> 00:04:55,930 It's a light gun. 86 00:04:55,930 --> 00:04:57,910 It's the mouse. 87 00:04:57,910 --> 00:05:00,940 So you start to see these iconic pictures of where all this 88 00:05:00,940 --> 00:05:04,120 stuff that we just live with-- 89 00:05:04,120 --> 00:05:05,920 where it all starts-- come from. 90 00:05:05,920 --> 00:05:11,530 And a lot of it came from the Whirlwind and SAGE systems. 91 00:05:11,530 --> 00:05:13,960 MIT doesn't want to go into the business of manufacturing 92 00:05:13,960 --> 00:05:14,560 computers. 93 00:05:14,560 --> 00:05:19,000 They make the initial computers for the SAGE system, 94 00:05:19,000 --> 00:05:21,370 but they don't want to be in the computer manufacturing 95 00:05:21,370 --> 00:05:22,540 business. 96 00:05:22,540 --> 00:05:29,710 So the SAGE system becomes the IBM 700 series. 97 00:05:29,710 --> 00:05:33,580 This is how IBM really starts to scale up 98 00:05:33,580 --> 00:05:35,380 in mainframe computing, and starts 99 00:05:35,380 --> 00:05:38,810 building the computing systems for this air defense system. 100 00:05:38,810 --> 00:05:42,100 So again, it's an example of long-term patient military 101 00:05:42,100 --> 00:05:44,200 support for technology development 102 00:05:44,200 --> 00:05:45,760 that we've seen before. 103 00:05:45,760 --> 00:05:48,010 And here it is again, only this time, 104 00:05:48,010 --> 00:05:49,240 it's sponsoring computing. 105 00:05:49,240 --> 00:05:53,830 So Mitch Waldrop's book The Dream Machine I think 106 00:05:53,830 --> 00:05:58,780 is really the classic on the evolution of the Information 107 00:05:58,780 --> 00:05:59,860 Age. 108 00:05:59,860 --> 00:06:03,370 It is full of wonderful stories, and I 109 00:06:03,370 --> 00:06:04,880 feel that it's really well written, 110 00:06:04,880 --> 00:06:07,160 and I couldn't resist giving you a big dose of it. 111 00:06:07,160 --> 00:06:10,127 But there's a lot of great MIT stories in here, but lots 112 00:06:10,127 --> 00:06:12,460 of great stories about the foundation of where computing 113 00:06:12,460 --> 00:06:14,470 came from and how it arrived-- 114 00:06:14,470 --> 00:06:17,860 that fits with the material that we've been covering in class. 115 00:06:17,860 --> 00:06:20,170 So we'll go through some of this key stuff, 116 00:06:20,170 --> 00:06:24,310 and then stuff will lead us in some Q and A. 117 00:06:24,310 --> 00:06:30,910 Computing before Licklider, it's non-interactive, it's not fun, 118 00:06:30,910 --> 00:06:32,500 it's not exciting. 119 00:06:32,500 --> 00:06:36,040 It's behind glass walls, and you'd 120 00:06:36,040 --> 00:06:40,330 be met by like these high priests and priestesses who are 121 00:06:40,330 --> 00:06:44,380 wearing white coats and gloves. 122 00:06:44,380 --> 00:06:50,260 And you'd carry your stack of punch cards in, 123 00:06:50,260 --> 00:06:54,100 and you'd hand over this offering of punch cards. 124 00:06:54,100 --> 00:06:55,527 Maybe the high priest or priestess 125 00:06:55,527 --> 00:06:57,610 paid attention to you, and took the cards from you 126 00:06:57,610 --> 00:06:58,570 in a few minutes. 127 00:06:58,570 --> 00:07:00,040 Maybe that happened. 128 00:07:00,040 --> 00:07:03,430 And then they would tell you, fine. 129 00:07:03,430 --> 00:07:07,060 You're 479th on line, come back in a week 130 00:07:07,060 --> 00:07:09,100 and we'll have run this for you. 131 00:07:09,100 --> 00:07:12,520 This is not real-time computing. 132 00:07:12,520 --> 00:07:13,990 That's literally the way was like. 133 00:07:13,990 --> 00:07:15,490 It was just very forbidding. 134 00:07:15,490 --> 00:07:16,360 It was very distant. 135 00:07:18,970 --> 00:07:23,240 It's definitely not what we have now. 136 00:07:23,240 --> 00:07:26,050 And it's built for a number of processing bureaucracies, 137 00:07:26,050 --> 00:07:31,432 and initially in government, but then in business. 138 00:07:31,432 --> 00:07:32,890 And that's not to knock mainframes. 139 00:07:32,890 --> 00:07:37,600 They're a huge remarkable technical advance. 140 00:07:37,600 --> 00:07:41,110 IBM starts to dominate that sector, 141 00:07:41,110 --> 00:07:44,230 and they incorporate the machinery 142 00:07:44,230 --> 00:07:47,680 that they're familiar with. 143 00:07:47,680 --> 00:07:49,750 They're building business machines that 144 00:07:49,750 --> 00:07:53,120 are punch card-run with batch processing, 145 00:07:53,120 --> 00:07:59,050 so they carry those technologies into the mainframe computers 146 00:07:59,050 --> 00:08:00,190 that they're building. 147 00:08:00,190 --> 00:08:03,960 Now, look, you come back after a week, 148 00:08:03,960 --> 00:08:07,570 and you get handed a stack of this weird green and white 149 00:08:07,570 --> 00:08:09,288 paper with holes punched in the side. 150 00:08:09,288 --> 00:08:11,080 You probably don't have any memory of this, 151 00:08:11,080 --> 00:08:12,910 but I remember it deeply. 152 00:08:12,910 --> 00:08:15,088 And you'd get handed this stack of paper, 153 00:08:15,088 --> 00:08:16,630 and then you'd go back to your office 154 00:08:16,630 --> 00:08:19,510 and you sit down and start go through. 155 00:08:19,510 --> 00:08:23,680 And then you realize that, after page 39, 156 00:08:23,680 --> 00:08:26,680 it's all gibberish, because one of your punch cards 157 00:08:26,680 --> 00:08:30,010 was out of order or upside-down, or something like that. 158 00:08:30,010 --> 00:08:32,012 So then you'd have to go back and stand in line 159 00:08:32,012 --> 00:08:33,429 with the high priest or priestess, 160 00:08:33,429 --> 00:08:35,740 and start the whole process all over again. 161 00:08:35,740 --> 00:08:38,559 That's really what it's like. 162 00:08:38,559 --> 00:08:43,090 And the emphasis was on bigger and bigger computers 163 00:08:43,090 --> 00:08:50,450 able to handle more and more large number batch processing. 164 00:08:50,450 --> 00:08:55,960 So Licklider comes along more than a decade 165 00:08:55,960 --> 00:09:03,610 before personal computing and 30 years before the internet. 166 00:09:03,610 --> 00:09:06,400 And the word computer still has this kind 167 00:09:06,400 --> 00:09:11,170 of ominous tone of a intimidating device that's 168 00:09:11,170 --> 00:09:15,190 kind of hidden away in this over-lit, super air 169 00:09:15,190 --> 00:09:19,750 conditioned, noisy, fan-ridden space. 170 00:09:19,750 --> 00:09:25,570 And it's about as far away from this as we can imagine. 171 00:09:25,570 --> 00:09:30,550 And this is really what that team at DARPA, that Licklider 172 00:09:30,550 --> 00:09:33,250 led, drives us to. 173 00:09:33,250 --> 00:09:36,070 So we'll talk about the film in a bit, 174 00:09:36,070 --> 00:09:39,040 but a lot of the leadership of that team 175 00:09:39,040 --> 00:09:40,480 goes on to do amazing stuff. 176 00:09:40,480 --> 00:09:43,930 So here's some Licklider photographs. 177 00:09:43,930 --> 00:09:52,210 This is DEC's PDP-8 that was the first mini-computer, 178 00:09:52,210 --> 00:09:57,340 an entire movement away from mainframe computing 179 00:09:57,340 --> 00:09:59,800 that really was key to spawning the revolution. 180 00:09:59,800 --> 00:10:03,010 So Licklider is playing with this machine. 181 00:10:03,010 --> 00:10:05,170 It came from Digital Equipment Corporation, 182 00:10:05,170 --> 00:10:09,010 which is an MIT spin-off, Lincoln Labs spin-off. 183 00:10:09,010 --> 00:10:10,630 And they're trying to start to pursue 184 00:10:10,630 --> 00:10:13,060 a more personalized kind of computing 185 00:10:13,060 --> 00:10:16,330 with these mini-computers. 186 00:10:16,330 --> 00:10:23,260 On the lower left is the team that does the internet. 187 00:10:26,200 --> 00:10:30,400 And the only book that Licklider wrote 188 00:10:30,400 --> 00:10:33,520 is called libraries of the future, 189 00:10:33,520 --> 00:10:39,900 and he essentially envisions digital storage of everything. 190 00:10:39,900 --> 00:10:42,082 It's amazing work. 191 00:10:42,082 --> 00:10:43,540 The librarians are trying to wonder 192 00:10:43,540 --> 00:10:45,430 what to do with computers, and he essentially 193 00:10:45,430 --> 00:10:49,000 gives them the entire picture. 194 00:10:49,000 --> 00:10:51,430 So a very innovative character. 195 00:10:54,160 --> 00:10:55,900 He grows up in St. Louis. 196 00:10:55,900 --> 00:11:04,840 Both his parents are ministers, and he's an only child. 197 00:11:04,840 --> 00:11:08,740 Responsible kid, but his parents let 198 00:11:08,740 --> 00:11:14,740 him do things like assemble automobiles in the living room. 199 00:11:14,740 --> 00:11:17,920 For most of his life, he never spent more than $50 on a car 200 00:11:17,920 --> 00:11:21,760 because he would essentially by junkers and parts 201 00:11:21,760 --> 00:11:23,120 and put them together himself. 202 00:11:23,120 --> 00:11:25,330 That's the way the family drove around. 203 00:11:25,330 --> 00:11:28,630 So he has a tremendous facility for-- 204 00:11:28,630 --> 00:11:32,780 remember we talked about mind in hand, 205 00:11:32,780 --> 00:11:35,200 when we talked about the Industrial Revolution. 206 00:11:35,200 --> 00:11:39,730 Licklider has got that tactile knowledge facility, 207 00:11:39,730 --> 00:11:42,130 and it's part of why he loves computers. 208 00:11:42,130 --> 00:11:46,270 He falls in love with psychology and becomes actually 209 00:11:46,270 --> 00:11:49,420 quite a significant early thinker about psychology, 210 00:11:49,420 --> 00:11:51,930 when that whole field is starting to change. 211 00:11:51,930 --> 00:11:55,750 His World War II experience is spent up 212 00:11:55,750 --> 00:12:01,030 at Memorial Hall at Harvard in the basement 213 00:12:01,030 --> 00:12:05,860 with acoustics researchers, studying 214 00:12:05,860 --> 00:12:12,280 how pilots and the crew of airplanes-- how they react 215 00:12:12,280 --> 00:12:14,320 to this unbelievable noise that they've 216 00:12:14,320 --> 00:12:17,580 got in the pre-pressurized cabins in World War II aircraft 217 00:12:17,580 --> 00:12:20,770 with these huge propeller engines next to their ears, 218 00:12:20,770 --> 00:12:22,660 trying to hear radio signals. 219 00:12:22,660 --> 00:12:26,440 And this is before there's any kind of navigation system. 220 00:12:26,440 --> 00:12:27,520 It's pre-LORAN. 221 00:12:27,520 --> 00:12:29,320 It's obviously long before GPS. 222 00:12:29,320 --> 00:12:32,630 So there's a terrible problem-- the communication systems 223 00:12:32,630 --> 00:12:34,387 and the way in which you navigate a lot-- 224 00:12:34,387 --> 00:12:35,970 typically, if you can't see the stars. 225 00:12:35,970 --> 00:12:39,280 So all kinds of plans get lost over oceans 226 00:12:39,280 --> 00:12:43,720 because the radio communications don't get picked up. 227 00:12:43,720 --> 00:12:48,580 So he's working on how can we improve the man machine 228 00:12:48,580 --> 00:12:52,840 interface around acoustics in the cabin of airplanes? 229 00:12:52,840 --> 00:12:58,360 So he's thinking a lot about how people and machines work 230 00:12:58,360 --> 00:12:59,900 together. 231 00:12:59,900 --> 00:13:03,370 And that's what he contributes. 232 00:13:03,370 --> 00:13:05,740 He's not a computer-- early computer scientists really, 233 00:13:05,740 --> 00:13:06,640 although he loves computing. 234 00:13:06,640 --> 00:13:07,960 He plays with them all the time, knows them 235 00:13:07,960 --> 00:13:09,280 backwards and forwards. 236 00:13:09,280 --> 00:13:15,520 He's bringing the ideas of a psychologist to computing, 237 00:13:15,520 --> 00:13:17,260 and that's what, in many ways, makes 238 00:13:17,260 --> 00:13:19,690 him a complete breakthrough thinker. 239 00:13:19,690 --> 00:13:22,120 He's able to take ideas from psychology 240 00:13:22,120 --> 00:13:25,673 and bring them into this early computing field, not just 241 00:13:25,673 --> 00:13:27,340 an engineering perspective-- although he 242 00:13:27,340 --> 00:13:30,730 has very good engineering and physics capability. 243 00:13:30,730 --> 00:13:38,710 And he writes an absolutely seminal article in 1960 244 00:13:38,710 --> 00:13:41,440 that essentially was still working through his agenda 245 00:13:41,440 --> 00:13:43,480 of what needs to happen. 246 00:13:43,480 --> 00:13:48,970 It's an amazing vision. 247 00:13:48,970 --> 00:13:52,210 He, for example, realizes that 85% of human time 248 00:13:52,210 --> 00:13:55,750 is spent getting into a position to be 249 00:13:55,750 --> 00:13:58,550 able to think about a problem-- arranging all the stuff, 250 00:13:58,550 --> 00:13:59,967 pulling all the material together, 251 00:13:59,967 --> 00:14:01,460 and then you can think. 252 00:14:01,460 --> 00:14:04,550 Suppose you could reduce that 85% by having the computer 253 00:14:04,550 --> 00:14:06,470 do it for you, by having the computer arrange 254 00:14:06,470 --> 00:14:08,480 a lot of that stuff for you. 255 00:14:08,480 --> 00:14:10,490 That's part of his concept. 256 00:14:10,490 --> 00:14:13,010 The central concept he has is the hope 257 00:14:13,010 --> 00:14:15,590 is that human brains and computing 258 00:14:15,590 --> 00:14:18,350 machines will be coupled together very tightly, 259 00:14:18,350 --> 00:14:20,660 and that the resulting partnership will 260 00:14:20,660 --> 00:14:23,990 think as no human is ever thought and process 261 00:14:23,990 --> 00:14:26,690 data in a way not approached by the information handling 262 00:14:26,690 --> 00:14:29,270 machines we have today. 263 00:14:29,270 --> 00:14:30,860 In other words, he has this vision 264 00:14:30,860 --> 00:14:34,100 of people and machines in this collaborative relationship. 265 00:14:34,100 --> 00:14:37,167 It's going to be a symbiotic kind of relationship. 266 00:14:37,167 --> 00:14:38,750 And let's think for a minute about how 267 00:14:38,750 --> 00:14:40,940 people thought about computing. 268 00:14:40,940 --> 00:14:44,150 Now, we're all sitting around wondering when the robots are 269 00:14:44,150 --> 00:14:45,800 going to take our jobs. 270 00:14:45,800 --> 00:14:49,010 That's the obsession of the moment. 271 00:14:49,010 --> 00:14:50,690 Licklider's coming out of a time period 272 00:14:50,690 --> 00:14:52,580 where there was a similar obsession. 273 00:14:52,580 --> 00:14:55,190 And Norbert Wiener-- who pops up in this book, 274 00:14:55,190 --> 00:14:57,500 and is another one of MIT's absolute greats, 275 00:14:57,500 --> 00:15:00,200 a brilliant mathematician-- 276 00:15:00,200 --> 00:15:03,530 Wiener is the one who invents the term cybernetics. 277 00:15:03,530 --> 00:15:08,407 And Wiener has this vision saying, look out-- 278 00:15:08,407 --> 00:15:09,740 and he writes in the late '40s-- 279 00:15:09,740 --> 00:15:12,380 the computers are coming, and they're 280 00:15:12,380 --> 00:15:15,110 going to be smarter than we are. 281 00:15:15,110 --> 00:15:19,550 It's a very dark picture of what computing is going to be. 282 00:15:19,550 --> 00:15:22,760 And you all seen 2001-- the Hal computer 283 00:15:22,760 --> 00:15:25,310 in 2001, Space Odyssey? 284 00:15:25,310 --> 00:15:26,300 AUDIENCE: I know of it. 285 00:15:26,300 --> 00:15:27,758 WILLIAM BONVILLIAN: You know of it. 286 00:15:27,758 --> 00:15:29,480 Well, shows my age. 287 00:15:29,480 --> 00:15:33,290 But that's the kind of vision, this dark vision of computers 288 00:15:33,290 --> 00:15:35,390 taking over our lives. 289 00:15:35,390 --> 00:15:38,390 And that was alive and well as early computing 290 00:15:38,390 --> 00:15:39,570 started to take off. 291 00:15:39,570 --> 00:15:41,810 And actually one of the conceptualizers 292 00:15:41,810 --> 00:15:45,080 who got people worrying about computing was Norbert Wiener. 293 00:15:45,080 --> 00:15:48,500 But along comes Licklider with his background in psychology, 294 00:15:48,500 --> 00:15:52,130 and he has a completely different vision. 295 00:15:52,130 --> 00:15:58,580 And his vision wins, thank heavens. 296 00:15:58,580 --> 00:16:03,260 So he understands that things like human perception, 297 00:16:03,260 --> 00:16:08,840 and goal-setting, and judgment, and insight, and intuition-- 298 00:16:08,840 --> 00:16:11,240 machines can't do that, that these are 299 00:16:11,240 --> 00:16:13,735 incredible human attributes. 300 00:16:13,735 --> 00:16:15,110 And the machine's just not can be 301 00:16:15,110 --> 00:16:18,590 able to equal us on that stuff, that certain kind of feel 302 00:16:18,590 --> 00:16:23,460 we have about how things fall together and shape up in place. 303 00:16:23,460 --> 00:16:25,820 So those qualities are going to surpass machines, 304 00:16:25,820 --> 00:16:28,730 but then machines are going to be able to do stuff 305 00:16:28,730 --> 00:16:31,220 that we can't do, that we're not particularly good at, 306 00:16:31,220 --> 00:16:33,480 and don't want to spend a lot of time on. 307 00:16:33,480 --> 00:16:35,510 So there's going to be a relationship. 308 00:16:35,510 --> 00:16:39,740 So Licklider is spending a lot of time at Lincoln Lab, 309 00:16:39,740 --> 00:16:43,940 playing on computers up at Lincoln Lab. 310 00:16:43,940 --> 00:16:46,370 He starts to master all those early machines, 311 00:16:46,370 --> 00:16:53,240 including the DEC PDP-8, the first mini-computer-- 312 00:16:53,240 --> 00:16:56,180 real mini-computer. 313 00:16:56,180 --> 00:17:00,830 He leads the MIT Psychology Department, 314 00:17:00,830 --> 00:17:08,119 and joins BBN, which is the premier acoustics consulting 315 00:17:08,119 --> 00:17:09,790 firm. 316 00:17:09,790 --> 00:17:13,520 And they do things like design the acoustics 317 00:17:13,520 --> 00:17:16,910 for Lincoln Center, the famous acoustics firm. 318 00:17:16,910 --> 00:17:19,160 And they had this idea, gee, maybe 319 00:17:19,160 --> 00:17:21,770 we ought to get in the computing business. 320 00:17:21,770 --> 00:17:24,710 Maybe these computers are actually here to stay. 321 00:17:24,710 --> 00:17:29,630 So Licklider is hired at BBN to bring them 322 00:17:29,630 --> 00:17:33,055 into the early computing age. 323 00:17:33,055 --> 00:17:34,055 And that's what he does. 324 00:17:37,950 --> 00:17:41,570 In this 1960 article, his key image 325 00:17:41,570 --> 00:17:45,170 is a fig tree pollinated by a wasp-- 326 00:17:45,170 --> 00:17:47,390 in other words, the relationship between the fig tree 327 00:17:47,390 --> 00:17:49,400 and the wasp is complete symbiotic, 328 00:17:49,400 --> 00:17:51,860 they're mutually supporting. 329 00:17:51,860 --> 00:17:53,510 And that's the vision that he has 330 00:17:53,510 --> 00:17:55,490 of what computing is going to be. 331 00:17:55,490 --> 00:18:00,350 So it's this partnership between computers and people 332 00:18:00,350 --> 00:18:03,440 with a complementary set of functions. 333 00:18:03,440 --> 00:18:05,960 And the partnership has to be in real time. 334 00:18:05,960 --> 00:18:08,270 We have to get rid of all this batch processing stuff. 335 00:18:08,270 --> 00:18:10,250 That has to go. 336 00:18:10,250 --> 00:18:13,190 It's got to be real time. 337 00:18:13,190 --> 00:18:15,020 The computer is to relieve people 338 00:18:15,020 --> 00:18:18,470 of that drudgery factor, the 85% of the set up time 339 00:18:18,470 --> 00:18:21,470 we spend getting ready to think. 340 00:18:21,470 --> 00:18:26,600 It is a time of what is being talked about as time-sharing. 341 00:18:26,600 --> 00:18:29,930 These big massive mainframes behind glass with the priests 342 00:18:29,930 --> 00:18:32,900 and priestesses dressed in white lab coats, 343 00:18:32,900 --> 00:18:35,780 people don't get access to them. 344 00:18:35,780 --> 00:18:39,290 It's like going to the Oracle of Delphi. 