1 00:00:10,670 --> 00:00:15,019 about privacy big data security 2 00:00:15,019 --> 00:00:15,769 we covered 3 00:00:15,769 --> 00:00:20,509 material the laying out some confessions some lessons a prediction or two 4 00:00:20,509 --> 00:00:23,900 I we're gonna have time for questions at the end so this is usually a pretty 5 00:00:23,900 --> 00:00:25,150 engaging topic 6 00:00:25,150 --> 00:00:28,659 how technology impacts our society so up 7 00:00:28,660 --> 00:00:32,829 compile them at during the talk and what we set up quite a bit I'm at the end 8 00:00:32,829 --> 00:00:36,250 to go through them I my name is in either on 9 00:00:36,250 --> 00:00:39,649 Michael said the product not X I 10 00:00:39,649 --> 00:00:43,960 meh did not exist quickly give you I from now love 11 00:00:43,960 --> 00:00:48,079 how and where we're merging were about the next generation big data to really 12 00:00:48,079 --> 00:00:51,489 understand data supplied things within the enterprise and across enterprises 13 00:00:51,489 --> 00:00:55,699 I to really discover the date that you have and 14 00:00:55,700 --> 00:01:00,120 compete with that efficiently and easily were slowly emerging at a stealth mode 15 00:01:00,120 --> 00:01:02,339 company was founded almost a year ago 16 00:01:02,340 --> 00:01:06,439 us a statement a lotta more interesting things coming out about organ 17 00:01:06,439 --> 00:01:10,439 not act up first confession I am NOT an attorney 18 00:01:10,439 --> 00:01:15,648 I'm an engineer I I have back on signal processing cryptography 19 00:01:15,649 --> 00:01:20,380 software I my under rousing like coins in the ring so I'm certainly more 20 00:01:20,380 --> 00:01:23,389 a geeky nerd and then-attorney although 21 00:01:23,389 --> 00:01:26,798 policy really does intrigue me I started early in my career 22 00:01:26,799 --> 00:01:30,959 designing advanced avionics system for launch vehicles I did a stint in about 23 00:01:30,959 --> 00:01:34,170 10 years developing online voting systems in the crypto protocols that 24 00:01:34,170 --> 00:01:34,749 were 25 00:01:34,749 --> 00:01:38,240 that are involved in that that was an amazing ride 26 00:01:38,240 --> 00:01:42,759 and now most recently I been the I was the first few privacy officer added 27 00:01:42,759 --> 00:01:45,770 hell yes i an off around their big 18 28 00:01:45,770 --> 00:01:49,039 so a that we've been doing the last 29 00:01:49,039 --> 00:01:52,359 for five years up which will be 30 00:01:52,359 --> 00:01:56,020 the the crux if the talk will be talking about what is 31 00:01:56,020 --> 00:02:00,969 the personal data supplied chain I we were at Intel is that the ended that day 32 00:02:00,969 --> 00:02:01,990 to supply chain 33 00:02:01,990 --> 00:02:06,009 we had about 20 billion public records we saw about 30 million visitors a month 34 00:02:06,009 --> 00:02:11,080 the company been around since 2003 the company lifetime is sold about $50 35 00:02:11,080 --> 00:02:12,239 million reports 36 00:02:12,239 --> 00:02:15,770 I we powered a power now only a few months 37 00:02:15,770 --> 00:02:19,140 my lap ob people search for AT&T 38 00:02:19,140 --> 00:02:23,179 a previously for ya who so usually for looking for someone online 39 00:02:23,180 --> 00:02:27,540 if they are very common and you can get away with Google search 40 00:02:27,540 --> 00:02:31,140 if they have a minutes or more common we're gonna find yourself and tell you 41 00:02:31,140 --> 00:02:32,140 for one of our 42 00:02:32,140 --> 00:02:36,730 proper wanna their properties I you know 43 00:02:36,730 --> 00:02:39,929 people asking joint ellios they were a 44 00:02:39,930 --> 00:02:43,790 data broker which is I usually interests have term 45 00:02:43,790 --> 00:02:49,750 I I when I joined they had a Better Business Bureau A+ rating up a nap they 46 00:02:49,750 --> 00:02:50,320 were 47 00:02:50,320 --> 00:02:53,489 I engaged in several investigation by 48 00:02:53,490 --> 00:02:57,350 investigation by Attorney General across the country by the Senate Commerce 49 00:02:57,350 --> 00:02:59,190 Committee was investigating them 50 00:02:59,190 --> 00:03:02,200 up I had to be there 51 00:03:02,200 --> 00:03:07,720 stupid or naive to jump into that but what I really found I really thought was 52 00:03:07,720 --> 00:03:08,060 that 53 00:03:08,060 --> 00:03:12,030 public data is an important element other stab with 54 00:03:12,030 --> 00:03:16,620 establishing trust between people and I felt there was a need to do it right 55 00:03:16,620 --> 00:03:20,750 and I thought I could do more from the inside an organization like that 56 00:03:20,750 --> 00:03:23,760 and just throwing rocks I I 57 00:03:23,760 --> 00:03:27,840 at them from the outside so during my tenure we improved our better business 58 00:03:27,840 --> 00:03:29,250 better business 59 00:03:29,250 --> 00:03:32,570 your rating from a nap now B-plus I 60 00:03:32,570 --> 00:03:36,090 I'm proud about the investigations are over I 61 00:03:36,090 --> 00:03:40,430 we've established the industry is something that I i 62 00:03:40,430 --> 00:03:43,930 is reputable and trustworthy I least in some 63 00:03:43,930 --> 00:03:47,630 eyes and but I didn't do that alone are far from it 64 00:03:47,630 --> 00:03:50,640 I when I first joined in Tally s I data 65 00:03:50,640 --> 00:03:54,600 privacy listening or of our toughest critics actually 66 00:03:54,600 --> 00:03:57,739 it was a much more modest or up 67 00:03:57,740 --> 00:04:01,720 we r look at people that we're really at risk 68 00:04:01,720 --> 00:04:05,910 for the kinda data that we provided in an hour risk groups included folks like 69 00:04:05,910 --> 00:04:06,420 victim 70 00:04:06,420 --> 00:04:11,489 victims of domestic violence and elected officials and police officers and judges 71 00:04:11,490 --> 00:04:16,139 those are the folks are often up the subject some stocking 72 00:04:16,139 --> 00:04:19,228 I and and offenders would use our data 73 00:04:19,228 --> 00:04:23,539 sometimes for odd those ill-gotten 74 00:04:23,539 --> 00:04:26,909 or are ill purposes I 75 00:04:26,910 --> 00:04:31,050 so I want to talk to families folks I the first on my list was 76 00:04:31,050 --> 00:04:35,240 the national network to end domestic violence I victims of domestic violence 77 00:04:35,240 --> 00:04:36,979 for typically stock 78 00:04:36,979 --> 00:04:41,430 typically when and I really wanted to understand what we can do 79 00:04:41,430 --> 00:04:44,680 I as a provider of personal information 80 00:04:44,680 --> 00:04:48,189 to these women and so I went to talk to Cindy Southwark 81 00:04:48,189 --> 00:04:52,009 who's an executive at the from there and messing around and I said 82 00:04:52,009 --> 00:04:56,240 what can we do better here and few pretty straightforward get my women out 83 00:04:56,240 --> 00:04:58,029 of your day 84 00:04:58,029 --> 00:05:01,710 and I I think that was a pretty clear 85 00:05:01,710 --> 00:05:05,419 directive and I think great in the lessons about 50,000 women 86 00:05:05,419 --> 00:05:08,770 United States that are victims of domestic violence would give us a lesson 87 00:05:08,770 --> 00:05:13,599 out and she went and tried to go compiled let's 88 00:05:13,599 --> 00:05:16,998 an she came back and women 89 00:05:16,999 --> 00:05:21,270 really don't like to be on anyway that's the last thing they want to be honest 90 00:05:21,270 --> 00:05:24,740 less I said okay haven't we get my pen and they could 91 00:05:24,740 --> 00:05:28,759 login opt out the record in that kinda 92 00:05:28,759 --> 00:05:33,870 kinda stopped that forbidden she came back you know what really be helpful 93 00:05:33,870 --> 00:05:37,499 why can't these women just suppressed their 94 00:05:37,499 --> 00:05:41,449 leaders contact information they don't want to opt out their entire profile 95 00:05:41,449 --> 00:05:45,639 Mike why because offenders will know that the women are proactively managing 96 00:05:45,639 --> 00:05:46,620 your profile 97 00:05:46,620 --> 00:05:50,800 and send them even more leave your incentive fender even more 98 00:05:50,800 --> 00:05:54,020 in San under more and so 99 00:05:54,020 --> 00:05:58,248 what would be nice if the trail different call the most contact 100 00:05:58,249 --> 00:06:02,999 mo safety sensitive information is the contact information just like that 101 