345 00:18:39,290 --> 00:18:40,640 You just can't get in there. 346 00:18:40,640 --> 00:18:44,010 And he has a vision of could we, in effect, 347 00:18:44,010 --> 00:18:46,820 take that computer back from the Oracle of Delphi 348 00:18:46,820 --> 00:18:48,380 and get people on this thing. 349 00:18:48,380 --> 00:18:49,790 And how are we going to do that with mainframes? 350 00:18:49,790 --> 00:18:52,123 We're going to do it through what's called, at the time, 351 00:18:52,123 --> 00:18:52,790 time-sharing. 352 00:18:52,790 --> 00:18:56,420 Let people have access to a share of time. 353 00:18:56,420 --> 00:19:01,000 And he has a vision of central computers, and then-- 354 00:19:01,000 --> 00:19:05,080 as a competing center with interconnected users 355 00:19:05,080 --> 00:19:07,300 tying into that central computing system. 356 00:19:10,630 --> 00:19:14,920 He understands things like building 357 00:19:14,920 --> 00:19:21,170 a network of people aided by computers 358 00:19:21,170 --> 00:19:25,370 that become in effect a network of thinking centers. 359 00:19:25,370 --> 00:19:29,660 And that network concept really is a core idea 360 00:19:29,660 --> 00:19:33,460 that is what his team at DARPA later establishes 361 00:19:33,460 --> 00:19:35,750 BBN as the internet. 362 00:19:35,750 --> 00:19:37,980 He understands that artificial intelligence 363 00:19:37,980 --> 00:19:42,320 isn't an analogous way of thinking about thinking-- 364 00:19:42,320 --> 00:19:46,640 that it can help us understand intellectual processes. 365 00:19:50,480 --> 00:19:54,320 He has a vision of a computer being a desk console, not 366 00:19:54,320 --> 00:19:55,400 a mainframe-- 367 00:19:55,400 --> 00:19:56,978 in other words, radically smaller 368 00:19:56,978 --> 00:19:58,520 than even the mini-computer that he's 369 00:19:58,520 --> 00:20:01,370 starting to play with from DEC. 370 00:20:04,130 --> 00:20:10,160 He just has a series of these concepts in this 1960 article 371 00:20:10,160 --> 00:20:13,910 that lays out what's going to be happening. 372 00:20:13,910 --> 00:20:17,150 He envisions interactive computing. 373 00:20:17,150 --> 00:20:20,900 He envisions programming is not just programming. 374 00:20:20,900 --> 00:20:24,590 He starts to understand that this is a science-- 375 00:20:24,590 --> 00:20:27,410 in a way, a linguistic kind of science in itself, 376 00:20:27,410 --> 00:20:30,320 and is deserving full respect-- 377 00:20:30,320 --> 00:20:34,370 that computer science is a true science like other sciences, 378 00:20:34,370 --> 00:20:37,100 that it's not just a new tool set. 379 00:20:37,100 --> 00:20:40,190 It's something organized around one of the fundamentals. 380 00:20:40,190 --> 00:20:43,467 Information is one of the fundamentals, 381 00:20:43,467 --> 00:20:45,800 one of the foundational building blocks of the universe. 382 00:20:45,800 --> 00:20:48,230 And you're organizing a whole field around information. 383 00:20:48,230 --> 00:20:50,030 Computing is going to be a science, 384 00:20:50,030 --> 00:20:53,420 and he recognizes that very, very early on, 385 00:20:53,420 --> 00:20:55,220 long before others. 386 00:20:55,220 --> 00:20:56,930 It'll have practical applications, 387 00:20:56,930 --> 00:21:03,920 but it's worthy of a whole intellectual field on its own. 388 00:21:03,920 --> 00:21:08,630 And he sees how closely tied computing 389 00:21:08,630 --> 00:21:09,922 can be to cognitive science. 390 00:21:09,922 --> 00:21:12,380 So he has a whole vision of what he calls dynamic modeling. 391 00:21:15,130 --> 00:21:17,100 In other words, this is an era where 392 00:21:17,100 --> 00:21:20,010 you get statistics on paper. 393 00:21:20,010 --> 00:21:23,220 But he has the vision that, if you put them on your computer, 394 00:21:23,220 --> 00:21:26,910 you can start to play with each element 395 00:21:26,910 --> 00:21:32,040 and begin to turn them into graphs and designs, 396 00:21:32,040 --> 00:21:36,540 and get a much more realistic feel for how the data is 397 00:21:36,540 --> 00:21:40,260 related, and start to play with the different variables. 398 00:21:40,260 --> 00:21:44,070 And you're going to be able to feed hundreds and thousands 399 00:21:44,070 --> 00:21:46,470 of variables into your computer in a way 400 00:21:46,470 --> 00:21:49,200 that you can never even think of, 401 00:21:49,200 --> 00:21:53,830 when people are doing statistics by pencil and paper. 402 00:21:53,830 --> 00:21:56,910 So he has a whole vision of this whole-- what 403 00:21:56,910 --> 00:21:58,830 he calls dynamic modeling, the modeling 404 00:21:58,830 --> 00:22:03,750 role that computing effect plays in a very dramatic way now. 405 00:22:03,750 --> 00:22:09,960 It's that computing can handle this problem of complexity 406 00:22:09,960 --> 00:22:12,210 because of its relentless ability 407 00:22:12,210 --> 00:22:16,170 to deal with thousands and now millions of variables. 408 00:22:16,170 --> 00:22:17,700 So economics is a profession that 409 00:22:17,700 --> 00:22:20,190 was organized around essentially you couldn't really 410 00:22:20,190 --> 00:22:23,280 handle much more than three, four, five variables 411 00:22:23,280 --> 00:22:25,140 operating at the same time. 412 00:22:25,140 --> 00:22:27,870 Computing would enable science to cope 413 00:22:27,870 --> 00:22:32,340 with a profound issue like climate change, 414 00:22:32,340 --> 00:22:35,010 and actually start to be able to model that. 415 00:22:35,010 --> 00:22:36,750 So he understands all this stuff, 416 00:22:36,750 --> 00:22:40,740 and he starts to lay this out piece after piece after piece 417 00:22:40,740 --> 00:22:42,280 for people to think about. 418 00:22:45,300 --> 00:22:48,510 I mentioned earlier the electronic library. 419 00:22:48,510 --> 00:22:50,490 He understands that all information 420 00:22:50,490 --> 00:22:52,217 is going to be online and stored online, 421 00:22:52,217 --> 00:22:54,300 and is going to be accessible to everybody who can 422 00:22:54,300 --> 00:22:58,170 get onto one of these things. 423 00:22:58,170 --> 00:23:00,480 And it will be the library and it 424 00:23:00,480 --> 00:23:03,328 will be the reference librarian for you. 425 00:23:03,328 --> 00:23:05,370 In other words, it's going to be able to organize 426 00:23:05,370 --> 00:23:08,220 your information in ways that could never have 427 00:23:08,220 --> 00:23:09,960 been previously anticipated. 428 00:23:14,820 --> 00:23:19,950 15 years before the PC and 30 years before the internet, 429 00:23:19,950 --> 00:23:24,510 he's seeing graphics, computing graphics, computer 430 00:23:24,510 --> 00:23:29,220 rich personal workstations, human computer symbiosis, 431 00:23:29,220 --> 00:23:32,220 interactive computing, networks, the idea 432 00:23:32,220 --> 00:23:34,920 of an online community, a thinking community, 433 00:23:34,920 --> 00:23:40,170 online libraries, instant retrieval, data organization, 434 00:23:40,170 --> 00:23:43,800 computing language, programming that's 435 00:23:43,800 --> 00:23:51,330 elevated from a tool set to a true science-- 436 00:23:51,330 --> 00:23:53,260 a whole digital medium of expression. 437 00:23:53,260 --> 00:23:54,970 So he is a visionary. 438 00:23:54,970 --> 00:23:55,950 He has this vision. 439 00:23:58,740 --> 00:24:03,180 He starts writing about the intergalactic network. 440 00:24:03,180 --> 00:24:08,940 It's the worldwide web decades before Tim Berners-Lee here 441 00:24:08,940 --> 00:24:10,770 pulls it off. 442 00:24:10,770 --> 00:24:17,500 He's a visionary, and of a remarkable degree. 443 00:24:17,500 --> 00:24:22,950 And there's no vision champion at the time, 444 00:24:22,950 --> 00:24:27,990 but then DARPA hires him out of BBN 445 00:24:27,990 --> 00:24:31,740 to lead essentially this whole development 446 00:24:31,740 --> 00:24:33,330 of personal computing. 447 00:24:33,330 --> 00:24:39,240 So it's a completely rare moment where a visionary becomes 448 00:24:39,240 --> 00:24:40,290 the vision enabler. 449 00:24:43,950 --> 00:24:45,450 What happens here? 450 00:24:45,450 --> 00:24:47,940 So 1957 is the Shock of Sputnik. 451 00:24:47,940 --> 00:24:50,340 So DARPA gets created. 452 00:24:50,340 --> 00:24:54,110 DARPA's originally given the space mission, because. 453 00:24:54,110 --> 00:24:55,950 Eisenhower's angry at the services 454 00:24:55,950 --> 00:24:59,010 because they weren't talking to each other, 455 00:24:59,010 --> 00:25:02,460 and Russians got the first satellite up. 456 00:25:02,460 --> 00:25:04,380 And it wasn't about a nice blinking thing, 457 00:25:04,380 --> 00:25:08,002 it was about first strike ballistic missile capability. 458 00:25:08,002 --> 00:25:08,835 It was very serious. 459 00:25:12,600 --> 00:25:15,570 But then he took space away and gave it to NASA, 460 00:25:15,570 --> 00:25:17,910 so DARPA had to start thinking of things to do. 461 00:25:17,910 --> 00:25:20,790 And they come up with a lot of creative stuff, 462 00:25:20,790 --> 00:25:25,050 but the big early DARPA set of accomplishments 463 00:25:25,050 --> 00:25:28,770 is really Licklider and his team. 464 00:25:28,770 --> 00:25:35,040 And so Jack Kennedy and Secretary 465 00:25:35,040 --> 00:25:38,410 of Defense Robert McNamara go through the Cuban Missile 466 00:25:38,410 --> 00:25:42,600 Crisis, which is a complete nightmare. 467 00:25:42,600 --> 00:25:46,230 And we do things like get within 20 minutes of launching 468 00:25:46,230 --> 00:25:48,818 a first strike by mistake. 469 00:25:48,818 --> 00:25:50,610 And they realize that they've got something 470 00:25:50,610 --> 00:25:54,780 that they call a really serious command and control problem. 471 00:25:54,780 --> 00:25:57,720 And Jack Ruina, who's head of DARPA, who's another MIT 472 00:25:57,720 --> 00:26:03,948 faculty member, hires Licklider to come in and deal 473 00:26:03,948 --> 00:26:05,490 with the command and control problem. 474 00:26:05,490 --> 00:26:09,030 And Licklider's response to Jack Ruina is, wait a minute, 475 00:26:09,030 --> 00:26:15,270 I'm working on computing and making computing personal. 476 00:26:15,270 --> 00:26:18,150 What do I know about the command and control problem? 477 00:26:18,150 --> 00:26:20,940 But they have a conversation, and Ruina 478 00:26:20,940 --> 00:26:23,640 starts to see that what Licklider is working on 479 00:26:23,640 --> 00:26:26,650 is exactly what he is concerned about the command and control 480 00:26:26,650 --> 00:26:27,150 problem. 481 00:26:27,150 --> 00:26:30,540 With all these capabilities that Licklider is thinking about, 482 00:26:30,540 --> 00:26:33,510 they need to tackle this command and control problem. 483 00:26:33,510 --> 00:26:36,930 So Licklider comes to DARPA, and because he's 484 00:26:36,930 --> 00:26:39,540 working on what DARPA calls a presidential-- 485 00:26:39,540 --> 00:26:42,870 in other words, this is the problem the president himself-- 486 00:26:42,870 --> 00:26:45,090 President Kennedy himself has posed-- 487 00:26:45,090 --> 00:26:47,700 fix the command and control problem. 488 00:26:47,700 --> 00:26:50,130 He has resources. 489 00:26:50,130 --> 00:26:53,400 So it's a very powerful moment. 490 00:26:53,400 --> 00:26:55,440 Ruina hadn't really thought about the linkage 491 00:26:55,440 --> 00:26:57,600 between behavior and computing and the kind of way 492 00:26:57,600 --> 00:27:00,240 that Licklider has really thought through, 493 00:27:00,240 --> 00:27:02,370 but begins to see what the possibilities are. 494 00:27:02,370 --> 00:27:06,330 So this is really when DARPA becomes DARPA. 495 00:27:09,540 --> 00:27:17,307 Some of the culture of DARPA-- we've talked a bit about they-- 496 00:27:17,307 --> 00:27:19,140 I'll just summarize some of the rules sets-- 497 00:27:28,600 --> 00:27:32,870 hire great program managers that are world-class, 498 00:27:32,870 --> 00:27:34,880 give them an enormous amount of power. 499 00:27:34,880 --> 00:27:37,430 They have the authority at DARPA. 500 00:27:37,430 --> 00:27:40,340 Unleash them. 501 00:27:40,340 --> 00:27:43,400 Let them develop a vision, and then find 502 00:27:43,400 --> 00:27:46,250 the best people in the country to do it, 503 00:27:46,250 --> 00:27:48,500 and they're on the problem. 504 00:27:48,500 --> 00:27:50,590 That is not the way R&D organizations are 505 00:27:50,590 --> 00:27:52,370 organized in the United States. 506 00:27:52,370 --> 00:27:55,220 That's not the way we do it. 507 00:27:55,220 --> 00:27:56,960 We have NSF. 508 00:27:56,960 --> 00:27:59,930 It has a program manager. 509 00:27:59,930 --> 00:28:04,820 The program manager is overseen by a peer review system. 510 00:28:04,820 --> 00:28:07,670 They wait for applications to come in from the scientists 511 00:28:07,670 --> 00:28:08,960 saying what they'd like to do. 512 00:28:08,960 --> 00:28:10,400 They review the applications. 513 00:28:10,400 --> 00:28:14,000 The peer review team makes a decision on who gets funded. 514 00:28:14,000 --> 00:28:15,830 The project manager is essentially 515 00:28:15,830 --> 00:28:18,530 doing the arranging here, but peer review 516 00:28:18,530 --> 00:28:19,880 makes the core decision. 517 00:28:19,880 --> 00:28:21,830 It's a much more bureaucratic process, 518 00:28:21,830 --> 00:28:23,490 and it's a bottom-up system. 519 00:28:23,490 --> 00:28:25,490 Look, there's a lot of strengths in that system. 520 00:28:25,490 --> 00:28:27,380 I don't want to knock that system. 521 00:28:27,380 --> 00:28:29,600 It's an important function to have for science 522 00:28:29,600 --> 00:28:32,480 so that scientists themselves can originate ideas. 523 00:28:32,480 --> 00:28:33,880 That's not what DARPA does. 524 00:28:33,880 --> 00:28:35,540 DARPA has strong program managers. 525 00:28:35,540 --> 00:28:40,610 They develop the ideas, and they go out and make them happen. 526 00:28:40,610 --> 00:28:43,670 And program managers don't last much more than four years 527 00:28:43,670 --> 00:28:44,200 at DARPA-- 528 00:28:44,200 --> 00:28:46,530 three to five years-- 529 00:28:46,530 --> 00:28:49,520 so they've got that period of time 530 00:28:49,520 --> 00:28:53,720 to get done with their project. 531 00:28:53,720 --> 00:28:56,150 They've got to get something happening within a relatively 532 00:28:56,150 --> 00:28:57,060 short period of time. 533 00:28:57,060 --> 00:28:59,330 So that's going to affect what they try to do. 534 00:28:59,330 --> 00:29:02,120 They're not working on 25-year projects at DARPA. 535 00:29:02,120 --> 00:29:04,670 They're working on stuff that can be stood up in a relatively 536 00:29:04,670 --> 00:29:06,410 short period of time. 537 00:29:06,410 --> 00:29:12,110 And it's just a very different organizational model 538 00:29:12,110 --> 00:29:14,060 for a science organization-- 539 00:29:14,060 --> 00:29:17,633 no peer review, two sign offs. 540 00:29:17,633 --> 00:29:20,300 If you're a program manager, you figure out what you want to do, 541 00:29:20,300 --> 00:29:21,470 what direction you want to go to. 542 00:29:21,470 --> 00:29:23,887 You sell that to your office director and to the director, 543 00:29:23,887 --> 00:29:24,530 and that's it. 544 00:29:24,530 --> 00:29:27,710 And it could be funded that day. 545 00:29:27,710 --> 00:29:31,520 Can move very, very fast. 546 00:29:31,520 --> 00:29:35,990 So they're organized around this whole grand challenge concept. 547 00:29:35,990 --> 00:29:38,090 DARPA doesn't want to do incremental advance. 548 00:29:38,090 --> 00:29:41,210 It only wants to do technology revolutionary breakthroughs 549 00:29:41,210 --> 00:29:43,940 that will change everything. 550 00:29:43,940 --> 00:29:45,160 So it doesn't want to waste-- 551 00:29:45,160 --> 00:29:46,910 let other people do the incremental stuff. 552 00:29:46,910 --> 00:29:48,800 That's not DARPA's job. 553 00:29:48,800 --> 00:29:50,332 And it's a right left model, so it 554 00:29:50,332 --> 00:29:52,040 wants to look at what's going to come out 555 00:29:52,040 --> 00:29:56,270 of the pipeline, what the technology is that it wants-- 556 00:29:56,270 --> 00:30:00,950 personal computing-- and then go back into the pipeline 557 00:30:00,950 --> 00:30:03,830 and figure out the research that's going to get this 558 00:30:03,830 --> 00:30:07,210 outcome, again, within the life of the program manager-- 559 00:30:07,210 --> 00:30:09,990 in this, say, four or five-year period. 560 00:30:09,990 --> 00:30:16,270 So it's just it's really different. 561 00:30:16,270 --> 00:30:20,300 Licklider pulls together different teams. 562 00:30:20,300 --> 00:30:24,290 So the other thing about DARPA is 563 00:30:24,290 --> 00:30:26,390 that essentially the role of the program manager 564 00:30:26,390 --> 00:30:29,690 is to find great groups and to support them. 565 00:30:29,690 --> 00:30:31,730 And that's what Licklider does. 566 00:30:31,730 --> 00:30:33,470 So DARPA's different because it's 567 00:30:33,470 --> 00:30:37,670 working as an institutional entity in the R&D space. 568 00:30:37,670 --> 00:30:41,610 It's another institutional R&D entity, 569 00:30:41,610 --> 00:30:47,420 but it is also organized around personal face-to-face great 570 00:30:47,420 --> 00:30:49,340 group creation. 571 00:30:49,340 --> 00:30:51,800 So it takes the two sides of the equation 572 00:30:51,800 --> 00:30:54,740 that we treated as two entirely different classes-- 573 00:30:54,740 --> 00:30:56,270 it is putting them together. 574 00:30:56,270 --> 00:30:58,640 That's why DARPA follows after our discussion 575 00:30:58,640 --> 00:30:59,930 in the great group class. 576 00:30:59,930 --> 00:31:02,540 That's who Licklider is finding. 577 00:31:02,540 --> 00:31:05,000 He's finding these amazing, great groups. 578 00:31:05,000 --> 00:31:07,400 And we've seen some of them before. 579 00:31:07,400 --> 00:31:13,040 These are all who's who of the history of computer science. 580 00:31:13,040 --> 00:31:16,610 We talked about Doug Engelbart. 581 00:31:16,610 --> 00:31:18,040 There he is. 582 00:31:18,040 --> 00:31:22,520 He's someone that Lick finds and supports, 583 00:31:22,520 --> 00:31:24,710 even though the Stanford Research Institute 584 00:31:24,710 --> 00:31:26,040 doesn't know what he's doing. 585 00:31:26,040 --> 00:31:27,623 They don't know what he's doing there. 586 00:31:30,380 --> 00:31:33,440 He does this you know unbelievable demonstration 587 00:31:33,440 --> 00:31:36,620 with all this breakthrough material. 588 00:31:36,620 --> 00:31:40,350 At MIT, project MAC is right there at Technology Square. 589 00:31:40,350 --> 00:31:44,190 You can kind of recognize that building on the lower left. 590 00:31:44,190 --> 00:31:49,130 So after Licklider leaves DARPA, he goes to join Project MAC. 591 00:31:49,130 --> 00:31:53,780 But meanwhile, he gets stood up by Ferdinand Corbato and Bob 592 00:31:53,780 --> 00:31:55,863 Fano. 593 00:31:55,863 --> 00:31:58,280 The history of what comes out of that place-- and it's all 594 00:31:58,280 --> 00:32:01,400 DARPA-funded-- is just breathtaking. 595 00:32:01,400 --> 00:32:05,780 So when you go buy that building at Technology Square-- 596 00:32:05,780 --> 00:32:08,060 I think it's 500 Technology Square-- 597 00:32:08,060 --> 00:32:11,900 take off your hat, because amazing stuff. 598 00:32:11,900 --> 00:32:14,450 Absolutely amazing stuff occurred there. 599 00:32:14,450 --> 00:32:18,390 Within MIT, there's this fascinating conflict. 600 00:32:18,390 --> 00:32:22,580 There are two competing visions of computer science 601 00:32:22,580 --> 00:32:25,910 fighting for the same DARPA money. 602 00:32:25,910 --> 00:32:29,990 So Marvin Minsky, great roboticist-- 603 00:32:29,990 --> 00:32:32,380 great, one of the all-time greats-- just died recently. 604 00:32:32,380 --> 00:32:35,630 Amazing figure. 605 00:32:35,630 --> 00:32:38,120 Minsky had this vision that computer thinking would be 606 00:32:38,120 --> 00:32:40,370 superior to people's thinking. 607 00:32:40,370 --> 00:32:44,420 So he has the Artificial Intelligence Laboratory 608 00:32:44,420 --> 00:32:47,180 and he's running that vision, and then 609 00:32:47,180 --> 00:32:49,820 Licklider and Corbato and Fano are 610 00:32:49,820 --> 00:32:52,240 running the computer is our friend, 611 00:32:52,240 --> 00:32:53,990 it's going to be a symbiotic relationship, 612 00:32:53,990 --> 00:32:56,540 we're going to do personal computing-- 613 00:32:56,540 --> 00:33:00,650 very different ideas battling each other for funding. 614 00:33:00,650 --> 00:33:05,000 It took years of negotiation to put together CSAIL 615 00:33:05,000 --> 00:33:08,810 because you had these two different visions that 616 00:33:08,810 --> 00:33:11,010 competed with each other. 617 00:33:11,010 --> 00:33:16,850 Just like we talked about the competition over the genome 618 00:33:16,850 --> 00:33:19,040 last week in the great groups session, 619 00:33:19,040 --> 00:33:21,020 this was one of those great competitions. 620 00:33:21,020 --> 00:33:24,130 Each side is pushing each other. 621 00:33:24,130 --> 00:33:28,040 And Minsky and team would do unbelievable robotics stuff 622 00:33:28,040 --> 00:33:31,680 and they would say, top this. 623 00:33:31,680 --> 00:33:34,680 And the MAC team would come back with the responses. 624 00:33:34,680 --> 00:33:37,470 So it was a wonderful competition going on here. 625 00:33:37,470 --> 00:33:42,390 But creating CSAIL required two co-directors at the outset-- 626 00:33:42,390 --> 00:33:44,400 somebody from computer science and somebody 627 00:33:44,400 --> 00:33:46,977 from artificial intelligence. 628 00:33:46,977 --> 00:33:48,060 It's kind of settled down. 629 00:33:48,060 --> 00:33:49,560 They're all on the same project now. 630 00:33:49,560 --> 00:33:54,720 But that was a 20-year or longer project to get there. 631 00:33:54,720 --> 00:34:01,500 So it's a remarkable set of communities, 632 00:34:01,500 --> 00:34:05,310 including project MAC at MIT, but that's just one of them 633 00:34:05,310 --> 00:34:06,160 that gets created. 634 00:34:06,160 --> 00:34:08,070 So all the early computer science departments 635 00:34:08,070 --> 00:34:09,330 are funded by DARPA. 636 00:34:09,330 --> 00:34:10,679 DARPA really creates them. 637 00:34:10,679 --> 00:34:12,659 That's why there is computer science that's 638 00:34:12,659 --> 00:34:14,242 so deep and rich in the United States. 639 00:34:18,630 --> 00:34:25,489 Licklider's three research priorities at DARPA are 640 00:34:25,489 --> 00:34:29,600 time-sharing-- which is really interactive computing, 641 00:34:29,600 --> 00:34:31,370 and that's like personal computing-- 642 00:34:31,370 --> 00:34:33,800 and then graphics and modeling-- 643 00:34:33,800 --> 00:34:35,540 and another great name, Ivan Sullivan 644 00:34:35,540 --> 00:34:39,139 ran that piece, and that's early computing graphics-- 645 00:34:39,139 --> 00:34:41,420 and then networking. 646 00:34:41,420 --> 00:34:43,730 And Bob Taylor, who we talked about last week, 647 00:34:43,730 --> 00:34:47,870 who goes to Xerox PARC, comes out of DARPA to do that. 648 00:34:47,870 --> 00:34:49,940 And that's where he gets his stuff. 649 00:34:49,940 --> 00:34:52,429 And he and Larry Roberts, as well, 650 00:34:52,429 --> 00:34:56,900 are the ones that lead the effort to stand up computing. 651 00:34:56,900 --> 00:34:59,480 And Luyao told us about him last week. 652 00:34:59,480 --> 00:35:02,240 We sent around his obit, because he just died this past week. 653 00:35:02,240 --> 00:35:03,410 It's a wonderful obit-- 654 00:35:03,410 --> 00:35:04,910 if you all get a chance to read it-- 655 00:35:04,910 --> 00:35:07,790 that appeared in New York Times. 656 00:35:07,790 --> 00:35:11,730 The ARPANET gets created. 657 00:35:11,730 --> 00:35:14,270 So Licklider creates this thing at DARPA. 658 00:35:14,270 --> 00:35:16,430 He has two different stints at DARPA. 659 00:35:16,430 --> 00:35:18,890 He's there for several years in each stint. 660 00:35:18,890 --> 00:35:21,948 But he creates a community that keeps carrying on the projects. 661 00:35:21,948 --> 00:35:23,990 In other words, they agree on what the agenda is, 662 00:35:23,990 --> 00:35:26,540 and they keep working down the list together. 663 00:35:26,540 --> 00:35:30,110 So DARPA doesn't-- isn't always able to do this. 664 00:35:30,110 --> 00:35:31,460 Often there's a real problem-- 665 00:35:31,460 --> 00:35:33,470 when a program manager leaves after their three 666 00:35:33,470 --> 00:35:36,950 to five-year stint, what happens to their work? 667 00:35:36,950 --> 00:35:38,360 Who does the follow one? 668 00:35:38,360 --> 00:35:41,180 But Licklider builds this community 669 00:35:41,180 --> 00:35:47,270 that keeps the project going for way over a decade and a half. 670 00:35:47,270 --> 00:35:54,350 And Taylor and Larry Roberts, out of Lincoln Lab, 671 00:35:54,350 --> 00:35:57,350 really lead the effort to do the internet. 672 00:35:57,350 --> 00:36:05,160 They hire BBN, Licklider's old firm, to actually carry it out. 673 00:36:05,160 --> 00:36:08,210 And that's done right here in Cambridge. 674 00:36:08,210 --> 00:36:11,240 So the internet gets done just up the road. 675 00:36:16,740 --> 00:36:20,420 I won't go into the whole story about it, but it's-- 676 00:36:24,920 --> 00:36:28,010 two great things occur in 1969-- 677 00:36:28,010 --> 00:36:31,670 we go to the Moon and we do the internet. 678 00:36:31,670 --> 00:36:37,430 And in 1969, nobody had ever heard of the internet. 679 00:36:37,430 --> 00:36:42,830 But in the end, which is the more significant? 680 00:36:42,830 --> 00:36:47,780 And that's going on just up the street here. 681 00:36:47,780 --> 00:36:49,250 It's a pretty breathtaking moment. 682 00:36:49,250 --> 00:36:51,290 These are some of the figures. 683 00:36:51,290 --> 00:36:57,110 This is Vint Cerf and Bob Kahn, who were-- 684 00:36:57,110 --> 00:36:59,580 lead the whole protocol effort-- 685 00:36:59,580 --> 00:37:02,092 TCPIP, the whole internet protocol system 686 00:37:02,092 --> 00:37:04,175 that's really at the heart of the internet system. 687 00:37:08,000 --> 00:37:10,670 Larry Roberts, who came out of Lincoln Lab, 688 00:37:10,670 --> 00:37:18,050 persuades the DARPA director to use email. 689 00:37:18,050 --> 00:37:20,290 So DARPA's like an early email user. 690 00:37:20,290 --> 00:37:22,340 So everybody in DARPA goes on email. 691 00:37:22,340 --> 00:37:25,430 They nurture this thing along. 692 00:37:25,430 --> 00:37:29,990 They're one of the early user bases. 693 00:37:29,990 --> 00:37:31,830 It becomes this system of communication, 694 00:37:31,830 --> 00:37:36,080 because DARPA's using it for the entire ARPANET community. 695 00:37:36,080 --> 00:37:45,110 It's originally stood up with about five universities linked. 696 00:37:45,110 --> 00:37:49,480 The real problem they've got is that, because DARPA's 697 00:37:49,480 --> 00:37:51,980 trying to stand up these early computer science departments, 698 00:37:51,980 --> 00:37:54,830 and lots of breakthroughs are happening in computing, 699 00:37:54,830 --> 00:37:56,690 DARPA's having to spend a fortune buying 700 00:37:56,690 --> 00:38:00,710 new computers for each one of its university research 701 00:38:00,710 --> 00:38:01,220 centers. 702 00:38:01,220 --> 00:38:02,553 And it's costing them a fortune. 703 00:38:02,553 --> 00:38:06,050 They're spending their money on the next generation 704 00:38:06,050 --> 00:38:07,910 of computer advance. 705 00:38:07,910 --> 00:38:12,140 And one of the reasons they have to do the internet 706 00:38:12,140 --> 00:38:15,020 is that, gee, if they only had to buy one computer 707 00:38:15,020 --> 00:38:17,698 and everybody could get on it, it 708 00:38:17,698 --> 00:38:19,490 would be a gigantic money saver, and they'd 709 00:38:19,490 --> 00:38:21,140 be able to do much more stuff. 710 00:38:21,140 --> 00:38:23,572 So they have this internal reason. 711 00:38:23,572 --> 00:38:25,280 Now, the external reason for the internet 712 00:38:25,280 --> 00:38:27,560 is much more complicated. 713 00:38:27,560 --> 00:38:29,260 One of the reasons is that it's going 714 00:38:29,260 --> 00:38:32,840 to be a secure communications system at the time 715 00:38:32,840 --> 00:38:33,890 of terrible threats. 716 00:38:33,890 --> 00:38:37,478 Now, you all are-- 717 00:38:37,478 --> 00:38:39,020 probably don't have much recollection 718 00:38:39,020 --> 00:38:41,720 of 9/11, but sure enough, the only way 719 00:38:41,720 --> 00:38:46,080 we knew about each other on 9/11 was because the internet still 720 00:38:46,080 --> 00:38:46,580 worked. 721 00:38:46,580 --> 00:38:50,990 The phone system in New York City went down in a flash. 722 00:38:50,990 --> 00:38:52,490 And it worked. 723 00:38:52,490 --> 00:38:55,700 It turned out to be an enormously important 724 00:38:55,700 --> 00:38:59,630 fallback communications system, because it is so resilient. 725 00:38:59,630 --> 00:39:02,810 In other word, you can establish those network lines essentially 726 00:39:02,810 --> 00:39:07,610 through any connection, not just a straight one-to-one 727 00:39:07,610 --> 00:39:09,750 connection like the phone system. 728 00:39:09,750 --> 00:39:13,670 So that was part of the rationale for developing 729 00:39:13,670 --> 00:39:17,950 the internet, but frankly, Licklider 730 00:39:17,950 --> 00:39:20,577 had a vision of what the internet was going to be, 731 00:39:20,577 --> 00:39:21,910 which is really why they did it. 732 00:39:21,910 --> 00:39:25,150 And then they had this excuse of trying to reduce their computer 733 00:39:25,150 --> 00:39:27,592 budget for buying fancy computers for all 734 00:39:27,592 --> 00:39:29,800 their computer science departments around the country 735 00:39:29,800 --> 00:39:31,420 at universities. 736 00:39:31,420 --> 00:39:36,260 So it's an amazing story. 737 00:39:36,260 --> 00:39:38,530 We heard the story from Luyao last week 738 00:39:38,530 --> 00:39:40,375 about how personal computing comes about. 739 00:39:42,880 --> 00:39:46,480 A lot of that community is being-- 740 00:39:46,480 --> 00:39:51,070 came right out of Licklider's support of university research 741 00:39:51,070 --> 00:40:02,560 groups, and then moves over to carry it out at Xerox PARC. 742 00:40:02,560 --> 00:40:04,690 And then, of course, Xerox PARC is unable to do it, 743 00:40:04,690 --> 00:40:07,210 as you explained. 744 00:40:07,210 --> 00:40:10,120 But luckily, Steve Jobs buys the rights 745 00:40:10,120 --> 00:40:11,770 to walk through and see the stuff. 746 00:40:11,770 --> 00:40:15,790 And he's a great implementer. 747 00:40:15,790 --> 00:40:17,130 All right. 748 00:40:17,130 --> 00:40:19,383 Steph, what did I miss? 749 00:40:19,383 --> 00:40:20,800 STEPH: The one point that I really 750 00:40:20,800 --> 00:40:25,420 emphasize to set up the questions is something 751 00:40:25,420 --> 00:40:29,000 that Bill touched on, which is that there 752 00:40:29,000 --> 00:40:36,220 was a very significant dispute between the scientists 753 00:40:36,220 --> 00:40:38,470 and engineers who were trying to create the integrated 754 00:40:38,470 --> 00:40:44,140 system versus the isolated, siloed systems. 755 00:40:44,140 --> 00:40:46,300 And that ends up playing a big role in, I think, 756 00:40:46,300 --> 00:40:48,467 how many of you perceived the readings and the kinds 757 00:40:48,467 --> 00:40:50,320 of questions that you asked. 758 00:40:50,320 --> 00:40:53,800 I think the place where I'd like to start our conversation 759 00:40:53,800 --> 00:40:57,280 is about the mechanisms by which the computing revolution was 760 00:40:57,280 --> 00:40:58,390 created. 761 00:40:58,390 --> 00:41:00,820 We have a question from Beth, and she poses, 762 00:41:00,820 --> 00:41:03,130 how much of a technological revolution 763 00:41:03,130 --> 00:41:05,920 is the scientific breakthrough versus the vision 764 00:41:05,920 --> 00:41:07,940 of how to use it? 765 00:41:07,940 --> 00:41:10,570 So I think that really gets at the heart of 766 00:41:10,570 --> 00:41:14,590 whether or not a technology can be disruptive or revolutionary 767 00:41:14,590 --> 00:41:16,600 versus whether the implementation is 768 00:41:16,600 --> 00:41:20,328 what actually matters, as you just noted with Steve Jobs. 769 00:41:24,167 --> 00:41:26,250 AUDIENCE: Just to clarify, when you vision, do you 770 00:41:26,250 --> 00:41:30,390 mean JCR's vision, or do you mean the visionary who comes 771 00:41:30,390 --> 00:41:32,865 after, which is Steve Jobs? 772 00:41:32,865 --> 00:41:34,590 Which vision are you talking about? 773 00:41:34,590 --> 00:41:37,530 STEPH: Well, I think we'll let Beth clarify. 774 00:41:37,530 --> 00:41:40,320 BETH: I think it is kind of a cumulative process. 775 00:41:40,320 --> 00:41:42,760 The reason that Steve Jobs was able to have a vision 776 00:41:42,760 --> 00:41:46,080 was because JCR had his vision. 777 00:41:46,080 --> 00:41:49,470 So it's kind of like at each step in the road, the reason 778 00:41:49,470 --> 00:41:51,780 we were able to progress, was it because of the vision, 779 00:41:51,780 --> 00:41:54,380 or was it because of technology, or was it 780 00:41:54,380 --> 00:41:56,767 this guy was at the right place at the right time? 781 00:41:56,767 --> 00:41:58,350 How else could things have progressed, 782 00:41:58,350 --> 00:42:01,290 if it had been under different circumstance? 783 00:42:04,842 --> 00:42:06,300 STEPH: So how much weight, I guess, 784 00:42:06,300 --> 00:42:09,480 are we giving the invention, the mechanism by which this 785 00:42:09,480 --> 00:42:11,111 was achieved on its own? 786 00:42:14,800 --> 00:42:17,580 AUDIENCE: Well, I think a good way to think about it is-- 787 00:42:17,580 --> 00:42:19,550 you know ancient dynasties? 788 00:42:19,550 --> 00:42:21,260 There's one that was-- 789 00:42:21,260 --> 00:42:23,120 took a while and didn't do anything 790 00:42:23,120 --> 00:42:25,340 that was that significant. 791 00:42:25,340 --> 00:42:26,930 But they built up the infrastructure 792 00:42:26,930 --> 00:42:28,555 so that a dynasty comes in right after, 793 00:42:28,555 --> 00:42:31,055 and they end up doing all these awesome things, and everyone 794 00:42:31,055 --> 00:42:32,330 talks about that one. 795 00:42:32,330 --> 00:42:34,160 So I think it's kind of like Steve Jobs, 796 00:42:34,160 --> 00:42:36,785 Bill Gates have a voice because, again, billionaires-- everyone 797 00:42:36,785 --> 00:42:38,060 is like, oh. 798 00:42:38,060 --> 00:42:40,760 There's a saying where, I wish I was rich so then people 799 00:42:40,760 --> 00:42:41,930 would listen to me. 800 00:42:41,930 --> 00:42:45,980 So people pay attention because they became wealthy. 801 00:42:45,980 --> 00:42:48,230 And they kind of popularized the idea of the engineer, 802 00:42:48,230 --> 00:42:50,690 and they gave a lot of respect to the engineer, 803 00:42:50,690 --> 00:42:52,230 because now they make money. 804 00:42:52,230 --> 00:42:55,190 But in the past, like in the '70s, there's a couple series 805 00:42:55,190 --> 00:42:57,710 and a couple of biographies where there's really, 806 00:42:57,710 --> 00:42:59,750 really brilliant people that might be-- 807 00:42:59,750 --> 00:43:02,120 I would even say more brilliant than some people today-- 808 00:43:02,120 --> 00:43:03,480 in those technologies. 809 00:43:03,480 --> 00:43:08,690 But they just weren't valued economically or socially. 810 00:43:08,690 --> 00:43:12,238 In the '70s, being a nerd was being a nerd. 811 00:43:12,238 --> 00:43:13,780 Nowadays, being a nerd is like, damn, 812 00:43:13,780 --> 00:43:14,385 he might be a billionaire-- 813 00:43:14,385 --> 00:43:15,552 I'm going to be nice to him. 814 00:43:19,760 --> 00:43:23,020 Now it's cool to be a nerd-- to be smart. 815 00:43:23,020 --> 00:43:25,270 STEPH: I think we'll hit that point again in a second, 816 00:43:25,270 --> 00:43:27,245 so keep that thought process. 817 00:43:27,245 --> 00:43:29,718 RASHEED: I think Steve and Bill Gates had a vision socially 818 00:43:29,718 --> 00:43:31,510 like, we're going to make being a nerd cool 819 00:43:31,510 --> 00:43:34,540 again because we can't make a whole bunch of money. 820 00:43:34,540 --> 00:43:36,280 You think they had that vision when 821 00:43:36,280 --> 00:43:38,510 they were being nerds and making money, 822 00:43:38,510 --> 00:43:40,532 that we're going to make-- 823 00:43:40,532 --> 00:43:42,490 basically being a nerd cool because [INAUDIBLE] 824 00:43:42,490 --> 00:43:44,530 AUDIENCE: What I'm saying is-- so it's the difference 825 00:43:44,530 --> 00:43:45,030 between-- 826 00:43:48,280 --> 00:43:51,700 like somebody gets a car-- somebody has a Rolls Royce 827 00:43:51,700 --> 00:43:55,270 and somebody has a Toyota. 828 00:43:55,270 --> 00:43:56,667 This first wave-- 829 00:43:56,667 --> 00:43:58,750 WILLIAM BONVILLIAN: Licklider would have a junker. 830 00:43:58,750 --> 00:43:59,680 AUDIENCE: Hm? 831 00:43:59,680 --> 00:44:01,050 WILLIAM BONVILLIAN: Licklider would have a junker-- 832 00:44:01,050 --> 00:44:01,350 AUDIENCE: Yeah. 833 00:44:01,350 --> 00:44:02,660 WILLIAM BONVILLIAN: --that he personally would build. 834 00:44:02,660 --> 00:44:04,243 AUDIENCE: Well, what I'm trying to say 835 00:44:04,243 --> 00:44:05,620 is that it's an item that doesn't 836 00:44:05,620 --> 00:44:07,340 touch the hands of a lot of people, 837 00:44:07,340 --> 00:44:09,358 so it's not valued, versus-- 838 00:44:09,358 --> 00:44:10,900 the personal computer was a big thing 839 00:44:10,900 --> 00:44:12,525 because everyone could interact with it 840 00:44:12,525 --> 00:44:14,485 and would be able to communicate with it. 841 00:44:14,485 --> 00:44:16,110 These big computers, it was only if you 842 00:44:16,110 --> 00:44:19,660 were in the institution, if you were probably a grad student 843 00:44:19,660 --> 00:44:21,700 or more brilliant-- that you've got 844 00:44:21,700 --> 00:44:23,460 to interact with the computer. 845 00:44:23,460 --> 00:44:24,880 And what they really did was just 846 00:44:24,880 --> 00:44:27,940 make it common for everybody to use. 847 00:44:27,940 --> 00:44:30,770 Really that's what the entrepreneur-- the Gates, 848 00:44:30,770 --> 00:44:31,333 Ford-- 849 00:44:31,333 --> 00:44:33,250 is somebody who takes a complicated technology 850 00:44:33,250 --> 00:44:34,750 and simplifies it to the point where 851 00:44:34,750 --> 00:44:36,930 [INAUDIBLE] a large subset of consumers, 852 00:44:36,930 --> 00:44:38,591 if not all consumers. 853 00:44:45,970 --> 00:44:48,068 Any other questions? 854 00:44:48,068 --> 00:44:49,860 STEPH: Sorry, I'm comfortable with silence, 855 00:44:49,860 --> 00:44:51,752 so I like it when people think. 856 00:44:51,752 --> 00:44:53,460 AUDIENCE: What was interesting about this 857 00:44:53,460 --> 00:44:55,980 though, that I connected, is when you 858 00:44:55,980 --> 00:45:00,120 think about startups or any-- 859 00:45:00,120 --> 00:45:02,880 I guess vision organization, it always starts-- 860 00:45:02,880 --> 00:45:05,710 Reid Hoffman talks about this, which is you 861 00:45:05,710 --> 00:45:09,740 want to have a key insight about society. 