00:06:02,999 --> 00:06:06,479 as a great idea I thought it was a just 102 00:06:06,479 --> 00:06:11,378 not as a good idea for at risk group as a good idea for our custom 103 00:06:11,379 --> 00:06:14,819 and so I really start to cook this idea that 104 00:06:14,819 --> 00:06:18,800 wanna be nice to be able to have access to your own background check and be able 105 00:06:18,800 --> 00:06:20,110 to manage 106 00:06:20,110 --> 00:06:23,889 privacy profile around your own background check 107 00:06:23,889 --> 00:06:27,460 and I started think about this more deeply and then I talk to another 108 00:06:27,460 --> 00:06:32,049 up privacy advocate side one by the name Sally happen 109 00:06:32,050 --> 00:06:35,409 is an anthropologist and a product designer design 110 00:06:35,409 --> 00:06:38,909 bill Gertz years ago as they look like bananas and 111 00:06:38,909 --> 00:06:43,039 so kids you know we love bananas so you put your grandson look like a banana and 112 00:06:43,039 --> 00:06:43,509 of course 113 00:06:43,509 --> 00:06:46,759 over 30 for I know they were for our boys 114 00:06:46,759 --> 00:06:50,199 they let go birds and so I ass out what you think a 115 00:06:50,199 --> 00:06:53,729 you know our product like working with same question is: 116 00:06:53,729 --> 00:06:57,058 sandy and she said you know the prominent LES 117 00:06:57,059 --> 00:07:01,800 you guys are all about voyeurism and not enough about narcissism 118 00:07:01,800 --> 00:07:05,520 Mike and it's the same 119 00:07:05,520 --> 00:07:08,599 no that's in the echo 120 00:07:08,599 --> 00:07:11,979 show me my own background check and so 121 00:07:11,979 --> 00:07:15,740 we did there's a site now culture Rep dot com 122 00:07:15,740 --> 00:07:19,770 it's a free site I you can look at your own back on track 123 00:07:19,770 --> 00:07:23,469 after you're properly authenticated through a knowledge-based authentication 124 00:07:23,469 --> 00:07:27,599 you can execute privacy controls on 125 00:07:27,599 --> 00:07:31,279 on that side you can suppress the latest contact information 126 00:07:31,279 --> 00:07:34,869 you you something that no one I in 127 00:07:34,869 --> 00:07:38,800 domestic violence community outreach group a private meeting after by I 128 00:07:38,800 --> 00:07:42,059 I was a really interesting idea came from a buddy of mine 129 00:07:42,059 --> 00:07:45,979 I when I started at Intelli as he looked up is back on track in 130 00:07:45,979 --> 00:07:49,039 the a criminal charges on his record 131 00:07:49,039 --> 00:07:53,969 in the charge was criminal trespassing 132 00:07:53,969 --> 00:07:58,020 as a criminal trespassing offense on my record 133 00:07:58,020 --> 00:08:01,318 all I did was jump the fence at a con at a concert 134 00:08:01,319 --> 00:08:04,379 in college somebody broke into a 135 00:08:04,379 --> 00:08:07,860 change I said well that's how the county reports 136 00:08:07,860 --> 00:08:11,770 we can't we don't we can change we change it 137 00:08:11,770 --> 00:08:15,099 I why can you at least let me comment on it give 138 00:08:15,099 --> 00:08:18,860 give provide my side of the story for so someone sees that record 139 00:08:18,860 --> 00:08:21,960 they will know my side of the story at a good night 140 00:08:21,960 --> 00:08:25,979 we could do that and so this I like to do that too you can actually comments 21 141 00:08:25,979 --> 00:08:27,969 that pulls your back on track 142 00:08:27,969 --> 00:08:31,819 well actually see your comment right next to 143 00:08:31,819 --> 00:08:35,300 any charges so I think 144 00:08:35,299 --> 00:08:38,789 this is maybe a little ahead of its time I 145 00:08:38,789 --> 00:08:43,120 I think there's going to be more and more up access your own information and 146 00:08:43,120 --> 00:08:44,730 the ability to manage 147 00:08:44,730 --> 00:08:47,750 affectively it's public information you should have 148 00:08:47,750 --> 00:08:51,310 a voice in I so that its perceived 149 00:08:51,310 --> 00:08:55,530 I correctly II and I think it's your responsibility 150 00:08:55,530 --> 00:08:59,680 i'd to manage and and you're right your own public data 151 00:08:59,680 --> 00:09:04,219 so popping up 11 for looking at data generally 152 00:09:04,220 --> 00:09:07,420 up not a big surprise is a lot of it 153 00:09:07,420 --> 00:09:10,680 I between the time in I have some you may know this 154 00:09:10,680 --> 00:09:15,010 from the time that the first cave people her on wall still about 20 155 00:09:15,010 --> 00:09:18,680 up to there's twenty exabyte up data 156 00:09:18,680 --> 00:09:22,109 transfer credits toward moved around that kinda thank 157 00:09:22,110 --> 00:09:25,900 four times that was moved around created 158 00:09:25,900 --> 00:09:29,180 stored in February this year 159 00:09:29,180 --> 00:09:32,459 ever by more than 160 00:09:32,460 --> 00:09:38,580 up hardest active data I wanna be stacked with a date with personal data 161 00:09:38,580 --> 00:09:42,830 and this is the public data supply chain that 162 00:09:42,830 --> 00:09:45,960 we were privy to Antalya and this is 163 00:09:45,960 --> 00:09:50,170 sorta where the data come from and where it goes come from three sources: 164 00:09:50,170 --> 00:09:54,500 predominately the headwaters of the supply chain 165 00:09:54,500 --> 00:09:57,510 are government typically 166 00:09:57,510 --> 00:10:00,950 vital records I property records 167 00:10:00,950 --> 00:10:04,370 licenses things like that I 168 00:10:04,370 --> 00:10:07,690 there's commercial record which come from 169 00:10:07,690 --> 00:10:11,760 businesses hi businesses that you actually to business where 170 00:10:11,760 --> 00:10:14,790 so when you get there really expensive oil change 171 00:10:14,790 --> 00:10:18,150 and you sign documents when you pay for 172 00:10:18,150 --> 00:10:21,709 your you may be signing away the data that you provided that 173 00:10:21,710 --> 00:10:27,980 oil change company and it'll be sold into this data supplied 174 00:10:27,980 --> 00:10:31,330 the last element of this is self-reported data that you put up and 175 00:10:31,330 --> 00:10:32,620 blogs that you might 176 00:10:32,620 --> 00:10:35,690 but in other public warrock I Twitter 177 00:10:35,690 --> 00:10:38,950 Facebook if your privacy settings on clampdown I 178 00:10:38,950 --> 00:10:42,060 so there's a lot of data I 179 00:10:42,060 --> 00:10:46,949 there's a analyst for the data that's in that comes in the state supply chain 180 00:10:46,950 --> 00:10:50,640 I and then it gets crimes together these big data engines 181 00:10:50,640 --> 00:10:55,480 and its use for a buncha stuff I in the risk environment its use for 182 00:10:55,480 --> 00:11:00,080 credit reports II and credit tax 183 00:11:00,080 --> 00:11:05,710 housing employment screening marketing behavioral targeting feeds this 184 00:11:05,710 --> 00:11:09,800 I is fed by the state supply chain ID background facts 185 00:11:09,800 --> 00:11:13,780 I as well directories that Intel is engaged in 186 00:11:13,780 --> 00:11:16,780 people ask white 187 00:11:16,780 --> 00:11:19,900 uses the people after this data I 188 00:11:19,900 --> 00:11:22,740 most of them are not regulated some of them are regulated by the Fair Credit 189 00:11:22,740 --> 00:11:25,770 Reporting Act the most of these are not regulated 190 00:11:25,770 --> 00:11:29,810 I they run from the mundane to the obscure up 191 00:11:29,810 --> 00:11:32,900 I someone put together Union and they wanna 192 00:11:32,900 --> 00:11:37,140 altogether addresses the people that they went to school with 193 00:11:37,140 --> 00:11:40,730 up love is actually a big used for consumer 194 00:11:40,730 --> 00:11:43,750 background checks check out today 195 00:11:43,750 --> 00:11:47,000 parents checking out a fiance I 196 00:11:47,000 --> 00:11:50,630 is a big use pro there some profound uses people 197 00:11:50,630 --> 00:11:53,830 adopted kids looking up there lot their birth parents 198 00:11:53,830 --> 00:11:59,050 is a common you up the one that is the most obscure i think is in 199 00:11:59,050 --> 00:12:04,930 sorta 3 o'clock fine owner up dog relative for trans 200 00:12:04,930 --> 00:12:09,060 this actually was use up an analyst at work for me 201 00:12:09,060 --> 00:12:12,459 and we always when are Alice interviewed with that we always have them 202 00:12:12,460 --> 00:12:15,150 look at our site and give them a critique are they would give us a 203 00:12:15,150 --> 00:12:16,260 presentation 204 00:12:16,260 --> 00:12:19,650 what they found with the site and she came in to do not use the site a couple 205 00:12:19,650 --> 00:12:23,689 years ago my dog needed a trans bone marrow transplants when you define 206 00:12:23,690 --> 00:12:27,330 my dog siblings we had find the owner Mike Dogg sibling when using tell you to 207 00:12:27,330 --> 00:12:28,450 do that 208 00:12:28,450 --> 00:12:31,790 i'd you are our service 209 00:12:31,790 --> 00:12:35,099 up so lots uses 210 00:12:35,100 --> 00:12:38,190 let's talk a little bit about how we I 211 00:12:38,190 --> 00:12:42,750 do all how do we do this linkage how do we 212 00:12:42,750 --> 00:12:46,180 accomplish this linkage problem up 213 00:12:46,180 --> 00:12:49,959 we take billions records that fifteen billion record cell 214 00:12:49,960 --> 00:12:53,750 and we have in this inning United States typically will 215 00:12:53,750 --> 00:12:57,120 and we have to take billion record after Maxim 216 00:12:57,120 --> 00:13:00,730 to the hundreds of millions of people United States 217 00:13:00,730 --> 00:13:03,790 and soul ocular same name I 218 00:13:03,790 --> 00:13:07,760 clearly Carol Brooks for example there's about 1300 people named Karen Blixen 219 00:13:07,760 --> 00:13:10,380 United States might have about 10,000 records 220 00:13:10,380 --> 00:13:13,100 for people named Carol Brooks and you have to find the right Carol Brooks 221 00:13:13,100 --> 00:13:14,100 records 222 00:13:14,100 --> 00:13:17,560 for the correct hell Brooks people I for me 223 00:13:17,560 --> 00:13:23,209 others about 200 records for the 35 yr sodium Adler's 224 00:13:23,210 --> 00:13:27,780 in in the country and so we do this record linkage problem 225 00:13:27,780 --> 00:13:31,180 and what comes out a bit of a cluster of record so you see at the bottom right 226 00:13:31,180 --> 00:13:32,400 there is this cluster 227 00:13:32,400 --> 00:13:36,439 records and up you see the gym now for about 228 00:13:36,440 --> 00:13:39,890 ten o'clock that an ambulance facing lawyer 229 00:13:39,890 --> 00:13:44,699 from Houston Texas who is the bane of my existence because he really punches 230 00:13:44,700 --> 00:13:48,070 records out there and he always shows up higher than Lauren 231 00:13:48,070 --> 00:13:51,090 I'm I'm the lame NABARD about 8 o'clock 232 00:13:51,090 --> 00:13:55,220 up from now Mountain View Calif 233 00:13:55,220 --> 00:13:58,860 I so these are all this is the cluster ball the gym abner 234 00:13:58,860 --> 00:14:01,990 up that are that exist 235 00:14:01,990 --> 00:14:05,980 and so these are available for purchase at places like Antalya 236 00:14:05,980 --> 00:14:10,300 other sites we can get these profiles 237 00:14:10,300 --> 00:14:14,430 we also offer opt out you can opt out not to suppress and relays contact 238 00:14:14,430 --> 00:14:15,199 information 239 00:14:15,200 --> 00:14:19,610 but for free you can opt out a profile and people ask 240 00:14:19,610 --> 00:14:23,860 when someone knocks out you delete the record believe the profile 241 00:14:23,860 --> 00:14:27,210 confession no we do not and there's 242 00:14:27,210 --> 00:14:30,720 something called the fair Information practice principles 243 00:14:30,720 --> 00:14:34,450 that talk about data minimisation and minimal use 244 00:14:34,450 --> 00:14:39,560 and things like that fair use a beta and people ask why don't you delete the data 245 00:14:39,560 --> 00:14:43,030 I because we feel that there we certainly suppress it 246 00:14:43,030 --> 00:14:46,480 we certainly honor the opt-out would be in big trouble with the Federal Trade 247 00:14:46,480 --> 00:14:47,150 Commission 248 00:14:47,150 --> 00:14:50,680 within on the opt-out I 249 00:14:50,680 --> 00:14:54,089 but sometimes you can do better by not pull the race in the day and here's an 250 00:14:54,090 --> 00:14:55,380 example 251 00:14:55,380 --> 00:14:58,930 so here's to record his a profile with a few record Jane Hampton 252 00:14:58,930 --> 00:15:02,209 from Peoria Illinois we have to record for 253 