862 00:45:09,740 --> 00:45:11,740 So not all people know this, but Mark Zuckerberg 863 00:45:11,740 --> 00:45:12,817 was a psychology major. 864 00:45:12,817 --> 00:45:14,900 He just happened to take computer science classes, 865 00:45:14,900 --> 00:45:16,360 and then would say that he studied computer science 866 00:45:16,360 --> 00:45:19,260 because it was easier when he got into the marketplace 867 00:45:19,260 --> 00:45:21,820 and he had to get investors. 868 00:45:21,820 --> 00:45:25,150 JCR was psychology, so he had a main insight 869 00:45:25,150 --> 00:45:27,970 into how the human brain worked, and also 870 00:45:27,970 --> 00:45:29,560 a secret about human nature-- 871 00:45:29,560 --> 00:45:32,170 that it seemed that he used that to his advantage in order 872 00:45:32,170 --> 00:45:35,170 to create computer systems, which I thought 873 00:45:35,170 --> 00:45:36,498 was really, really interesting. 874 00:45:36,498 --> 00:45:39,040 RASHEED: Yeah, I think I'm going to echo Martin's point here. 875 00:45:39,040 --> 00:45:43,095 I think a lot of this kind of visionary insight 876 00:45:43,095 --> 00:45:44,970 was built off of a lot of things that he did. 877 00:45:44,970 --> 00:45:46,870 So he had all this tactile knowledge, 878 00:45:46,870 --> 00:45:49,870 skill to build cars and things like that, but as you can see, 879 00:45:49,870 --> 00:45:54,100 probably used that psychology background 880 00:45:54,100 --> 00:45:57,210 a lot more when he was trying to figure out how are we 881 00:45:57,210 --> 00:45:59,920 going to put together and really move 882 00:45:59,920 --> 00:46:01,420 personal computing forward. 883 00:46:01,420 --> 00:46:04,920 And I think another understated point here was, 884 00:46:04,920 --> 00:46:08,440 in this great groups model, I think he 885 00:46:08,440 --> 00:46:11,410 and Licklider definitely not only underscores, 886 00:46:11,410 --> 00:46:14,410 but kind of creates that great groups model for DARPA. 887 00:46:14,410 --> 00:46:16,420 But not only bringing it together 888 00:46:16,420 --> 00:46:18,657 and finding these people, but-- 889 00:46:18,657 --> 00:46:20,740 I think you mentioned like a lot of these projects 890 00:46:20,740 --> 00:46:23,120 weren't on the normal three to five years. 891 00:46:23,120 --> 00:46:25,660 But he took maybe four or five different collaborators 892 00:46:25,660 --> 00:46:27,460 and said, we're going to build out 893 00:46:27,460 --> 00:46:31,125 10 years, 15 years, and then just 894 00:46:31,125 --> 00:46:32,500 stay working on the same projects 895 00:46:32,500 --> 00:46:34,030 and solve all these five different problems 896 00:46:34,030 --> 00:46:35,738 so he could move the whole field forward, 897 00:46:35,738 --> 00:46:40,060 instead of just focusing on one or two problems. 898 00:46:40,060 --> 00:46:42,610 In this great groups model, he creates 899 00:46:42,610 --> 00:46:46,030 great groups within DARPA, and then that leadership, as well, 900 00:46:46,030 --> 00:46:47,420 and moves that forward. 901 00:46:47,420 --> 00:46:52,540 And so he's obviously using his psychology insight and research 902 00:46:52,540 --> 00:46:54,790 maybe a little bit more than his computing aptitude. 903 00:46:57,390 --> 00:46:59,716 AUDIENCE: I think [INAUDIBLE] I remember 904 00:46:59,716 --> 00:47:02,091 in the reading we talked about the cool think about being 905 00:47:02,091 --> 00:47:03,870 an entrepreneur in DARPA is that you 906 00:47:03,870 --> 00:47:06,310 get to be an entrepreneur in multiple things at once. 907 00:47:06,310 --> 00:47:08,685 So if you're an entrepreneur, you only have your business 908 00:47:08,685 --> 00:47:10,590 and you can only really focus on that. 909 00:47:10,590 --> 00:47:13,410 But an entrepreneur in DARPA who leads a vision, 910 00:47:13,410 --> 00:47:15,640 gets to touch multiple products at once-- you 911 00:47:15,640 --> 00:47:20,220 can see that with JCR, where do you have the three sections. 912 00:47:20,220 --> 00:47:23,070 But what I find is interesting is why isn't he more famous? 913 00:47:23,070 --> 00:47:26,930 Because this is the first time I heard of him, honestly. 914 00:47:26,930 --> 00:47:28,650 He probably should get more credit 915 00:47:28,650 --> 00:47:32,400 than a jobs or Bill Gates. 916 00:47:32,400 --> 00:47:36,060 And even in the papers about Bill Gates, he was like, well, 917 00:47:36,060 --> 00:47:38,650 yeah, we really did this revolution in the '70s. 918 00:47:38,650 --> 00:47:42,890 And it's like, you did the last 2%. 919 00:47:42,890 --> 00:47:45,440 So it was pretty interesting. 920 00:47:45,440 --> 00:47:48,110 WILLIAM BONVILLIAN: He is an absolutely remarkable figure. 921 00:47:48,110 --> 00:47:54,350 And the core was this vision that he had, and that-- 922 00:47:57,330 --> 00:48:00,900 and this image of a crossover. 923 00:48:00,900 --> 00:48:04,080 Often, great technologies are crossover technologies. 924 00:48:04,080 --> 00:48:11,790 When Jobs creates the iPod, he's taking an MP3 player 925 00:48:11,790 --> 00:48:14,670 of high quality and certain new attributes, 926 00:48:14,670 --> 00:48:17,460 but he's joining with something entirely different-- 927 00:48:17,460 --> 00:48:20,790 a remarkable way of organizing the music industry. 928 00:48:20,790 --> 00:48:22,860 And it's two pieces that get joined together, 929 00:48:22,860 --> 00:48:25,530 and that's at the heart, I think, 930 00:48:25,530 --> 00:48:27,150 of what Licklider is able to do. 931 00:48:27,150 --> 00:48:30,390 He's able to take his capability at computing 932 00:48:30,390 --> 00:48:32,460 and his engineering and computing capabilities, 933 00:48:32,460 --> 00:48:34,965 but he crosses it with his whole background 934 00:48:34,965 --> 00:48:37,620 and set of understandings that he comes to with psychology. 935 00:48:37,620 --> 00:48:40,200 And it's that crossover that makes him 936 00:48:40,200 --> 00:48:41,380 an absolutely unique figure. 937 00:48:41,380 --> 00:48:45,540 So when you all are thinking about how to do great stuff, 938 00:48:45,540 --> 00:48:48,300 just keep that crossover idea. 939 00:48:48,300 --> 00:48:51,840 Bringing something from here and moving it over here 940 00:48:51,840 --> 00:48:54,600 is really at the heart of a tremendous amount 941 00:48:54,600 --> 00:48:55,290 of creativity. 942 00:48:55,290 --> 00:48:59,700 It may be the definition of creativity. 943 00:48:59,700 --> 00:49:02,250 It's what he pulls off. 944 00:49:02,250 --> 00:49:05,880 By the time Licklider finishes his second stint-- 945 00:49:05,880 --> 00:49:10,332 which he's not as happy at DARPA the second time around. 946 00:49:10,332 --> 00:49:12,540 He has a director that doesn't get along with as well 947 00:49:12,540 --> 00:49:15,180 as he did with Jack Ruina and another famous DARPA director, 948 00:49:15,180 --> 00:49:16,020 named Heilmeier-- 949 00:49:18,800 --> 00:49:19,890 Heilmeier, excuse me. 950 00:49:24,540 --> 00:49:27,150 That team of has worked their way through those three 951 00:49:27,150 --> 00:49:30,270 critical projects. 952 00:49:30,270 --> 00:49:32,880 They've done the internet by '69. 953 00:49:32,880 --> 00:49:36,690 They've done desktop computing, carried out 954 00:49:36,690 --> 00:49:41,580 with some DARPA support by largely ex-DARPA funded folks 955 00:49:41,580 --> 00:49:43,020 at Xerox PARC. 956 00:49:43,020 --> 00:49:45,870 And then they've done computer graphics, and a lot of that 957 00:49:45,870 --> 00:49:49,050 moves over to Xerox PARC, as well, and then from there 958 00:49:49,050 --> 00:49:50,340 to Apple. 959 00:49:50,340 --> 00:49:54,155 But they've worked their way through those pieces. 960 00:49:54,155 --> 00:49:55,905 Let me just mention the film here briefly. 961 00:50:00,060 --> 00:50:03,590 Rasheed, you can-- we'll get to this in a second, 962 00:50:03,590 --> 00:50:06,800 although we'll let Steph do a couple more questions. 963 00:50:06,800 --> 00:50:09,890 The film was this desperate attempt 964 00:50:09,890 --> 00:50:11,870 by these people who have created the internet. 965 00:50:11,870 --> 00:50:13,250 And you saw them. 966 00:50:13,250 --> 00:50:15,200 This is an amazing collection. 967 00:50:15,200 --> 00:50:16,850 A lot of on the film. 968 00:50:16,850 --> 00:50:19,380 Larry Roberts is there; and Frank Hart, who 969 00:50:19,380 --> 00:50:23,010 led the BBN team; and Bob Kahn, who's on the BBN team, 970 00:50:23,010 --> 00:50:26,720 then goes to DARPA and works with Vint Cerf on the TCPIP 971 00:50:26,720 --> 00:50:31,610 internet protocol system; and Paul Baran who 972 00:50:31,610 --> 00:50:33,620 has a vision for what-- how to put together 973 00:50:33,620 --> 00:50:36,350 the whole packet switching system to begin with, 974 00:50:36,350 --> 00:50:38,750 that's at the core of the internet. 975 00:50:38,750 --> 00:50:41,690 All these famous internet pioneers 976 00:50:41,690 --> 00:50:45,480 are all put together on the film, and they're desperate. 977 00:50:45,480 --> 00:50:49,340 They are desperate because the internet has only 978 00:50:49,340 --> 00:50:52,530 gone to about 50 different university computer science 979 00:50:52,530 --> 00:50:53,030 departments. 980 00:50:53,030 --> 00:50:55,370 And they realize that they've got this earth-shattering 981 00:50:55,370 --> 00:50:58,550 communication system, and they're trying to explain 982 00:50:58,550 --> 00:51:01,220 desperately to-- 983 00:51:01,220 --> 00:51:04,760 on a 16-millimeter film what they've accomplished. 984 00:51:08,900 --> 00:51:11,270 I think it's '72 they do the film. 985 00:51:11,270 --> 00:51:15,860 It's not until really 1993, 1994 that the internet completely 986 00:51:15,860 --> 00:51:16,730 scales up. 987 00:51:16,730 --> 00:51:20,330 So it's a long hiatus. 988 00:51:20,330 --> 00:51:23,850 And NSF actually plays a very, very important role 989 00:51:23,850 --> 00:51:28,760 in the scale up, because it gets turned over by DARPA and NSF-- 990 00:51:28,760 --> 00:51:33,650 and NSF under Erich Bloch, who is the only NSF director who 991 00:51:33,650 --> 00:51:34,850 came out of industry. 992 00:51:34,850 --> 00:51:41,180 So Erich Bloch was one of the leaders of the IBM 360 project 993 00:51:41,180 --> 00:51:43,127 and was a winner of the president's 994 00:51:43,127 --> 00:51:45,210 Medal of Technology, but he was an IBM officially. 995 00:51:45,210 --> 00:51:46,760 Wasn't a scientist. 996 00:51:46,760 --> 00:51:49,790 So he gets leadership of NSF, and he 997 00:51:49,790 --> 00:51:55,160 creates NSF's very strong computing organization, 998 00:51:55,160 --> 00:52:00,320 pulls one of the leaders of deck to run it. 999 00:52:00,320 --> 00:52:03,110 And they treat computing as a full science. 1000 00:52:03,110 --> 00:52:05,650 That's another moment in Licklider's vision 1001 00:52:05,650 --> 00:52:07,460 is when NSF starts to treat computer 1002 00:52:07,460 --> 00:52:09,500 science as a true science, not just 1003 00:52:09,500 --> 00:52:12,945 a nice tool for other scientists to play with. 1004 00:52:12,945 --> 00:52:14,570 So that's a pretty critical moment too. 1005 00:52:14,570 --> 00:52:18,980 And Erich Bloch and some others understand the power 1006 00:52:18,980 --> 00:52:20,570 of the internet as a communication 1007 00:52:20,570 --> 00:52:23,240 system in science, and then they spread it to universities 1008 00:52:23,240 --> 00:52:24,320 all over the place. 1009 00:52:24,320 --> 00:52:27,950 That creates a fabulous early audience, and then it scales. 1010 00:52:27,950 --> 00:52:30,500 So that's kind of what plays out there. 1011 00:52:30,500 --> 00:52:33,530 But it's a remarkable-- 1012 00:52:33,530 --> 00:52:36,770 another remarkable is part of the story here. 1013 00:52:36,770 --> 00:52:39,085 Steph, let's go to a question or two more. 1014 00:52:39,085 --> 00:52:39,627 STEPH: Great. 1015 00:52:39,627 --> 00:52:43,260 I have precisely two more questions, 1016 00:52:43,260 --> 00:52:49,590 the first is with a sentiment that 1017 00:52:49,590 --> 00:52:52,767 was shared in the last couple of pages in the reading. 1018 00:52:52,767 --> 00:52:53,850 And I'll read a quote now. 1019 00:52:53,850 --> 00:52:56,040 It says, "Technology isn't destiny, 1020 00:52:56,040 --> 00:52:58,830 no matter how inexorable its evolution may seem. 1021 00:52:58,830 --> 00:53:00,300 The way its capabilities are used 1022 00:53:00,300 --> 00:53:03,120 is much a matter of cultural choice and historical accident 1023 00:53:03,120 --> 00:53:05,620 as it's politics or fashion." 1024 00:53:05,620 --> 00:53:10,050 And I think back to maybe the first or the second course 1025 00:53:10,050 --> 00:53:13,080 where we began to reflect about the role of culture 1026 00:53:13,080 --> 00:53:15,780 and technology transfer and technology adoption, 1027 00:53:15,780 --> 00:53:19,200 and in thinking about the sort of functionalist ethics that we 1028 00:53:19,200 --> 00:53:21,300 might have-- this is by Amartya Sen-- 1029 00:53:21,300 --> 00:53:23,520 he's a very famous economists who came out of India 1030 00:53:23,520 --> 00:53:26,370 in developmental economics-- 1031 00:53:26,370 --> 00:53:31,380 and thinking about the ways in which we're evaluating vision. 1032 00:53:31,380 --> 00:53:35,530 And it seems-- and I have a couple of notes here from which 1033 00:53:35,530 --> 00:53:37,120 I will read-- 1034 00:53:37,120 --> 00:53:41,440 that many of the people who we have truly 1035 00:53:41,440 --> 00:53:45,130 regarded as innovators and as transformative members 1036 00:53:45,130 --> 00:53:47,650 of society have created things that 1037 00:53:47,650 --> 00:53:49,390 are effectively public goods. 1038 00:53:49,390 --> 00:53:51,760 They've really approached technology and engineering 1039 00:53:51,760 --> 00:53:54,400 from a utility ethics framework-- not so 1040 00:53:54,400 --> 00:53:56,830 much about having that bias to commercialization, 1041 00:53:56,830 --> 00:53:58,600 like maybe Steve Jobs did, but it 1042 00:53:58,600 --> 00:54:01,450 was the Lickliders and the Edisons who really thought 1043 00:54:01,450 --> 00:54:03,940 about how they could create a product that 1044 00:54:03,940 --> 00:54:06,445 was going to transform society for everyone. 1045 00:54:06,445 --> 00:54:09,070 It wasn't going to be a product that was going to be accessible 1046 00:54:09,070 --> 00:54:10,570 only to the few. 1047 00:54:10,570 --> 00:54:14,530 And so in thinking about our ethics of commercialization 1048 00:54:14,530 --> 00:54:18,520 today, perhaps I thought it would 1049 00:54:18,520 --> 00:54:21,070 be worth it to become a little philosophical, in the sense 1050 00:54:21,070 --> 00:54:23,860 that Licklider was very philosophical about the role 1051 00:54:23,860 --> 00:54:26,650 that technology played in the development of society-- 1052 00:54:26,650 --> 00:54:29,440 and ask ourselves about the ways in which our commercialization 1053 00:54:29,440 --> 00:54:31,990 bias may be precluding us from creating 1054 00:54:31,990 --> 00:54:33,610 the so-called disruptive inventions 1055 00:54:33,610 --> 00:54:34,887 that we still wish to pursue. 1056 00:54:39,408 --> 00:54:41,200 RASHEED: So I think this is a cool question 1057 00:54:41,200 --> 00:54:43,650 because you get to do both-- 1058 00:54:43,650 --> 00:54:47,260 bring in, just like you said, a lot of the cultural aspects 1059 00:54:47,260 --> 00:54:50,810 that might be overlooked when you have this commercialization 1060 00:54:50,810 --> 00:54:52,570 bias. 1061 00:54:52,570 --> 00:54:54,460 But I'm not sure-- 1062 00:54:54,460 --> 00:54:57,210 or in this context, I think Licklider's got a head up 1063 00:54:57,210 --> 00:54:58,768 because the field is wide open. 1064 00:54:58,768 --> 00:55:01,060 And so nobody's really thought about personal computing 1065 00:55:01,060 --> 00:55:02,935 before, so he gets to really define the field 1066 00:55:02,935 --> 00:55:05,340 and make it general enough so that it can be applicable 1067 00:55:05,340 --> 00:55:05,840 everywhere. 1068 00:55:05,840 --> 00:55:08,913 But in every sense, you get the idea that-- time-sharing 1069 00:55:08,913 --> 00:55:10,330 and all these things-- he's trying 1070 00:55:10,330 --> 00:55:14,020 to make computing a lot more accessible. 1071 00:55:14,020 --> 00:55:16,690 And so I think that's a really cool and seminal thought 1072 00:55:16,690 --> 00:55:18,880 to make things more accessible, and that'll 1073 00:55:18,880 --> 00:55:22,570 give you the innovative, transformative, and really 1074 00:55:22,570 --> 00:55:26,470 disruptive ideals-- to bring this technology from one 1075 00:55:26,470 --> 00:55:28,570 and spread it around. 1076 00:55:28,570 --> 00:55:30,550 And so the spreading effect, I'm not 1077 00:55:30,550 --> 00:55:34,500 sure how to quantify it like Licklider 1078 00:55:34,500 --> 00:55:38,320 did, but I'm interested in what projects that DARPA might be 1079 00:55:38,320 --> 00:55:39,910 running now or might be looking at, 1080 00:55:39,910 --> 00:55:42,760 to see if there's still those people within the organization 1081 00:55:42,760 --> 00:55:46,090 that are looking at bringing technologies that might exist 1082 00:55:46,090 --> 00:55:49,480 now that might just be [INAUDIBLE] or high risk, 1083 00:55:49,480 --> 00:55:52,570 and spreading that to a more normative phase, 1084 00:55:52,570 --> 00:55:54,692 maybe like Licklider did. 1085 00:55:54,692 --> 00:55:56,900 WILLIAM BONVILLIAN: It's an interesting point, Steph, 1086 00:55:56,900 --> 00:55:59,270 that you raise, and Rasheed, that you comment on, 1087 00:55:59,270 --> 00:56:02,840 because there's a very strong culture in computing, including 1088 00:56:02,840 --> 00:56:06,050 a very strong culture in MIT computing, 1089 00:56:06,050 --> 00:56:08,510 that computing should be free, and computing 1090 00:56:08,510 --> 00:56:11,630 should be in the commons, and computing should serve us all. 1091 00:56:11,630 --> 00:56:16,430 And that's actually a very strong ethos that, in a way-- 1092 00:56:16,430 --> 00:56:18,620 and I think it's your point-- 1093 00:56:18,620 --> 00:56:22,610 comes out of the way in which Licklider 1094 00:56:22,610 --> 00:56:25,700 and that remarkable team evolved this technology. 1095 00:56:25,700 --> 00:56:29,840 Because this is not a for-profit operation. 1096 00:56:29,840 --> 00:56:35,470 They are essentially trying to spring 1097 00:56:35,470 --> 00:56:38,920 a transformative technology onto the world 1098 00:56:38,920 --> 00:56:42,565 by building it in the commons, and creating it 1099 00:56:42,565 --> 00:56:45,340 into the commons, and making it universally accessible. 1100 00:56:45,340 --> 00:56:47,350 That's a remarkable project. 1101 00:56:47,350 --> 00:56:48,820 That's not normally the way we do 1102 00:56:48,820 --> 00:56:53,350 stuff in a capitalist country. 1103 00:56:53,350 --> 00:56:55,410 You've identified an interesting kind of strand 1104 00:56:55,410 --> 00:56:58,990 here that's a different kind of technology development story. 1105 00:56:58,990 --> 00:57:01,750 And it's a different story than the Edison story. 1106 00:57:07,180 --> 00:57:09,640 It enables, Martin, as you point out, 1107 00:57:09,640 --> 00:57:12,940 Gates and Jobs to get extremely rich. 1108 00:57:12,940 --> 00:57:16,000 But all the foundational stuff was 1109 00:57:16,000 --> 00:57:17,920 thrown into the commons long ahead of them, 1110 00:57:17,920 --> 00:57:21,580 which was an enabler for them, as well. 1111 00:57:21,580 --> 00:57:22,720 So interesting. 1112 00:57:22,720 --> 00:57:24,310 Thanks for bringing the philosophy in. 1113 00:57:24,310 --> 00:57:29,140 STEPH: [INAUDIBLE] for what it's worth, I find-- 1114 00:57:29,140 --> 00:57:30,940 or I personally think that doing good 1115 00:57:30,940 --> 00:57:35,290 is not exclusive mutually with doing well 1116 00:57:35,290 --> 00:57:36,832 for yourself and for other people. 1117 00:57:36,832 --> 00:57:38,290 I think that there's a lot of value 1118 00:57:38,290 --> 00:57:40,480 in thinking about the ways in which you 1119 00:57:40,480 --> 00:57:43,450 can reach a wide audience and transform society, and at 1120 00:57:43,450 --> 00:57:45,310 the same time, do well for yourself, 1121 00:57:45,310 --> 00:57:49,000 and then return that value onto society. 1122 00:57:49,000 --> 00:57:51,240 But to what extent? 1123 00:57:51,240 --> 00:57:53,080 The kinds of people who are capitalizing 1124 00:57:53,080 --> 00:57:56,140 on the gains of the Lickliders is-- 1125 00:57:56,140 --> 00:58:00,292 I'm not sure that they may all share my vision for the world. 1126 00:58:00,292 --> 00:58:01,750 AUDIENCE: I'll add a point to that. 1127 00:58:01,750 --> 00:58:06,280 I think a lot of it matters-- it matters how you define success. 1128 00:58:06,280 --> 00:58:08,120 What channel's creating wealth? 1129 00:58:08,120 --> 00:58:12,040 Because you'd say, yeah, these innovators didn't become 1130 00:58:12,040 --> 00:58:15,850 wealthy, or the US did become wealthy, 1131 00:58:15,850 --> 00:58:19,780 Bill Gates and Steve Jobs became wealthy, and their employees. 1132 00:58:19,780 --> 00:58:22,780 But at the same time, how much the net worth was increased 1133 00:58:22,780 --> 00:58:26,290 in the US based on the economic output, based on these people, 1134 00:58:26,290 --> 00:58:27,250 and also their legacy? 1135 00:58:27,250 --> 00:58:29,888 Because most likely in the future, we'll re-look into this 1136 00:58:29,888 --> 00:58:31,680 and say, actually, these were the pioneers. 1137 00:58:31,680 --> 00:58:33,013 And they get even more traction. 1138 00:58:35,650 --> 00:58:37,610 Bob Metcalfe, who made the internet, 1139 00:58:37,610 --> 00:58:39,230 made the original one. 1140 00:58:39,230 --> 00:58:41,157 And then other people implemented it and made 1141 00:58:41,157 --> 00:58:43,240 it better and better and better, so he's like, oh, 1142 00:58:43,240 --> 00:58:44,290 it's awesome, because I get credit 1143 00:58:44,290 --> 00:58:46,690 for all these brilliant things people did after me. 1144 00:58:46,690 --> 00:58:49,640 So I wouldn't have been able to do any of this, 1145 00:58:49,640 --> 00:58:52,750 but I get all this credit. 1146 00:58:52,750 --> 00:58:55,990 And I think that another sector that I could compare to, 1147 00:58:55,990 --> 00:58:57,620 which is space. 1148 00:58:57,620 --> 00:59:04,760 So NASA went very fast and used a lot of resources in order 1149 00:59:04,760 --> 00:59:09,262 to go into space, and they really set the framework. 1150 00:59:09,262 --> 00:59:10,970 I listened to an astronaut talk about it, 1151 00:59:10,970 --> 00:59:14,270 and he'd said, our job is to widen the horizons, 1152 00:59:14,270 --> 00:59:16,910 because it's going to be a while until it's commercially 1153 00:59:16,910 --> 00:59:20,270 viable for other people to do. 1154 00:59:20,270 --> 00:59:22,280 But that's our job, to really go into the places 1155 00:59:22,280 --> 00:59:24,080 where a capitalist can't go. 1156 00:59:24,080 --> 00:59:26,902 But there is a net positive on our society because of it. 1157 00:59:26,902 --> 00:59:28,610 And that's kind of the job of government, 1158 00:59:28,610 --> 00:59:30,560 to do the stuff that individuals can't 1159 00:59:30,560 --> 00:59:33,120 do for the net positive of the society. 1160 00:59:33,120 --> 00:59:34,610 So I think there is a lot of-- 1161 00:59:34,610 --> 00:59:36,530 there is a wealth creation there. 1162 00:59:36,530 --> 00:59:38,000 They create a wealth of knowledge. 1163 00:59:38,000 --> 00:59:42,920 They get respected and honored much higher maybe. 1164 00:59:42,920 --> 00:59:45,250 So I think it really depends how you define it. 1165 00:59:47,718 --> 00:59:50,010 WILLIAM BONVILLIAN: How about a closing thought, Steph? 1166 00:59:50,010 --> 00:59:52,590 STEPH: Yeah, that sets us up for the closing thought, 1167 00:59:52,590 --> 00:59:55,103 which is really about Licklider's legacy. 1168 00:59:55,103 --> 00:59:56,520 I really appreciated that Bill had 1169 00:59:56,520 --> 00:59:59,790 us read the component in which the author is having 1170 00:59:59,790 --> 01:00:05,550 his wife Linda reflect on Lick and the ways in which he passed 1171 01:00:05,550 --> 01:00:11,550 away, and the reaction from the computing community about 1172 01:00:11,550 --> 01:00:15,402 not only his legacy, but the ways in which he had been 1173 01:00:15,402 --> 01:00:16,860 such an incredible not only mentor, 1174 01:00:16,860 --> 01:00:20,070 but champion-- which was a word that you used today 1175 01:00:20,070 --> 01:00:21,790 maybe for the first time. 1176 01:00:21,790 --> 01:00:23,790 The kinds of mentors that one can have-- there's 1177 01:00:23,790 --> 01:00:25,830 the people who will help you learn things, 1178 01:00:25,830 --> 01:00:28,080 and then there's the people who will advocate for you. 1179 01:00:28,080 --> 01:00:30,390 And Lick was really the champion of computing, 1180 01:00:30,390 --> 01:00:34,110 not only for so many people, but for the field itself, 1181 01:00:34,110 --> 01:00:36,540 with ARPA being his champion. 1182 01:00:36,540 --> 01:00:40,050 And I think there's a lot to be said for finding those mentors, 1183 01:00:40,050 --> 01:00:43,380 and I hope that we in this room have found those mentors 1184 01:00:43,380 --> 01:00:46,080 in our own lives, or on the path to finding those people, 1185 01:00:46,080 --> 01:00:47,730 and that MIT and our institutions 1186 01:00:47,730 --> 01:00:50,340 generally help to connect us with the people who 1187 01:00:50,340 --> 01:00:51,860 will help us make the revolutions 1188 01:00:51,860 --> 01:00:54,472 that we want to see happen. 1189 01:00:54,472 --> 01:00:57,600 WILLIAM BONVILLIAN: Great, thank you. 1190 01:00:57,600 --> 01:01:00,720 Rasheed, a couple of thoughts about the film clip? 1191 01:01:00,720 --> 01:01:01,980 RASHEED: Yeah. 1192 01:01:01,980 --> 01:01:05,580 So I think the film clip is really a little bit weird 1193 01:01:05,580 --> 01:01:10,290 to watch, because it was really putting a face, and a name, 1194 01:01:10,290 --> 01:01:14,580 and a voice to all these people that we just read about. 1195 01:01:14,580 --> 01:01:17,402 And they came out about as I expected, especially because I 1196 01:01:17,402 --> 01:01:19,110 think the audience of the film was really 1197 01:01:19,110 --> 01:01:20,947 trying to make the internet ideas 1198 01:01:20,947 --> 01:01:23,130 and computing a lot more palatable 1199 01:01:23,130 --> 01:01:25,780 for the general audience. 1200 01:01:25,780 --> 01:01:27,900 And so it was nice to see them kind of struggle 1201 01:01:27,900 --> 01:01:30,165 with translating what they do on the daily basis 1202 01:01:30,165 --> 01:01:32,040 in their research, which is highly technical, 1203 01:01:32,040 --> 01:01:34,680 to general population. 1204 01:01:34,680 --> 01:01:37,390 And I think it's why be needed time to ask, 1205 01:01:37,390 --> 01:01:40,200 are these sort of efforts that DARPA should be doing 1206 01:01:40,200 --> 01:01:44,100 normally-- to take these high, very technical, 1207 01:01:44,100 --> 01:01:45,510 research-based-- 1208 01:01:45,510 --> 01:01:47,280 you might need a PhD to understand 1209 01:01:47,280 --> 01:01:49,680 what's going on on a project. 1210 01:01:49,680 --> 01:01:52,410 And then really translating it for that public good 1211 01:01:52,410 --> 01:01:54,644 of disseminating scientific information. 1212 01:01:57,307 --> 01:01:58,890 AUDIENCE: So to rephrase the question, 1213 01:01:58,890 --> 01:02:01,410 you're asking basically should DARPA be engaging in 1214 01:02:01,410 --> 01:02:02,410 high risk, high rewards? 1215 01:02:02,410 --> 01:02:04,330 WILLIAM BONVILLIAN: Little bit louder. 1216 01:02:04,330 --> 01:02:05,967 AUDIENCE: I'm trying to figure out-- 1217 01:02:05,967 --> 01:02:08,050 so I was thinking, is the question basically that, 1218 01:02:08,050 --> 01:02:14,350 should DARPA be engaging in relatively technical projects 1219 01:02:14,350 --> 01:02:18,250 that might have really huge impacts to society? 1220 01:02:18,250 --> 01:02:20,048 RASHEED: I think more so, should DARPA, 1221 01:02:20,048 --> 01:02:22,090 with each project that they take on-- should they 1222 01:02:22,090 --> 01:02:24,687 be advocating to the general public as much as possible? 1223 01:02:24,687 --> 01:02:26,770 STEPH: Should there be a communication component-- 1224 01:02:26,770 --> 01:02:27,010 AUDIENCE: OK. 1225 01:02:27,010 --> 01:02:28,120 STEPH: --to the research that they do, 1226 01:02:28,120 --> 01:02:29,980 much like they'll be doing this weekend at the Science 1227 01:02:29,980 --> 01:02:31,990 March or the Climate March next weekend. 1228 01:02:31,990 --> 01:02:32,840 AUDIENCE: Is DARPA going to be there? 1229 01:02:32,840 --> 01:02:33,673 STEPH: I don't know. 1230 01:02:33,673 --> 01:02:36,740 Will they? 1231 01:02:36,740 --> 01:02:39,370 WILLIAM BONVILLIAN: It's an interesting point about DARPA, 1232 01:02:39,370 --> 01:02:42,745 because in the IT revolution, DARPA 1233 01:02:42,745 --> 01:02:45,940 is working to stand up that revolution essentially 1234 01:02:45,940 --> 01:02:50,530 in the civilian economy, not in a closed-system defense 1235 01:02:50,530 --> 01:02:53,320 economy. 1236 01:02:53,320 --> 01:02:55,600 And there were conscious decisions 1237 01:02:55,600 --> 01:02:57,620 that DARPA made there. 1238 01:02:57,620 --> 01:03:00,640 So the military is often pretty good at getting 1239 01:03:00,640 --> 01:03:05,830 that first prototype out, the initial breakthrough prototype. 1240 01:03:05,830 --> 01:03:09,010 What it doesn't really have the resources to do 1241 01:03:09,010 --> 01:03:13,660 are the relentless series of follow-on incremental advances. 1242 01:03:13,660 --> 01:03:18,190 And clearly, the IT revolution that we've been living through 1243 01:03:18,190 --> 01:03:23,560 for the last 25 years is predominantly 1244 01:03:23,560 --> 01:03:25,900 incremental advances, some of which 1245 01:03:25,900 --> 01:03:28,270 can be obviously quite significant. 1246 01:03:28,270 --> 01:03:32,320 But the basic framework was kind of in place 1247 01:03:32,320 --> 01:03:35,410 more than 20 years ago, and we'd been in effect building 1248 01:03:35,410 --> 01:03:36,700 on that. 1249 01:03:36,700 --> 01:03:39,550 And DARPA understood that the military would never 1250 01:03:39,550 --> 01:03:41,440 have the resources-- that it can really 1251 01:03:41,440 --> 01:03:44,920 use this personal computing system 1252 01:03:44,920 --> 01:03:48,190 and the internet system-- they would be incredible tools 1253 01:03:48,190 --> 01:03:51,070 to create a whole new organizational set 1254 01:03:51,070 --> 01:03:52,840 of mechanisms for the military. 1255 01:03:52,840 --> 01:03:55,580 But it was going to stand it up in the civilian sector, 1256 01:03:55,580 --> 01:03:57,670 and then the military could buy it back 1257 01:03:57,670 --> 01:04:00,640 from the civilian sector after the civilian sector 1258 01:04:00,640 --> 01:04:03,160 did all the incremental work and flooded 1259 01:04:03,160 --> 01:04:04,690 the sector with investment. 1260 01:04:04,690 --> 01:04:09,040 So for a long period of time, the military 1261 01:04:09,040 --> 01:04:13,870 was the principal spender on computing and on software. 1262 01:04:13,870 --> 01:04:16,088 It dominated as the big buyer in that sector 1263 01:04:16,088 --> 01:04:18,130 for actually a fairly significant period of time. 1264 01:04:18,130 --> 01:04:19,835 That era is long gone. 1265 01:04:19,835 --> 01:04:21,460 The military's control of these sectors 1266 01:04:21,460 --> 01:04:26,740 is infinitesimally small now, because they've been stood up, 1267 01:04:26,740 --> 01:04:28,270 and the military buys them back. 1268 01:04:28,270 --> 01:04:31,210 So that's a different kind of way for DARPA to operate. 1269 01:04:31,210 --> 01:04:33,630 In the next piece we're going to do in a second, 1270 01:04:33,630 --> 01:04:35,380 we're going to talk about DARPA developing 1271 01:04:35,380 --> 01:04:38,710 technologies within the military and how 1272 01:04:38,710 --> 01:04:42,820 it operated in that way, which is, in some ways, 1273 01:04:42,820 --> 01:04:46,030 quite challenging. 1274 01:04:46,030 --> 01:04:48,620 Closing thought on the film? 1275 01:04:48,620 --> 01:04:50,095 Question? 1276 01:04:50,095 --> 01:04:51,220 AUDIENCE: Just very short-- 1277 01:04:51,220 --> 01:04:55,205 I do think there is stuff for DARPA to do, at the very least, 1278 01:04:55,205 --> 01:04:56,940 in today's society. 1279 01:04:56,940 --> 01:04:58,880 I worked on a DARPA-funded project. 1280 01:04:58,880 --> 01:05:01,970 People I generally talked to-- oh yeah, 1281 01:05:01,970 --> 01:05:04,220 DARPA is just a money sinkhole. 1282 01:05:04,220 --> 01:05:05,630 They just have a ton of money. 1283 01:05:05,630 --> 01:05:07,560 A lot of people don't necessarily 1284 01:05:07,560 --> 01:05:10,150 see how a defense agency's research comes down 1285 01:05:10,150 --> 01:05:11,150 to benefit them. 1286 01:05:18,645 --> 01:05:20,020 RASHEED: So as a closing thought, 1287 01:05:20,020 --> 01:05:22,630 I think DARPA has the advantage where they can 1288 01:05:22,630 --> 01:05:25,475 do this high risk, high reward. 1289 01:05:25,475 --> 01:05:27,850 Under the guise of defense, I think they kind of get away 1290 01:05:27,850 --> 01:05:30,100 with not being able to say-- 1291 01:05:30,100 --> 01:05:32,465 or being able to be like, it's classified, 1292 01:05:32,465 --> 01:05:35,865 or we don't really have to tell you under defense parameters 1293 01:05:35,865 --> 01:05:37,570 or protocols, or the reasons why we 1294 01:05:37,570 --> 01:05:40,420 want to look into these projects can 1295 01:05:40,420 --> 01:05:41,890 be a little more complicated. 1296 01:05:41,890 --> 01:05:43,530 But I do see-- 1297 01:05:43,530 --> 01:05:45,940 and in cases like this, where to have ARPANET, 1298 01:05:45,940 --> 01:05:48,323 which obviously has great potential, 1299 01:05:48,323 --> 01:05:49,990 if you do that type transfer and move it 1300 01:05:49,990 --> 01:05:51,573 into the civilian sector, and then you 1301 01:05:51,573 --> 01:05:54,470 get this IT revolution that follows on after. 1302 01:05:54,470 --> 01:05:56,740 s so I think it's important for DARPA 1303 01:05:56,740 --> 01:05:59,920 to keep initiatives like this for maybe bigger projects that 1304 01:05:59,920 --> 01:06:02,770 are a lot easier to stand up in civilian sectors. 1305 01:06:02,770 --> 01:06:06,220 WILLIAM BONVILLIAN: Yeah, an important point, Rasheed. 1306 01:06:06,220 --> 01:06:08,410 DARPA gets a lot of slack because it's 1307 01:06:08,410 --> 01:06:11,320 under that national security mission banner. 1308 01:06:11,320 --> 01:06:14,410 So it's able to take the kind of risks and aim 1309 01:06:14,410 --> 01:06:15,910 for a set of breakthroughs that's 1310 01:06:15,910 --> 01:06:20,110 harder for a civilian agency to undertake. 1311 01:06:20,110 --> 01:06:22,570 This is wild stuff. 1312 01:06:22,570 --> 01:06:25,630 Think about coming up with this stuff in the 1960s. 1313 01:06:25,630 --> 01:06:28,390 It's pretty amazing. 1314 01:06:28,390 --> 01:06:31,108 And that national security justification was pretty key. 1315 01:06:31,108 --> 01:06:32,650 All right, so now we're going to look 1316 01:06:32,650 --> 01:06:35,110 at a different side of DARPA. 1317 01:06:35,110 --> 01:06:38,470 This is a chapter from our textbook. 1318 01:06:38,470 --> 01:06:41,590 And there's a couple of topics here. 1319 01:06:41,590 --> 01:06:43,623 DARPA's a very unique institutional model 1320 01:06:43,623 --> 01:06:45,790 that operates at both the personal and institutional 1321 01:06:45,790 --> 01:06:46,980 levels, as we talked about. 1322 01:06:46,980 --> 01:06:49,565 And then secondly, DARPA's organizational rule set. 1323 01:06:52,180 --> 01:06:58,050 DARPA also creates lots of the foundational technologies 1324 01:06:58,050 --> 01:07:02,480 behind what is called the revolution in military affairs. 1325 01:07:02,480 --> 01:07:04,155 So here's the setting. 1326 01:07:07,410 --> 01:07:11,910 Harold Brown And Bill Perry are running Jimmy Carter's Defense 1327 01:07:11,910 --> 01:07:12,440 Department. 1328 01:07:12,440 --> 01:07:14,090 So Harold Brown's Secretary of Defense. 1329 01:07:14,090 --> 01:07:15,173 He's president of Caltech. 1330 01:07:17,370 --> 01:07:19,980 And Perry is out of Stanford. 1331 01:07:19,980 --> 01:07:21,930 They are great technologists. 1332 01:07:21,930 --> 01:07:25,110 Imagine the top leadership of the Defense Department 1333 01:07:25,110 --> 01:07:28,170 and the hands of really hot-shot technologists. 1334 01:07:28,170 --> 01:07:30,970 Doesn't usually happen. 1335 01:07:30,970 --> 01:07:32,940 And they've got a big problem on their hands. 1336 01:07:32,940 --> 01:07:37,170 The problem is the American army, in particular, 1337 01:07:37,170 --> 01:07:39,840 has been completely demoralized by the Vietnam War, 1338 01:07:39,840 --> 01:07:42,360 and has largely disintegrated. 1339 01:07:42,360 --> 01:07:47,340 It is not, in many ways, a truly functioning force. 1340 01:07:47,340 --> 01:07:50,010 And the Cold War is heating up, and there's 1341 01:07:50,010 --> 01:07:53,970 a monster presence of Soviet and Allied armies 1342 01:07:53,970 --> 01:07:55,410 in Eastern Europe. 1343 01:07:55,410 --> 01:07:58,860 And the problem that Brown and Perry have got 1344 01:07:58,860 --> 01:08:02,730 is that, because the US Army is weak and perceived as weak, 1345 01:08:02,730 --> 01:08:07,860 there's kind of nothing in the way of that Soviet army having 1346 01:08:07,860 --> 01:08:12,060 the capability of moving right into Western Europe. 1347 01:08:12,060 --> 01:08:17,080 And that's not a good situation because then you 1348 01:08:17,080 --> 01:08:22,130 have to decide whether to use nuclear weapons very quickly. 1349 01:08:22,130 --> 01:08:23,410 You don't have a lot of time. 1350 01:08:23,410 --> 01:08:24,910 You've got to decide whether you use 1351 01:08:24,910 --> 01:08:26,890 nuclear weapons, which essentially 1352 01:08:26,890 --> 01:08:29,290 means ending the world. 1353 01:08:29,290 --> 01:08:32,770 It's not the kind of decision you want to have to make. 1354 01:08:32,770 --> 01:08:38,330 So Perry and Brown started thinking about what 1355 01:08:38,330 --> 01:08:40,130 they call an offset strategy. 1356 01:08:40,130 --> 01:08:43,850 Can we start to restore the strength 1357 01:08:43,850 --> 01:08:45,649 of the American military? 1358 01:08:45,649 --> 01:08:49,160 They're not going to do it with a mass army, and a citizen 1359 01:08:49,160 --> 01:08:54,680 army, and building thousands of tanks, 1360 01:08:54,680 --> 01:08:57,950 and adding millions to the military. 1361 01:08:57,950 --> 01:09:00,140 That's unaffordable and they're just not-- 1362 01:09:00,140 --> 01:09:02,180 it's not going to happen. 1363 01:09:02,180 --> 01:09:03,740 And they'd be putting those people 1364 01:09:03,740 --> 01:09:05,870 into a broken institution anyway. 1365 01:09:05,870 --> 01:09:08,021 So the Army's in the process of restoring itself 1366 01:09:08,021 --> 01:09:09,979 through leadership of people like Colin Powell, 1367 01:09:09,979 --> 01:09:13,160 and a whole generation of new army leaders 1368 01:09:13,160 --> 01:09:15,500 that really restore that operation 1369 01:09:15,500 --> 01:09:17,600 in a remarkable kind of way. 1370 01:09:17,600 --> 01:09:18,740 It restores itself. 1371 01:09:18,740 --> 01:09:22,760 But that is an underway project at that time. 1372 01:09:22,760 --> 01:09:27,020 And what Perry and Brown start is a whole set 1373 01:09:27,020 --> 01:09:31,220 of thinking about can we get a significant on-the-ground 1374 01:09:31,220 --> 01:09:35,810 conventional warfare superiority so that the Soviet Union won't 1375 01:09:35,810 --> 01:09:39,470 even think about coming through the Fulda Gap 1376 01:09:39,470 --> 01:09:41,990 and moving on Western Europe. 1377 01:09:41,990 --> 01:09:47,330 And that's what they embark on, and that becomes DARPA's job. 1378 01:09:47,330 --> 01:09:56,060 So DARPA develops precision strike, they develop stealth, 1379 01:09:56,060 --> 01:09:59,090 and they develop UAVs or drones as core. 1380 01:09:59,090 --> 01:10:02,030 And then they're also doing a lot of stuff with submarines 1381 01:10:02,030 --> 01:10:03,380 that nobody ever talks about. 1382 01:10:03,380 --> 01:10:05,150 So I don't know what that is. 1383 01:10:05,150 --> 01:10:08,510 Something's going on there with submarines, undersea warfare. 1384 01:10:08,510 --> 01:10:10,170 But we know about the other three. 1385 01:10:10,170 --> 01:10:12,530 We have a pretty good idea what those look like. 1386 01:10:12,530 --> 01:10:20,690 And these are not technologies that the services want. 1387 01:10:20,690 --> 01:10:23,180 The Air Force is dominated by pilots. 1388 01:10:23,180 --> 01:10:25,160 They don't want drones. 1389 01:10:25,160 --> 01:10:27,970 That's like the last thing they want. 1390 01:10:27,970 --> 01:10:31,150 The Air Force wants airplanes that can actually fly. 1391 01:10:31,150 --> 01:10:33,340 They don't want these stealth things 1392 01:10:33,340 --> 01:10:36,742 that, if the computers weren't continually adjusting 1393 01:10:36,742 --> 01:10:38,200 the exact movement of the aircraft, 1394 01:10:38,200 --> 01:10:40,030 it would fall like a rock. 1395 01:10:40,030 --> 01:10:41,440 It's completely maneuverable. 1396 01:10:41,440 --> 01:10:43,740 It's like flying a bus. 1397 01:10:43,740 --> 01:10:44,965 No, really. 1398 01:10:44,965 --> 01:10:46,090 These are not maneuverable. 1399 01:10:46,090 --> 01:10:48,940 Nobody can see them, but it's still a bus. 1400 01:10:48,940 --> 01:10:52,420 So the Air Force doesn't want this thing. 1401 01:10:52,420 --> 01:10:55,480 But DARPA, in the case of stealth, 1402 01:10:55,480 --> 01:10:58,270 working with Lockheed-- which does some fascinating 1403 01:10:58,270 --> 01:11:01,150 translation of actually Russian scientific literature 1404 01:11:01,150 --> 01:11:02,830 and develops the concept of stealth-- 1405 01:11:02,830 --> 01:11:07,450 and then DARPA jumps on it with Lockheed and his Skunk Works-- 1406 01:11:07,450 --> 01:11:10,900 works on developing stealth. 1407 01:11:10,900 --> 01:11:13,750 It's working on precision strike. 1408 01:11:13,750 --> 01:11:15,430 Completely changes warfare. 1409 01:11:15,430 --> 01:11:18,400 Completely changes warfare. 1410 01:11:18,400 --> 01:11:20,980 In the Gulf War-- 1411 01:11:20,980 --> 01:11:23,862 which was where all this stuff got shown and demonstrated, 1412 01:11:23,862 --> 01:11:26,320 and that's where this phrase revolution in military affairs 1413 01:11:26,320 --> 01:11:27,220 comes from-- 1414 01:11:27,220 --> 01:11:28,720 the Russians saw this stuff. 1415 01:11:28,720 --> 01:11:30,670 Revolution in military affairs is their term 1416 01:11:30,670 --> 01:11:33,370 for what we pulled off. 1417 01:11:33,370 --> 01:11:37,540 They were stunned that these technologies created 1418 01:11:37,540 --> 01:11:41,020 a whole new kind of capability on the ground for the US 1419 01:11:41,020 --> 01:11:42,133 military that was just-- 1420 01:11:42,133 --> 01:11:43,300 they had no defense against. 1421 01:11:43,300 --> 01:11:47,560 We were essentially up against a Soviet-trained and style army 1422 01:11:47,560 --> 01:11:51,310 in the Gulf War in Iraq, and their air defense system 1423 01:11:51,310 --> 01:11:53,865 fell apart in a very short period of time. 1424 01:11:53,865 --> 01:11:55,240 They could not deal with stealth. 1425 01:11:55,240 --> 01:11:57,640 They could not deal with drones. 1426 01:11:57,640 --> 01:12:00,820 This continuous battlefield awareness, the airline battle 1427 01:12:00,820 --> 01:12:03,460 doctrine that was enabled by these different technologies 1428 01:12:03,460 --> 01:12:07,360 created an entirely different kind of military force. 1429 01:12:07,360 --> 01:12:10,510 So DARPA's doing this stuff, but this stuff 1430 01:12:10,510 --> 01:12:13,667 gets stood up within the military. 1431 01:12:13,667 --> 01:12:15,750 It's hard to think what the civilian application-- 1432 01:12:15,750 --> 01:12:18,520 we're now working away in civilian applications 1433 01:12:18,520 --> 01:12:20,830 on drones, so that's transferring over. 1434 01:12:20,830 --> 01:12:22,790 That could be pretty interesting. 1435 01:12:22,790 --> 01:12:24,790 It's already starting to get interesting. 1436 01:12:24,790 --> 01:12:27,070 GPS is another part of that story that's already 1437 01:12:27,070 --> 01:12:30,580 long since transferred over, that DARPA 1438 01:12:30,580 --> 01:12:33,910 works on helping to nurture. 1439 01:12:33,910 --> 01:12:38,260 Maybe stealth will transfer over, 1440 01:12:38,260 --> 01:12:43,510 because if we do something resembling autonomous vehicles, 1441 01:12:43,510 --> 01:12:47,710 you really want inverse stealth so that the radar systems can 1442 01:12:47,710 --> 01:12:49,690 really see your vehicle, that they're not 1443 01:12:49,690 --> 01:12:53,200 going to miss you in the noise. 1444 01:12:53,200 --> 01:12:57,130 Cars could look a lot weirder if they go stealth, but-- 1445 01:12:57,130 --> 01:12:58,720 inverse stealth, but that may be what 1446 01:12:58,720 --> 01:13:01,540 we need to do in order to make these autonomous vehicle 1447 01:13:01,540 --> 01:13:03,427 systems really operable. 1448 01:13:03,427 --> 01:13:05,320 AUDIENCE: Oh, you're saying autonomous cars. 1449 01:13:05,320 --> 01:13:06,940 WILLIAM BONVILLIAN: Yeah, autonomous cars. 1450 01:13:06,940 --> 01:13:07,480 I'm sorry. 1451 01:13:07,480 --> 01:13:08,990 Not drones-- cars. 1452 01:13:08,990 --> 01:13:09,746 AUDIENCE: I was thinking, why would you 1453 01:13:09,746 --> 01:13:10,720 want your drones to be visible? 1454 01:13:10,720 --> 01:13:12,137 WILLIAM BONVILLIAN: Yeah, so you'd 1455 01:13:12,137 --> 01:13:14,365 have a car that looks like the other side of stealth. 1456 01:13:21,850 --> 01:13:24,175 The military is a legacy sector. 1457 01:13:26,700 --> 01:13:28,440 It's been around since 1787. 1458 01:13:28,440 --> 01:13:30,570 It's extremely resistant to change. 1459 01:13:30,570 --> 01:13:32,880 It has its routines and doctrines, 1460 01:13:32,880 --> 01:13:38,190 and it has its doctrines that govern how it fights 1461 01:13:38,190 --> 01:13:39,040 and so forth. 1462 01:13:39,040 --> 01:13:43,470 And then here are Perry, and Brown, and DARPA 1463 01:13:43,470 --> 01:13:45,360 landing in the middle, forcing them 1464 01:13:45,360 --> 01:13:48,480 to adopt completely disruptive technologies 1465 01:13:48,480 --> 01:13:50,910 from what they've been doing. 1466 01:13:50,910 --> 01:13:55,050 So DARPA's operating here as a change agent within a legacy 1467 01:13:55,050 --> 01:13:56,490 sector military. 1468 01:13:56,490 --> 01:14:01,080 It's able to do the change agent role because of the island 1469 01:14:01,080 --> 01:14:02,630 bridge model. 1470 01:14:02,630 --> 01:14:06,180 Remember, we talked about this at Xerox PARC. 1471 01:14:06,180 --> 01:14:11,700 DARPA's got a bridge to the Secretary of Defense, 1472 01:14:11,700 --> 01:14:13,650 and the Secretary of Defense is able to make 1473 01:14:13,650 --> 01:14:16,950 it stick with the military. 1474 01:14:16,950 --> 01:14:20,460 And that's a really key DARPA capability. 1475 01:14:20,460 --> 01:14:23,520 That's also a lesson about how to bring change 1476 01:14:23,520 --> 01:14:25,620 into legacy sectors. 1477 01:14:25,620 --> 01:14:30,570 So DARPA rules of small and flexible, 1478 01:14:30,570 --> 01:14:32,550 and it's a flat organization, and it 1479 01:14:32,550 --> 01:14:34,620 allows a lot of freedom and autonomy 1480 01:14:34,620 --> 01:14:37,800 outside bureaucratic impediments to its program managers. 1481 01:14:37,800 --> 01:14:41,400 And it has an incredibly talented technical staff, 1482 01:14:41,400 --> 01:14:44,490 and it works off the challenge model. 1483 01:14:44,490 --> 01:14:46,830 The end of a project is the end of the project. 1484 01:14:46,830 --> 01:14:48,490 DARPA shuts things down. 1485 01:14:48,490 --> 01:14:50,460 It just doesn't persevere forever. 1486 01:14:50,460 --> 01:14:54,870 If it doesn't seem to be working, it'll close it down. 1487 01:14:54,870 --> 01:15:02,380 It's got all those rules sets, but in addition, it's 1488 01:15:02,380 --> 01:15:04,510 got some other capabilities that I 1489 01:15:04,510 --> 01:15:07,120 think are important to understanding how to bring 1490 01:15:07,120 --> 01:15:08,920 change into legacy sectors. 1491 01:15:11,590 --> 01:15:15,520 Form critical innovation institutions-- 1492 01:15:15,520 --> 01:15:17,920 that's a lesson that we're seeing here. 1493 01:15:17,920 --> 01:15:19,870 The Defense Department would not have 1494 01:15:19,870 --> 01:15:22,540 done the revolution in military affairs 1495 01:15:22,540 --> 01:15:25,420 if there wasn't a critical innovation system there 1496 01:15:25,420 --> 01:15:27,970 to do the innovation, ie DARPA. 1497 01:15:27,970 --> 01:15:31,450 So that's kind of a new rule for us to think about right. 1498 01:15:31,450 --> 01:15:33,760 What's going to be that critical innovation system that 1499 01:15:33,760 --> 01:15:36,700 can operate as a change agent? 1500 01:15:36,700 --> 01:15:40,750 Because most of the economy is legacy sectors. 1501 01:15:40,750 --> 01:15:43,600 Point two, use this island-bridge model. 1502 01:15:43,600 --> 01:15:44,980 It's really critical. 1503 01:15:44,980 --> 01:15:47,860 Remember, Edison can call up JP Morgan 1504 01:15:47,860 --> 01:15:51,880 and get JP Morgan to fund the electrical system. 1505 01:15:51,880 --> 01:15:53,960 That island-bridge thing is absolutely critical. 1506 01:15:53,960 --> 01:15:57,015 Get your innovation team on a protected island, 1507 01:15:57,015 --> 01:15:58,390 make sure they have a route back. 1508 01:15:58,390 --> 01:16:01,330 And DARPA's a great example of that. 1509 01:16:01,330 --> 01:16:03,400 Build a thinking community-- 1510 01:16:03,400 --> 01:16:06,100 that's what the Licklider story is about. 1511 01:16:06,100 --> 01:16:08,420 This is not just Licklider having a vision. 1512 01:16:08,420 --> 01:16:12,430 He's able to create a whole thought community within DARPA 1513 01:16:12,430 --> 01:16:14,710 and then by all the people that are working for DARPA, 1514 01:16:14,710 --> 01:16:18,850 like Doug angle and that whole world out there. 1515 01:16:18,850 --> 01:16:21,225 So they're able to scale up this community so 1516 01:16:21,225 --> 01:16:22,850 that it reaches a certain kind of mass, 1517 01:16:22,850 --> 01:16:25,183 and it can get a lot of stuff done and a lot of thinking 1518 01:16:25,183 --> 01:16:27,408 done, and develop a lot and nurture a lot of ideas. 1519 01:16:27,408 --> 01:16:29,950 So when you're thinking about implementing your technologies, 1520 01:16:29,950 --> 01:16:32,830 think about building that thinking community. 1521 01:16:32,830 --> 01:16:35,980 And then another key piece here is linking the technologies 1522 01:16:35,980 --> 01:16:38,690 to the operators. 1523 01:16:38,690 --> 01:16:41,717 So if the technologists are only talking to themselves 1524 01:16:41,717 --> 01:16:43,300 and they are not talking to the people 1525 01:16:43,300 --> 01:16:46,790 actually operating the system, nothing will happen. 1526 01:16:46,790 --> 01:16:48,460 So you've got to make sure, as you're 1527 01:16:48,460 --> 01:16:50,770 working on bringing your technologies about, 1528 01:16:50,770 --> 01:16:54,110 that you've got linkages to the actual operators. 1529 01:16:54,110 --> 01:16:57,418 So these are a few more lessons from DARPA. 1530 01:16:57,418 --> 01:16:59,710 And look, I'm going to go ahead and do Glenn Fong here, 1531 01:16:59,710 --> 01:17:04,650 because he's got some pieces about. 1532 01:17:04,650 --> 01:17:06,140 And Max, you've got Glenn, right? 1533 01:17:06,140 --> 01:17:06,640 OK. 1534 01:17:09,470 --> 01:17:13,600 Glenn just kind of tells us what we talked about earlier, 1535 01:17:13,600 --> 01:17:15,940 that so much of the IT revolution 1536 01:17:15,940 --> 01:17:18,760 really came out of this DARPA-supported thinking 1537 01:17:18,760 --> 01:17:20,080 community. 1538 01:17:20,080 --> 01:17:24,670 So Glenn argues that the Apple Lisa, and then its Mac 1539 01:17:24,670 --> 01:17:28,210 really came from Xerox PARC's Alto, as we've discussed. 1540 01:17:28,210 --> 01:17:31,360 And then in turn, Microsoft's Windows really 1541 01:17:31,360 --> 01:17:34,420 came from Apple's Mac, which in turn, came back 1542 01:17:34,420 --> 01:17:38,920 from Xerox PARC, which in turn came from this DARPA community. 1543 01:17:38,920 --> 01:17:43,510 So the key staff from PARC go to Apple 1544 01:17:43,510 --> 01:17:49,150 and to Microsoft as implementers on the pieces of the vision 1545 01:17:49,150 --> 01:17:52,330 that they're pursuing. 1546 01:17:52,330 --> 01:17:55,720 And it's all Lick's kids. 1547 01:17:55,720 --> 01:17:58,310 It's this community that Licklider starts to establish. 1548 01:17:58,310 --> 01:18:02,620 So Bob Taylor, as Luyao led us through last week, 1549 01:18:02,620 --> 01:18:05,260 he leads PARC. 1550 01:18:05,260 --> 01:18:07,150 The DARPA-funded university research 1551 01:18:07,150 --> 01:18:09,820 centers at places like MIT, and Berkeley, and Stanford, 1552 01:18:09,820 --> 01:18:15,370 and Carnegie Mellon, and SRI, and Utah, 1553 01:18:15,370 --> 01:18:19,990 they support and train a lot of that Xerox PARC staff. 1554 01:18:19,990 --> 01:18:22,420 So by creating the early computer science departments 1555 01:18:22,420 --> 01:18:26,410 and funding them, DARPA's training a whole thinking 1556 01:18:26,410 --> 01:18:29,080 community here. 1557 01:18:29,080 --> 01:18:34,930 And this is just a list of famous computer scientists 1558 01:18:34,930 --> 01:18:37,780 that got their support from DARPA. 1559 01:18:37,780 --> 01:18:41,830 And this is who did it. 1560 01:18:41,830 --> 01:18:43,810 This is that community. 1561 01:18:43,810 --> 01:18:46,810 And it's a stunning set of realizations 1562 01:18:46,810 --> 01:18:50,980 as to just how central DARPA is at the whole follow-on 1563 01:18:50,980 --> 01:18:52,660 to the IT revolution. 1564 01:18:52,660 --> 01:18:58,930 So Glenn Fong brings us to this set of realizations. 1565 01:18:58,930 --> 01:19:06,250 Things like Ed Catmull at Lucas Films, it's a remarkable-- 1566 01:19:06,250 --> 01:19:08,470 you see where the strands lead in off and all kinds 1567 01:19:08,470 --> 01:19:09,880 of directions. 1568 01:19:09,880 --> 01:19:12,520 So you get a sense for just how-- 1569 01:19:12,520 --> 01:19:15,130 what a great foundational institution this is. 1570 01:19:15,130 --> 01:19:19,390 So let's do some questions and points about these two 1571 01:19:19,390 --> 01:19:20,890 readings. 1572 01:19:20,890 --> 01:19:22,660 And Rasheed, do you want to lead us off 1573 01:19:22,660 --> 01:19:25,720 on that chapter from the textbook? 1574 01:19:25,720 --> 01:19:26,920 And then we'll go to Max. 1575 01:19:26,920 --> 01:19:28,500 RASHEED: Yeah. 1576 01:19:28,500 --> 01:19:31,330 So I think it's really important in this chapter. 1577 01:19:31,330 --> 01:19:33,490 as it highlights the elements of DARPA that 1578 01:19:33,490 --> 01:19:37,200 make it so fantastic and able to do all the things it's 1579 01:19:37,200 --> 01:19:38,518 been able to do. 1580 01:19:38,518 --> 01:19:40,060 So we talked about the island-bridge, 1581 01:19:40,060 --> 01:19:41,030 and things like that-- 1582 01:19:41,030 --> 01:19:46,450 but really the success of also the program managers 1583 01:19:46,450 --> 01:19:48,552 acting as sort of mini change agents 1584 01:19:48,552 --> 01:19:50,260 to be able to be that small and flexible. 1585 01:19:54,220 --> 01:19:57,460 I think somebody asked actually a pretty good question, which 1586 01:19:57,460 --> 01:20:03,520 was, is there a way to bring these elements of DARPA-- 1587 01:20:03,520 --> 01:20:05,835 giving these dynamic program managers 1588 01:20:05,835 --> 01:20:07,960 the freedom to do the things that they want to do-- 1589 01:20:07,960 --> 01:20:11,280 is there a way to do that outside of this DARPA model, 1590 01:20:11,280 --> 01:20:12,710 maybe enterprise to this model? 1591 01:20:12,710 --> 01:20:15,586 I think Martin asked that question. 1592 01:20:15,586 --> 01:20:16,584 AUDIENCE: [INAUDIBLE] 1593 01:20:16,584 --> 01:20:18,038 RASHEED: Yeah. 1594 01:20:18,038 --> 01:20:18,580 AUDIENCE: OK. 1595 01:20:18,580 --> 01:20:21,020 Yeah, I just thought it was really interesting, because 1596 01:20:21,020 --> 01:20:22,520 in business, a lot of time, you talk 1597 01:20:22,520 --> 01:20:24,978 about focusing on the alphas and getting rid of your betas. 1598 01:20:24,978 --> 01:20:26,853 The stuff that you're really, really good at, 1599 01:20:26,853 --> 01:20:29,420 focus on doing that, so if you're not good don't do it. 1600 01:20:29,420 --> 01:20:31,220 So you try to outsource everything. 1601 01:20:31,220 --> 01:20:35,485 So it'd be an interesting model because you might move faster. 1602 01:20:35,485 --> 01:20:36,860 And so that's what I was thinking 1603 01:20:36,860 --> 01:20:38,895 about-- a way of people with good ideas that 1604 01:20:38,895 --> 01:20:41,870 are very capable being linked with resources 1605 01:20:41,870 --> 01:20:45,560 in a much better manner to maybe being able to do more. 1606 01:20:51,162 --> 01:20:52,745 AUDIENCE: So I just want to interject. 1607 01:20:52,745 --> 01:20:56,098 I think you just described the MIT faculty. 1608 01:20:56,098 --> 01:20:58,750 We can hire postdocs and students 1609 01:20:58,750 --> 01:21:03,930 to do things that might not want to do, and they specialize. 1610 01:21:03,930 --> 01:21:06,553 AUDIENCE: The way I would think about it differently is-- 1611 01:21:06,553 --> 01:21:08,220 that would mean they're very book smart. 1612 01:21:08,220 --> 01:21:10,780 And like what kind of person would come into-- 1613 01:21:10,780 --> 01:21:14,560 through MIT faculty versus like people that might be-- 1614 01:21:14,560 --> 01:21:15,430 look at the people-- 1615 01:21:15,430 --> 01:21:17,170 Apple-- Steve Jobs-- 1616 01:21:17,170 --> 01:21:18,730 drop out-- Bill Gates-- 1617 01:21:18,730 --> 01:21:21,290 drop out. 1618 01:21:21,290 --> 01:21:23,200 You're definitely right, but I guess the way 1619 01:21:23,200 --> 01:21:24,010 I was trying to say it is more-- 1620 01:21:24,010 --> 01:21:25,552 AUDIENCE: You don't have to say that. 1621 01:21:25,552 --> 01:21:27,370 AUDIENCE: Well, you are right. 1622 01:21:27,370 --> 01:21:31,575 I meant more of a spectrum of personality types. 1623 01:21:31,575 --> 01:21:32,950 AUDIENCE: Yeah, it's interesting, 1624 01:21:32,950 --> 01:21:35,050 because I don't know-- 1625 01:21:35,050 --> 01:21:37,030 you guys see the research that they do. 1626 01:21:37,030 --> 01:21:39,250 A lot of these faculty that I come across 1627 01:21:39,250 --> 01:21:40,930 have spin out companies. 1628 01:21:40,930 --> 01:21:44,500 And no, they're not all becoming Bill Gates overnight, 1629 01:21:44,500 --> 01:21:49,923 but they're all very much into their own specialty. 1630 01:21:49,923 --> 01:21:51,340 AUDIENCE: Well, the research model 1631 01:21:51,340 --> 01:21:53,560 is a very productive model because you have somebody 1632 01:21:53,560 --> 01:21:56,060 who has key insights but didn't have the time to go and make 1633 01:21:56,060 --> 01:21:58,160 a company off of it, so you get somebody 1634 01:21:58,160 --> 01:21:59,930 from-- who's one of their grad students. 1635 01:21:59,930 --> 01:22:00,710 AUDIENCE: Exactly. 1636 01:22:00,710 --> 01:22:03,893 AUDIENCE: [INAUDIBLE] the professor [INAUDIBLE] 1637 01:22:03,893 --> 01:22:05,060 AUDIENCE: Oh god, which one? 1638 01:22:05,060 --> 01:22:06,300 Langer? 1639 01:22:06,300 --> 01:22:10,740 [INTERPOSING VOICES] 1640 01:22:10,740 --> 01:22:14,030 AUDIENCE: But also, research institutions 1641 01:22:14,030 --> 01:22:15,880 try to focus more on-- 1642 01:22:15,880 --> 01:22:20,370 they're called frog innovations, [INAUDIBLE] you move forward. 1643 01:22:20,370 --> 01:22:23,040 And so I was thinking more of hawk innovations, which 1644 01:22:23,040 --> 01:22:24,900 is a completely different field that no one's looking into, 1645 01:22:24,900 --> 01:22:26,817 because it doesn't seem to be that productive. 1646 01:22:26,817 --> 01:22:28,000 AUDIENCE: OK. 1647 01:22:28,000 --> 01:22:30,420 There are a bunch of frogs, is that what you're saying? 1648 01:22:30,420 --> 01:22:31,128 AUDIENCE: No, no. 1649 01:22:31,128 --> 01:22:33,630 Just in general the type of research that goes through. 1650 01:22:33,630 --> 01:22:36,410 Or it has to have a purpose that's already defined 1651 01:22:36,410 --> 01:22:37,675 based on a current paradigm. 1652 01:22:37,675 --> 01:22:38,300 AUDIENCE: Yeah. 1653 01:22:38,300 --> 01:22:39,860 AUDIENCE: So it's very risky to do anything different 1654 01:22:39,860 --> 01:22:41,480 because you're kind of ruining your thesis [INAUDIBLE] 1655 01:22:41,480 --> 01:22:41,570 AUDIENCE: Yeah. 1656 01:22:41,570 --> 01:22:43,815 And have to get money from somebody to do it, 1657 01:22:43,815 --> 01:22:45,440 and convince them it's going somewhere. 1658 01:22:45,440 --> 01:22:47,120 Yeah, OK. 1659 01:22:47,120 --> 01:22:47,840 Got it. 1660 01:22:47,840 --> 01:22:48,833 RASHEED: How does that fit in with what 1661 01:22:48,833 --> 01:22:50,583 you were saying with the alphas and betas, 1662 01:22:50,583 --> 01:23:00,310 where focus in on what you're good at [INAUDIBLE] 1663 01:23:00,310 --> 01:23:01,470 AUDIENCE: [INAUDIBLE] 1664 01:23:01,470 --> 01:23:04,530 AUDIENCE: I think the idea that he's trying to get at 1665 01:23:04,530 --> 01:23:06,802 is that you focus on what you're good at, 1666 01:23:06,802 --> 01:23:08,760 but in order to figure out what you're good at, 1667 01:23:08,760 --> 01:23:11,460 you have to have some failures along the way. 1668 01:23:11,460 --> 01:23:13,620 So you have to try and do different things, 1669 01:23:13,620 --> 01:23:17,250 and then maybe you find out, oh, this idea's really amazing he. 1670 01:23:17,250 --> 01:23:19,760 It'll make lots of money, and it-- 1671 01:23:19,760 --> 01:23:21,960 and we happen to know what we're doing. 1672 01:23:21,960 --> 01:23:23,370 But you might have a dozen things 1673 01:23:23,370 --> 01:23:26,203 that don't go so well before that, so it's a process. 1674 01:23:26,203 --> 01:23:28,620 AUDIENCE: You might come up with the key insight, in terms 1675 01:23:28,620 --> 01:23:30,150 of manufacturing, but you haven't 1676 01:23:30,150 --> 01:23:31,500 done large-scale manufacturing. 1677 01:23:31,500 --> 01:23:33,570 So how are you going to connect to a firm that already 1678 01:23:33,570 --> 01:23:36,112 understands that really well so that you can execute quickly? 1679 01:23:36,112 --> 01:23:37,330 So you try to figure it out. 1680 01:23:37,330 --> 01:23:38,250 But it'll take you a lot of capital, 1681 01:23:38,250 --> 01:23:40,620 will take you two to three years and you're screwed. 1682 01:23:40,620 --> 01:23:42,750 I was thinking more of like-- 1683 01:23:42,750 --> 01:23:45,060 in software, you have Amazon Web Services, 1684 01:23:45,060 --> 01:23:47,550 you have databases that are already huge that you can just 1685 01:23:47,550 --> 01:23:49,708 go port and you're through. 1686 01:23:49,708 --> 01:23:52,000 And it'd be interesting for that for like manufacturing 1687 01:23:52,000 --> 01:23:52,570 or biology. 1688 01:23:52,570 --> 01:23:54,195 Because also biology's super difficult, 1689 01:23:54,195 --> 01:24:00,630 because say you want to do a startup [INAUDIBLE] 1690 01:24:00,630 --> 01:24:01,350 RASHEED: Yeah. 1691 01:24:01,350 --> 01:24:05,120 So I guess obviously it's going to vary along different fields, 1692 01:24:05,120 --> 01:24:07,960 but I think manufacturing might be an interesting one, or even 1693 01:24:07,960 --> 01:24:10,660 biology and the intricacies there. 1694 01:24:10,660 --> 01:24:13,780 But I think this brings up a larger point, which 1695 01:24:13,780 --> 01:24:16,535 is an even bigger abstract question, 1696 01:24:16,535 --> 01:24:18,910 but follows along with this idea of creating these change 1697 01:24:18,910 --> 01:24:19,740 agents. 1698 01:24:19,740 --> 01:24:23,530 And so somebody asked, is there this sense that these change 1699 01:24:23,530 --> 01:24:26,050 agents, like the ones that are found in DARPA-- 1700 01:24:26,050 --> 01:24:27,760 is it important to create these change 1701 01:24:27,760 --> 01:24:33,070 agents by giving you a whole bunch of money and letting do 1702 01:24:33,070 --> 01:24:35,020 you do-- 1703 01:24:35,020 --> 01:24:36,950 use your own discretion, as product managers, 1704 01:24:36,950 --> 01:24:39,970 to figure out what you want to do, or to find 1705 01:24:39,970 --> 01:24:43,667 these people that are doing these fringe research 1706 01:24:43,667 --> 01:24:45,250 opportunities that might be high risk, 1707 01:24:45,250 --> 01:24:47,750 and they might not just have the funding, and go from there? 1708 01:24:47,750 --> 01:24:50,830 And so it's the do you want to create these change 1709 01:24:50,830 --> 01:24:52,990 agents by giving it to the people 1710 01:24:52,990 --> 01:24:55,990 that you think are capable, or do you want to go out and find 1711 01:24:55,990 --> 01:25:00,730 these people who may not be proven in their field yet, 1712 01:25:00,730 --> 01:25:04,853 or might not have research that is still at this high risk 1713 01:25:04,853 --> 01:25:05,353 phase? 1714 01:25:08,180 --> 01:25:10,300 STEPH: Could I ask clarifying question? 1715 01:25:10,300 --> 01:25:11,980 Do you mean in the context of DARPA, 1716 01:25:11,980 --> 01:25:15,370 or do you mean generally? 1717 01:25:15,370 --> 01:25:18,037 RASHEED: I would say generally, because DARPA-- 1718 01:25:18,037 --> 01:25:19,870 you have a model that's established already. 1719 01:25:19,870 --> 01:25:20,620 STEPH: OK. 1720 01:25:20,620 --> 01:25:23,215 The reason I ask that is because in one 1721 01:25:23,215 --> 01:25:24,340 of the readings for today-- 1722 01:25:24,340 --> 01:25:27,430 I believe it was the-- 1723 01:25:27,430 --> 01:25:29,460 it might have been the Fong reading actually-- 1724 01:25:29,460 --> 01:25:31,210 there's a quote from that technologist who 1725 01:25:31,210 --> 01:25:32,627 said that he wanted the government 1726 01:25:32,627 --> 01:25:34,390 to not touch engineers at all. 1727 01:25:34,390 --> 01:25:38,252 He said, "If the US wants [INAUDIBLE] jobs, better lives, 1728 01:25:38,252 --> 01:25:40,210 and a stronger economy, the best thing lobbyist 1729 01:25:40,210 --> 01:25:42,943 bureaucrats and politicians can do is leave us alone." 1730 01:25:42,943 --> 01:25:44,860 And so that's why I thought it was interesting 1731 01:25:44,860 --> 01:25:46,660 if you were asking in the context of DARPA 1732 01:25:46,660 --> 01:25:49,030 or in the context of an organization that was funded 1733 01:25:49,030 --> 01:25:50,470 by government dollars. 1734 01:25:50,470 --> 01:25:53,860 Because maybe what the implication 1735 01:25:53,860 --> 01:25:55,900 is, in this instance, for this person who 1736 01:25:55,900 --> 01:25:58,000 is rejecting the notion that government played 1737 01:25:58,000 --> 01:26:00,370 any role in the computational revolution of all, 1738 01:26:00,370 --> 01:26:02,740 is we don't want government help. 1739 01:26:02,740 --> 01:26:05,590 But maybe the model that they actually want 1740 01:26:05,590 --> 01:26:09,910 is give us the money, and then don't mess with us at all. 1741 01:26:09,910 --> 01:26:11,050 AUDIENCE: [INAUDIBLE] 1742 01:26:11,050 --> 01:26:13,535 STEPH: Yeah, exactly. 1743 01:26:13,535 --> 01:26:14,410 AUDIENCE: [INAUDIBLE] 1744 01:26:14,410 --> 01:26:15,785 WILLIAM BONVILLIAN: That's spoken 1745 01:26:15,785 --> 01:26:18,970 like someone who's done some annual reports before. 1746 01:26:18,970 --> 01:26:20,320 AUDIENCE: Is that the VC? 1747 01:26:20,320 --> 01:26:22,750 Is that from [INAUDIBLE]? 1748 01:26:22,750 --> 01:26:23,960 STEPH: This is from Draper. 1749 01:26:23,960 --> 01:26:24,668 AUDIENCE: Draper? 1750 01:26:26,830 --> 01:26:29,230 STEPH: It's on page 234, if you're 1751 01:26:29,230 --> 01:26:30,400 interested in looking at it. 1752 01:26:30,400 --> 01:26:34,680 AUDIENCE: [INAUDIBLE] I think what 1753 01:26:34,680 --> 01:26:38,268 he meant is more like please don't put more 1754 01:26:38,268 --> 01:26:39,810 regulations that I have to deal with, 1755 01:26:39,810 --> 01:26:40,977 when I'm trying to innovate. 1756 01:26:40,977 --> 01:26:42,730 It keeps getting in the way. 1757 01:26:45,700 --> 01:26:47,331 STEPH: I like Beth's point too, though. 1758 01:26:53,260 --> 01:26:54,410 I'm so sorry, Lily. 1759 01:26:54,410 --> 01:26:56,100 I like your point too. 1760 01:26:56,100 --> 01:26:58,650 Could you maybe elaborate in your experience 1761 01:26:58,650 --> 01:27:02,790 what it's like to innovate under the auspices of working 1762 01:27:02,790 --> 01:27:04,800 in a team with other people? 1763 01:27:04,800 --> 01:27:06,660 AUDIENCE: Yeah, so I've had experience 1764 01:27:06,660 --> 01:27:10,110 with National Science Foundation grants, 1765 01:27:10,110 --> 01:27:13,890 huge NASA Astrobiology Institute grants. 1766 01:27:13,890 --> 01:27:18,390 And their reporting requirements are intensive. 1767 01:27:18,390 --> 01:27:20,400 Usually they're on an annual basis. 1768 01:27:20,400 --> 01:27:26,250 And I think that the intention behind the annual reporting 1769 01:27:26,250 --> 01:27:30,490 is necessary and good. 1770 01:27:30,490 --> 01:27:35,000 But often, they're extremely cumbersome, 1771 01:27:35,000 --> 01:27:39,040 and the first year, really people are just making-- 1772 01:27:39,040 --> 01:27:40,930 including stuff that they've already done, 1773 01:27:40,930 --> 01:27:46,480 or even inflating what they've accomplished 1774 01:27:46,480 --> 01:27:47,620 to look good on the report. 1775 01:27:47,620 --> 01:27:49,630 So those are the only experiences 1776 01:27:49,630 --> 01:27:53,550 that I've had is with those two funding agencies. 1777 01:27:53,550 --> 01:27:57,012 AUDIENCE: Yeah, I think DARPA has the more [INAUDIBLE] 1778 01:27:57,012 --> 01:27:59,220 WILLIAM BONVILLIAN: Yeah, DARPA runs a different kind 1779 01:27:59,220 --> 01:28:00,970 of opposite operation. 1780 01:28:00,970 --> 01:28:05,970 There's there's, frankly, less paper, but the DARPA program 1781 01:28:05,970 --> 01:28:08,730 manager-- 1782 01:28:08,730 --> 01:28:12,970 at a traditional science agency, the whole decision process 1783 01:28:12,970 --> 01:28:16,940 is organized around whether or not you get the grant award. 1784 01:28:16,940 --> 01:28:19,958 And it's getting that award out the door. 1785 01:28:19,958 --> 01:28:22,250 That's how these organizations are really put together. 1786 01:28:22,250 --> 01:28:24,470 That's what the peer review process and selection 1787 01:28:24,470 --> 01:28:26,720 process is built around. 1788 01:28:26,720 --> 01:28:28,840 In DARPA, and in the-- 1789 01:28:28,840 --> 01:28:31,750 excuse me-- in some of the DARPA clones, 1790 01:28:31,750 --> 01:28:36,260 the program manager, that's like just getting the grant award. 1791 01:28:36,260 --> 01:28:37,700 That's only stage one. 1792 01:28:37,700 --> 01:28:40,970 Then the program manager moves into your home. 1793 01:28:40,970 --> 01:28:43,250 They're going to be on you all the time. 1794 01:28:43,250 --> 01:28:46,670 Someone has described DARPA as a set of geniuses 1795 01:28:46,670 --> 01:28:48,350 connected by a travel agent. 1796 01:28:48,350 --> 01:28:50,960 They spend at least 1/3 of their time on the road 1797 01:28:50,960 --> 01:28:57,080 out meeting with their teams, bringing ideas around. 1798 01:28:57,080 --> 01:28:59,840 They'll typically have a number of teams standing up 1799 01:28:59,840 --> 01:29:02,690 working on a portfolio of related projects, 1800 01:29:02,690 --> 01:29:05,330 and they're going to bring ideas from those different teams. 1801 01:29:05,330 --> 01:29:07,997 They're going to put those teams together to talk to each other, 1802 01:29:07,997 --> 01:29:10,310 and sometimes that's not easy, because they don't want 1803 01:29:10,310 --> 01:29:12,380 to really share their secrets. 1804 01:29:12,380 --> 01:29:14,370 So it's a complicated process to manage, 1805 01:29:14,370 --> 01:29:18,200 but the DARPA manager is there to help 1806 01:29:18,200 --> 01:29:20,960 the awardee get through the process 1807 01:29:20,960 --> 01:29:23,690 and work in any way possible to expedite things and move 1808 01:29:23,690 --> 01:29:24,540 things along. 1809 01:29:24,540 --> 01:29:27,800 So it's a completely different organizational model 1810 01:29:27,800 --> 01:29:31,100 for the program manager compared to most science agencies. 1811 01:29:31,100 --> 01:29:33,380 And the DARPA clones, like ARPA-E or IARPA 1812 01:29:33,380 --> 01:29:35,690 do the same thing. 1813 01:29:35,690 --> 01:29:38,420 And we'll talk about ARPA-E at a later class. 1814 01:29:38,420 --> 01:29:41,280 But it's just a different model. 1815 01:29:41,280 --> 01:29:47,210 And there's another capability-- we talked earlier about DARPA 1816 01:29:47,210 --> 01:29:50,750 capability of working at both the institutional level 1817 01:29:50,750 --> 01:29:53,690 and at the great groups level. 1818 01:29:53,690 --> 01:29:56,600 The purpose of this reading about its work 1819 01:29:56,600 --> 01:29:59,720 within the defense legacy sector is 1820 01:29:59,720 --> 01:30:02,510 to say that DARPA has been able to do innovation 1821 01:30:02,510 --> 01:30:06,360 within a legacy sector to a remarkable degree, 1822 01:30:06,360 --> 01:30:09,830 as well as do innovation, like in IT, 1823 01:30:09,830 --> 01:30:12,770 on a complete technology frontier. 1824 01:30:12,770 --> 01:30:16,700 That's a remarkable set of capabilities 1825 01:30:16,700 --> 01:30:20,330 to build into one place, and yet its model 1826 01:30:20,330 --> 01:30:22,730 has been able to fit both. 1827 01:30:22,730 --> 01:30:25,820 So it's another interesting achievement 1828 01:30:25,820 --> 01:30:28,170 that it's been able to pull off. 1829 01:30:28,170 --> 01:30:29,420 AUDIENCE: Just to go back to-- 1830 01:30:29,420 --> 01:30:31,030 I know in your book, you talk about how, 1831 01:30:31,030 --> 01:30:32,189 when you go to a legacy sector, it's 1832 01:30:32,189 --> 01:30:34,564 like you're parachuting down and they're shooting at you. 1833 01:30:34,564 --> 01:30:37,670 I think, for the military [INAUDIBLE] fields 1834 01:30:37,670 --> 01:30:40,600 and new areas, it might easier, and so it's easier to form. 1835 01:30:40,600 --> 01:30:45,660 But I think for energy, how do you make that distinction? 1836 01:30:45,660 --> 01:30:47,992 Because I feel other legacy sectors 1837 01:30:47,992 --> 01:30:49,863 would be a lot harder to come in. 1838 01:30:49,863 --> 01:30:51,530 WILLIAM BONVILLIAN: Well, the military's 1839 01:30:51,530 --> 01:30:53,180 a hard legacy sector, because remember, 1840 01:30:53,180 --> 01:30:54,800 the services have the money. 1841 01:30:54,800 --> 01:30:56,750 They have their procurement budgets. 1842 01:30:56,750 --> 01:30:59,330 And if they don't want to do it, it's really hard 1843 01:30:59,330 --> 01:31:00,210 to get them to do it. 1844 01:31:00,210 --> 01:31:03,620 So that's why that island-bridge model's so crucial for DARPA 1845 01:31:03,620 --> 01:31:06,020 being able to rely on the Secretary of Defense 1846 01:31:06,020 --> 01:31:08,580 to kind of override some of the directions 1847 01:31:08,580 --> 01:31:09,830 where the services want to go. 1848 01:31:09,830 --> 01:31:14,900 It's still not easy, but all of these three technology areas-- 1849 01:31:14,900 --> 01:31:17,420 position strike, UAVs, and stealth-- 1850 01:31:17,420 --> 01:31:20,930 were all stood up despite some pretty significant service 1851 01:31:20,930 --> 01:31:22,250 opposition. 1852 01:31:22,250 --> 01:31:25,220 Not uniformly-- different parts of the services 1853 01:31:25,220 --> 01:31:27,980 saw what the possibilities were, so I want to make that uniform. 1854 01:31:27,980 --> 01:31:29,600 But it was still tough. 1855 01:31:29,600 --> 01:31:31,950 These were not easy to implement. 1856 01:31:31,950 --> 01:31:35,790 AUDIENCE: So [INAUDIBLE] so would you 1857 01:31:35,790 --> 01:31:38,370 say that ARPA-E has been as successful [INAUDIBLE] 1858 01:31:38,370 --> 01:31:41,340 DARPA did, or how do you think about it? 1859 01:31:41,340 --> 01:31:44,600 WILLIAM BONVILLIAN: Yeah, we'll go through ARPA-E 1860 01:31:44,600 --> 01:31:49,310 when we get to that class, but I think ARPA-E is very young. 1861 01:31:49,310 --> 01:31:50,390 It's still very new. 1862 01:31:50,390 --> 01:31:51,930 AUDIENCE: Like 2009, right? 1863 01:31:51,930 --> 01:31:54,410 WILLIAM BONVILLIAN: Yeah, it's like seven years old. 1864 01:31:54,410 --> 01:31:56,870 So in some ways, the jury's still out, 1865 01:31:56,870 --> 01:31:59,600 and we'll see if they survive. 1866 01:31:59,600 --> 01:32:02,240 But looks very interesting. 1867 01:32:02,240 --> 01:32:05,330 So Max, let's do some of Glenn Fong's work. 1868 01:32:05,330 --> 01:32:06,410 AUDIENCE: Yeah, sure. 1869 01:32:06,410 --> 01:32:08,870 OK, so basically, Fong was talking 1870 01:32:08,870 --> 01:32:12,438 a lot about PARC, and Xerox, and how 1871 01:32:12,438 --> 01:32:14,480 a lot of the concepts that were developed at PARC 1872 01:32:14,480 --> 01:32:17,750 were eventually taken by Microsoft and Apple, 1873 01:32:17,750 --> 01:32:21,620 and basically created the PC revolution that we all 1874 01:32:21,620 --> 01:32:24,620 know and love. 1875 01:32:24,620 --> 01:32:28,430 Yeah, Apple fanboy. 1876 01:32:28,430 --> 01:32:31,580 So one of the questions that I had while I 1877 01:32:31,580 --> 01:32:34,400 was reading the Fong reading-- 1878 01:32:34,400 --> 01:32:36,020 maybe someone can clarify this-- 1879 01:32:36,020 --> 01:32:38,660 I didn't really understand-- so PARC's-- 1880 01:32:38,660 --> 01:32:41,000 a lot of their ideas were taken-- 1881 01:32:41,000 --> 01:32:43,850 virtually everything-- so how did PARC not 1882 01:32:43,850 --> 01:32:48,320 respond in any sort of legal way? 1883 01:32:48,320 --> 01:32:50,473 They could have sued. 1884 01:32:50,473 --> 01:32:52,640 Was it possibly because they didn't patent anything? 1885 01:32:52,640 --> 01:32:54,880 WILLIAM BONVILLIAN: [INAUDIBLE] Jobs bought the rights-- 1886 01:32:54,880 --> 01:32:55,220 AUDIENCE: OK. 1887 01:32:55,220 --> 01:32:56,870 WILLIAM BONVILLIAN: --to go through. 1888 01:32:56,870 --> 01:33:00,380 So he acquired those rights. 1889 01:33:00,380 --> 01:33:02,922 AUDIENCE: I thought it was just the rights to visit, or the-- 1890 01:33:02,922 --> 01:33:04,255 WILLIAM BONVILLIAN: Do you know? 1891 01:33:04,255 --> 01:33:05,160 Go ahead. 1892 01:33:05,160 --> 01:33:07,535 AUDIENCE: So I don't know if it was he bought the rights. 1893 01:33:07,535 --> 01:33:09,330 I know he funded them somehow. 1894 01:33:09,330 --> 01:33:11,000 And part of that funding was that, oh, 1895 01:33:11,000 --> 01:33:12,740 I get to check out your projects, 1896 01:33:12,740 --> 01:33:16,178 and I get to see [INAUDIBLE] and take those ideas how it is. 1897 01:33:16,178 --> 01:33:16,720 AUDIENCE: OK. 1898 01:33:19,823 --> 01:33:21,240 AUDIENCE: I'll clarify this later. 1899 01:33:21,240 --> 01:33:23,330 But he can do that because of the rights. 1900 01:33:23,330 --> 01:33:26,660 And then Gates could do it because-- 1901 01:33:26,660 --> 01:33:29,885 Jobs and Gates were friends, so Jobs showed Gates. 1902 01:33:29,885 --> 01:33:31,760 He's like, look at this cool thing I'm doing. 1903 01:33:31,760 --> 01:33:33,718 Because they're kind of in a different sectors. 1904 01:33:33,718 --> 01:33:35,780 So he's like, check it out. 1905 01:33:35,780 --> 01:33:36,530 What do you think? 1906 01:33:36,530 --> 01:33:37,560 Because he's trying to impress his friend, 1907 01:33:37,560 --> 01:33:39,170 who's also very successful. 1908 01:33:39,170 --> 01:33:43,387 And Gates is like, whoa, this is pretty great. 1909 01:33:43,387 --> 01:33:45,470 And then he goes and he copies, and he [INAUDIBLE] 1910 01:33:45,470 --> 01:33:46,690 WILLIAM BONVILLIAN: You're talking about Windows? 1911 01:33:46,690 --> 01:33:47,315 AUDIENCE: Yeah. 1912 01:33:47,315 --> 01:33:48,570 So he comes out with-- 1913 01:33:48,570 --> 01:33:49,375 was it Windows DOS? 1914 01:33:49,375 --> 01:33:51,500 He comes out with that right before [INAUDIBLE]---- 1915 01:33:51,500 --> 01:33:53,530 because this is going to be part of the new Apple Computer. 1916 01:33:53,530 --> 01:33:55,113 And he comes out with it right before, 1917 01:33:55,113 --> 01:33:57,780 so it looks like Gates did it, instead of Jobs. 1918 01:33:57,780 --> 01:33:59,810 And then Jobs literally goes like, we're done. 1919 01:34:02,690 --> 01:34:04,717 This is the reason Apple became so secretive. 1920 01:34:04,717 --> 01:34:06,050 They used to be a lot more open. 1921 01:34:06,050 --> 01:34:08,840 They became secretive because after that, he's like-- 1922 01:34:08,840 --> 01:34:11,412 any little thing can go out if somebody finds out, 1923 01:34:11,412 --> 01:34:12,620 we're not doing this anymore. 1924 01:34:12,620 --> 01:34:15,380 So it completely switched the hierarchy. 1925 01:34:15,380 --> 01:34:18,410 So that's the rights part, in terms of the-- 1926 01:34:18,410 --> 01:34:21,640 Jobs does it, and then when Gates copies-- 1927 01:34:21,640 --> 01:34:24,470 when Jobs takes Gates to court, Gates 1928 01:34:24,470 --> 01:34:28,130 can say, actually, we both stole it, but it's OK for him, 1929 01:34:28,130 --> 01:34:29,750 so it's OK for me. 1930 01:34:29,750 --> 01:34:30,320 [INAUDIBLE] 1931 01:34:30,320 --> 01:34:32,570 RASHEED: But like you just said, Jobs didn't steal it. 1932 01:34:32,570 --> 01:34:34,560 He paid for it. 1933 01:34:34,560 --> 01:34:36,380 He paid for the rights. 1934 01:34:36,380 --> 01:34:38,060 AUDIENCE: When they go to court, Gates 1935 01:34:38,060 --> 01:34:40,910 says, yeah, we both just took it from the same source. 1936 01:34:40,910 --> 01:34:42,590 It's not like it was originally Apple's. 1937 01:34:42,590 --> 01:34:45,350 So that's the element of the legality. 1938 01:34:45,350 --> 01:34:46,940 The second part-- the reason PARC 1939 01:34:46,940 --> 01:34:49,107 wouldn't do it is because you have this thing called 1940 01:34:49,107 --> 01:34:52,640 the innovator's dilemma, which is, in any company, 1941 01:34:52,640 --> 01:34:55,910 you can have a product line-- a great example's Kodak. 1942 01:34:55,910 --> 01:34:58,250 They came out with the first digital camera, 1943 01:34:58,250 --> 01:35:00,650 but you're already making so much money, 1944 01:35:00,650 --> 01:35:05,180 and you probably have a large revenue streams doing-- 1945 01:35:05,180 --> 01:35:06,860 taking pictures of Polaroids. 1946 01:35:06,860 --> 01:35:08,860 So this is like, if you're an Ethernet company-- 1947 01:35:08,860 --> 01:35:11,510 if you're 3Com, and you have these-- 1948 01:35:11,510 --> 01:35:13,010 you're putting wires everywhere. 1949 01:35:13,010 --> 01:35:15,430 It's not the next day it's like, yo, we 1950 01:35:15,430 --> 01:35:16,640 got to take all this out. 1951 01:35:19,658 --> 01:35:21,950 Another thing too is, once you become a public company, 1952 01:35:21,950 --> 01:35:23,600 every single quarter matters. 1953 01:35:23,600 --> 01:35:24,770 So you can never go back. 1954 01:35:27,282 --> 01:35:29,490 You can know you're wrong but you have to go forward, 1955 01:35:29,490 --> 01:35:30,620 or else you're just going to get kicked off 1956 01:35:30,620 --> 01:35:32,900 or you're going to get sued by shareholders. 1957 01:35:32,900 --> 01:35:35,320 It's a very complicated thing to do. 1958 01:35:35,320 --> 01:35:37,985 RASHEED: Also, I wonder how much PARC staff or Xerox staff 1959 01:35:37,985 --> 01:35:40,610 there is left to really benefit after all the staff is actually 1960 01:35:40,610 --> 01:35:43,190 just going to Apple and Microsoft afterward. 1961 01:35:43,190 --> 01:35:47,673 It's not that I didn't benefit because I was at Xerox 1962 01:35:47,673 --> 01:35:48,590 and I missed the wave. 1963 01:35:48,590 --> 01:35:49,090 It's like. 1964 01:35:49,090 --> 01:35:51,547 I just left Xerox to go work at Apple or Microsoft, 1965 01:35:51,547 --> 01:35:52,880 and then got the benefits later. 1966 01:35:52,880 --> 01:35:55,338 WILLIAM BONVILLIAN: So Luyao, you led us through Xerox PARC 1967 01:35:55,338 --> 01:35:56,390 very nicely last week. 1968 01:35:56,390 --> 01:36:00,680 Do you have thoughts on how come Xerox didn't 1969 01:36:00,680 --> 01:36:03,470 get the Alto, and personal computing, 1970 01:36:03,470 --> 01:36:07,520 and graphic interfaces, and the mouse out the door? 1971 01:36:07,520 --> 01:36:08,300 AUDIENCE: Yeah. 1972 01:36:08,300 --> 01:36:11,390 [INAUDIBLE] last week that they were 1973 01:36:11,390 --> 01:36:13,800 so much into the research they're doing, and so 1974 01:36:13,800 --> 01:36:17,080 passion that they didn't consider 1975 01:36:17,080 --> 01:36:21,940 commercializing or bringing it to the general audience at all. 1976 01:36:24,620 --> 01:36:29,110 One of the investors came to their meeting room 1977 01:36:29,110 --> 01:36:34,060 and sat on a beanbag saying, OK, I'm the person [INAUDIBLE] 1978 01:36:34,060 --> 01:36:39,353 Can you guys show me more, and I will commercialize all this 1979 01:36:39,353 --> 01:36:40,270 by making it possible? 1980 01:36:40,270 --> 01:36:41,440 But they didn't. 1981 01:36:41,440 --> 01:36:43,300 They totally ignored him. 1982 01:36:43,300 --> 01:36:46,390 So I think they had a chance, but they just 1983 01:36:46,390 --> 01:36:47,830 didn't the awareness, I think. 1984 01:36:47,830 --> 01:36:48,913 WILLIAM BONVILLIAN: Right. 1985 01:36:48,913 --> 01:36:51,000 It was a really serious island-bridge connection 1986 01:36:51,000 --> 01:36:51,815 problem. 1987 01:36:51,815 --> 01:36:55,780 AUDIENCE: They were probably missing one or two [INAUDIBLE] 1988 01:36:55,780 --> 01:36:59,500 in their group that were able to connect 1989 01:36:59,500 --> 01:37:01,752 with the market with their innovations. 1990 01:37:01,752 --> 01:37:03,210 WILLIAM BONVILLIAN: Part of this is 1991 01:37:03,210 --> 01:37:04,690 something Martin mentioned earlier, 1992 01:37:04,690 --> 01:37:07,440 which is Xerox was focused on its core competency, which 1993 01:37:07,440 --> 01:37:09,960 is photocopies. 1994 01:37:09,960 --> 01:37:15,180 And they didn't understand this explosion of personal computing 1995 01:37:15,180 --> 01:37:17,550 capability, because it didn't fit their core competency 1996 01:37:17,550 --> 01:37:18,750 model. 1997 01:37:18,750 --> 01:37:26,130 So they missed a staggering opportunity. 1998 01:37:26,130 --> 01:37:29,100 But fortunately, Jobs saved us, so it all worked out. 1999 01:37:29,100 --> 01:37:32,513 AUDIENCE: [INAUDIBLE] call it the beanbag effect. 2000 01:37:32,513 --> 01:37:34,680 WILLIAM BONVILLIAN: Too much time in those beanbags, 2001 01:37:34,680 --> 01:37:36,930 is that your point, Lily? 2002 01:37:36,930 --> 01:37:38,490 Right, that was a great photograph 2003 01:37:38,490 --> 01:37:39,900 of the crowd on beanbags. 2004 01:37:39,900 --> 01:37:44,370 You had a great picture for us, Luyao. 2005 01:37:44,370 --> 01:37:47,230 How about a closing thought, Max, on Fong-- 2006 01:37:47,230 --> 01:37:49,390 or a couple of closing thoughts? 2007 01:37:49,390 --> 01:37:52,390 AUDIENCE: Let's see. 2008 01:37:52,390 --> 01:37:53,850 How about a closing question? 2009 01:37:53,850 --> 01:37:54,420 WILLIAM BONVILLIAN: Sure. 2010 01:37:54,420 --> 01:37:54,770 AUDIENCE: All right. 2011 01:37:54,770 --> 01:37:56,445 WILLIAM BONVILLIAN: That's fine. 2012 01:37:56,445 --> 01:37:58,415 I only let you get one out, so it's all yours. 2013 01:37:58,415 --> 01:37:59,790 AUDIENCE: So one of the questions 2014 01:37:59,790 --> 01:38:03,540 that someone had previously mentioned this basically 2015 01:38:03,540 --> 01:38:05,790 there's this stereotype that government bureaucracy is 2016 01:38:05,790 --> 01:38:08,970 kind of slow, and sluggish, and full of red tape, 2017 01:38:08,970 --> 01:38:11,190 and it's horrible. 2018 01:38:11,190 --> 01:38:13,860 And is it possible that you could use-- 2019 01:38:13,860 --> 01:38:15,450 like DARPA as an example-- to say 2020 01:38:15,450 --> 01:38:19,920 that maybe this is unfounded, maybe it's not as realistic 2021 01:38:19,920 --> 01:38:21,770 as people think? 2022 01:38:21,770 --> 01:38:23,770 And to continue that-- there's a couple of parts 2023 01:38:23,770 --> 01:38:26,490 to this question-- is it unfounded, 2024 01:38:26,490 --> 01:38:32,880 and should the government try to redo its image, 2025 01:38:32,880 --> 01:38:35,550 for lack of a better phrase? 2026 01:38:35,550 --> 01:38:38,750 And is there really any advantage they can get from it? 2027 01:38:38,750 --> 01:38:40,750 AUDIENCE: [INAUDIBLE] 2028 01:38:40,750 --> 01:38:41,400 AUDIENCE: What? 2029 01:38:41,400 --> 01:38:43,033 Can you repeat [INAUDIBLE] 2030 01:38:43,033 --> 01:38:43,950 AUDIENCE: Yeah, sorry. 2031 01:38:43,950 --> 01:38:45,617 I know, I started rambling a little bit. 2032 01:38:48,210 --> 01:38:51,600 Is the government slow? 2033 01:38:51,600 --> 01:38:54,240 AUDIENCE: Is DARPA an instance where it's not slow? 2034 01:38:54,240 --> 01:38:56,040 AUDIENCE: Yes. 2035 01:38:56,040 --> 01:38:58,410 And does the government get any advantage 2036 01:38:58,410 --> 01:39:02,490 from making itself not look slow? 2037 01:39:02,490 --> 01:39:04,880 WILLIAM BONVILLIAN: That's a really important point, Max. 2038 01:39:04,880 --> 01:39:10,350 DARPA is stuck in one of the most notorious bureaucracies 2039 01:39:10,350 --> 01:39:15,540 on the planet, that dates back to like 1787. 2040 01:39:15,540 --> 01:39:18,480 So how is it able to do this stuff? 2041 01:39:18,480 --> 01:39:22,110 It has some very interesting rules and some capabilities 2042 01:39:22,110 --> 01:39:22,650 that it got. 2043 01:39:22,650 --> 01:39:27,600 One of its key rules is that-- 2044 01:39:27,600 --> 01:39:31,020 it's called other transactions authority. 2045 01:39:31,020 --> 01:39:37,260 And DARPA's argument is, look, we're not buying jet aircraft. 2046 01:39:37,260 --> 01:39:40,823 We're just buying what's in people's heads to do research. 2047 01:39:40,823 --> 01:39:42,240 So we don't have to go through all 2048 01:39:42,240 --> 01:39:43,980 that elaborate procurement process 2049 01:39:43,980 --> 01:39:46,140 to protect the taxpayers. 2050 01:39:46,140 --> 01:39:48,780 We're out of the Federal Procurement System. 2051 01:39:48,780 --> 01:39:52,320 So DARPA's able to contract with enormous freedom, 2052 01:39:52,320 --> 01:39:54,540 and very quickly. 2053 01:39:54,540 --> 01:39:58,890 And then another DARPA power that they've acquired 2054 01:39:58,890 --> 01:40:00,010 is that they can-- 2055 01:40:00,010 --> 01:40:02,700 they have what's called IPA authority-- 2056 01:40:02,700 --> 01:40:05,880 Intergovernmental Personnel Act authority. 2057 01:40:05,880 --> 01:40:08,580 So ironically, at one point, DARPA 2058 01:40:08,580 --> 01:40:12,750 was attempting to hire me. 2059 01:40:12,750 --> 01:40:14,900 And I don't know why they wanted me, 2060 01:40:14,900 --> 01:40:18,510 because I'm not a technologist, but they seemed to want this. 2061 01:40:18,510 --> 01:40:21,448 And I wasn't going to go to DARPA, 2062 01:40:21,448 --> 01:40:22,740 because I'm not a technologist. 2063 01:40:22,740 --> 01:40:25,080 I wouldn't be at the heart of the operation. 2064 01:40:25,080 --> 01:40:27,000 So it wasn't going to work. 2065 01:40:27,000 --> 01:40:29,580 But they said they can hire me in a day. 2066 01:40:29,580 --> 01:40:31,410 They said, you want to do this? 2067 01:40:31,410 --> 01:40:32,820 You're on tomorrow. 2068 01:40:32,820 --> 01:40:36,870 This Intergovernmental Personnel Act 2069 01:40:36,870 --> 01:40:39,120 enables them essentially to enter 2070 01:40:39,120 --> 01:40:42,720 into a contract with whatever organization 2071 01:40:42,720 --> 01:40:45,270 you're working for-- maybe it's a university. 2072 01:40:45,270 --> 01:40:49,890 They buy your services using this other transactions 2073 01:40:49,890 --> 01:40:52,380 authority, so that they can put you on the job 2074 01:40:52,380 --> 01:40:53,640 almost immediately. 2075 01:40:55,140 --> 01:40:59,550 The use this IPA authority with companies or with universities 2076 01:40:59,550 --> 01:41:03,060 to be able to hire in lickety-split time. 2077 01:41:03,060 --> 01:41:07,230 Corporations can't hire this fast, much less universities, 2078 01:41:07,230 --> 01:41:08,970 much less government. 2079 01:41:08,970 --> 01:41:14,250 So they're outside of two of the traditional banes 2080 01:41:14,250 --> 01:41:16,680 of governmental bureaucratic nightmares. 2081 01:41:16,680 --> 01:41:20,220 They got themselves free of that authority. 2082 01:41:20,220 --> 01:41:25,590 And interestingly, both ARPA-E and IARPA, the DARPA clones, 2083 01:41:25,590 --> 01:41:29,200 have the same capabilities. 2084 01:41:29,200 --> 01:41:32,232 So that's part of why they're able to act like this. 2085 01:41:32,232 --> 01:41:33,690 I don't think it's the whole story, 2086 01:41:33,690 --> 01:41:35,250 but that's a couple of pieces that they've 2087 01:41:35,250 --> 01:41:36,210 been able to acquire. 2088 01:41:39,480 --> 01:41:41,592 RASHEED: I would answer and say, I 2089 01:41:41,592 --> 01:41:43,050 guess, traditionally, you would say 2090 01:41:43,050 --> 01:41:45,980 government and bureaucracy [INAUDIBLE] the highlight 2091 01:41:45,980 --> 01:41:46,740 of what's slow is. 2092 01:41:46,740 --> 01:41:48,115 When you think of something slow, 2093 01:41:48,115 --> 01:41:51,330 you think of a congressional session or something like that. 2094 01:41:51,330 --> 01:41:52,680 STEPH: It's deliberate. 2095 01:41:52,680 --> 01:41:54,097 RASHEED: And it's deliberately so, 2096 01:41:54,097 --> 01:41:56,347 and then there's an entire argument that can get into. 2097 01:41:56,347 --> 01:41:57,900 Is government supposed to move slow 2098 01:41:57,900 --> 01:42:00,540 because they need to be this giant political machine that 2099 01:42:00,540 --> 01:42:04,710 shouldn't be swayed by just one charismatic convincing guy? 2100 01:42:04,710 --> 01:42:07,080 But anyway, I think it's important to understand 2101 01:42:07,080 --> 01:42:09,545 that DARPA doesn't move slow because often it 2102 01:42:09,545 --> 01:42:10,920 doesn't have this opportunity to, 2103 01:42:10,920 --> 01:42:12,870 for national security reasons. 2104 01:42:12,870 --> 01:42:17,340 It's just not prudent for us to move slow to develop, 2105 01:42:17,340 --> 01:42:19,570 I guess, like drones and things like that, 2106 01:42:19,570 --> 01:42:23,075 which might have been drastic-- 2107 01:42:23,075 --> 01:42:24,700 things would look drastically different 2108 01:42:24,700 --> 01:42:26,117 if someone else had done it first, 2109 01:42:26,117 --> 01:42:28,860 or if we went into the Gulf War without these capabilities. 2110 01:42:28,860 --> 01:42:31,430 STEPH: Transferring that example to something like genomics, 2111 01:42:31,430 --> 01:42:33,360 could you imagine the kinds of implications 2112 01:42:33,360 --> 01:42:39,210 that it could have on just fundamental human ethics 2113 01:42:39,210 --> 01:42:42,035 if they were able to create genetic engineering overnight 2114 01:42:42,035 --> 01:42:44,160 in the way that DARPA's been able to create things? 2115 01:42:46,002 --> 01:42:47,460 WILLIAM BONVILLIAN: Well, I do want 2116 01:42:47,460 --> 01:42:50,040 to tell you that DARPA now has a very strong 2117 01:42:50,040 --> 01:42:52,920 biological technologies office. 2118 01:42:52,920 --> 01:42:55,800 So look out, Steph. 2119 01:42:55,800 --> 01:42:59,423 AUDIENCE: [INAUDIBLE] 2120 01:42:59,423 --> 01:43:00,090 STEPH: Oh, yeah. 2121 01:43:00,090 --> 01:43:02,460 Well, it was interesting, because at the AAAS meeting 2122 01:43:02,460 --> 01:43:06,270 in February, one of the sessions that I went to-- actually with 2123 01:43:06,270 --> 01:43:09,570 Chuck Weiss, who is Bill's co-author for many 2124 01:43:09,570 --> 01:43:11,400 of the things-- 2125 01:43:11,400 --> 01:43:14,760 was a panel that included this guy named Gary Marchant who's 2126 01:43:14,760 --> 01:43:19,770 a lawyer ethicist and genomics engineer, who was talking 2127 01:43:19,770 --> 01:43:22,890 about the necessity to include a participatory ethics 2128 01:43:22,890 --> 01:43:25,770 in decision making for genomics. 2129 01:43:25,770 --> 01:43:28,260 And one of his primary arguments, as I recall it, 2130 01:43:28,260 --> 01:43:30,960 is truly that, if we want to have-- 2131 01:43:30,960 --> 01:43:33,690 if we want to do a good job of implementing this 2132 01:43:33,690 --> 01:43:36,000 at scale, nationally or globally, 2133 01:43:36,000 --> 01:43:39,150 we also have to do a good job of listening meaningfully 2134 01:43:39,150 --> 01:43:41,280 to populations and then, at the same time, 2135 01:43:41,280 --> 01:43:46,350 meaningfully incorporating their feedback into their creation 2136 01:43:46,350 --> 01:43:46,890 process. 2137 01:43:46,890 --> 01:43:49,380 And that's very difficult. And I think that's precisely why 2138 01:43:49,380 --> 01:43:51,960 government is slow-- or the word that I choose use-- 2139 01:43:51,960 --> 01:43:55,530 deliberate-- but it obviously does make innovation. 2140 01:43:55,530 --> 01:43:59,070 So to Max's point, I don't think that DARPA applies, 2141 01:43:59,070 --> 01:44:01,550 unfortunately, to most things.