00:15:02,210 --> 00:15:05,280 one has our date of birth about 254 00:15:05,280 --> 00:15:08,290 just turned 15 years old a minor 255 00:15:08,290 --> 00:15:12,079 might ask why minor data in your in your information 256 00:15:12,080 --> 00:15:16,040 cash at all sometimes he gets in 257 00:15:16,040 --> 00:15:20,089 I as much as we don't want date on minor some time to get into the state supply 258 00:15:20,090 --> 00:15:21,640 chain and 259 00:15:21,640 --> 00:15:26,150 we don't went nowhere comes from we are races are entices and other data brokers 260 00:15:26,150 --> 00:15:28,720 are gracious acquires a date and sometimes 261 00:15:28,720 --> 00:15:31,900 data from miners getting there and so 262 00:15:31,900 --> 00:15:34,990 we put it together we put these two records together one records data for 263 00:15:34,990 --> 00:15:37,500 the other records not the day to Berkeley link them 264 00:15:37,500 --> 00:15:41,480 you're like me then and then we would opt out 265 00:15:41,480 --> 00:15:45,230 because we don't serve profiles on minor 266 00:15:45,230 --> 00:15:48,570 so then if if we did follow the data minimisation principal 267 00:15:48,570 --> 00:15:51,779 a affair the fam made from practice principles 268 00:15:51,779 --> 00:15:57,510 we would then a race this profile but we've got another record the next month 269 00:15:57,510 --> 00:16:01,370 but had no data for it was the same Hampton Inn there's her 270 00:16:01,370 --> 00:16:04,950 email address if we hit totally 271 00:16:04,950 --> 00:16:09,290 a race the opt-out profile we would be we have nothing the match against 272 00:16:09,290 --> 00:16:13,800 wouldn't know that Hampton is a minor and we would end up serving the profile 273 00:16:13,800 --> 00:16:15,439 with your email address 274 00:16:15,440 --> 00:16:18,740 however if we kept that we would now 275 00:16:18,740 --> 00:16:22,340 have two records in a profile for a minor 276 00:16:22,340 --> 00:16:26,440 and the new record would snap into that profile because we build a maximally 277 00:16:26,440 --> 00:16:29,900 by now and now we can opt out the entire profile 278 00:16:29,900 --> 00:16:33,260 and now no information on vein Hampton the minor 279 00:16:33,260 --> 00:16:36,750 would actually leak out onto the site so I think 280 00:16:36,750 --> 00:16:39,870 there sometime sometimes there are reasons why 281 00:16:39,870 --> 00:16:44,950 you if you follow the letter from the principal the end up delivering poor 282 00:16:44,950 --> 00:16:45,560 service 283 00:16:45,560 --> 00:16:48,790 and I think that's where up a little 284 00:16:48,790 --> 00:16:52,260 innovation a little thought for a objective 285 00:16:52,260 --> 00:16:55,900 reasoning comes into play so that's a little bit 286 00:16:55,900 --> 00:16:58,920 detail on on the 287 00:16:58,920 --> 00:17:03,630 personal data supplied let's get some altitude on the history of this 288 00:17:03,630 --> 00:17:07,459 I at least how I've been thinking about it for a while 289 00:17:07,459 --> 00:17:11,360 up while ago under fifty years ago or so 290 00:17:11,359 --> 00:17:15,799 house are pretty small I love this Oscar Wilde quote quote quote the only thing 291 00:17:15,799 --> 00:17:19,000 worse than being talked about is not being talked about in this is a a 292 00:17:19,000 --> 00:17:20,390 painting by 293 00:17:20,390 --> 00:17:24,079 Norman Rockwell call the gossip and II spent about a 294 00:17:24,079 --> 00:17:27,099 couple years my you think greensboro alabama 295 00:17:27,099 --> 00:17:30,490 and I think the population now 2500 I think when I was there 296 00:17:30,490 --> 00:17:34,730 i play like 1200 the town is small everybody knew everybody else 297 00:17:34,730 --> 00:17:37,930 and when you cut school the whole town knows by dinner literally 298 00:17:37,930 --> 00:17:41,270 dinner you cut you cut school there 299 00:17:41,270 --> 00:17:45,480 there was the reasonable expectation of privacy which is actually legal term 300 00:17:45,480 --> 00:17:49,210 was very low up a friend 301 00:17:49,210 --> 00:17:54,390 a minor op Monroe I whose was secretary stay in Washington for a lot years 302 00:17:54,390 --> 00:17:57,740 he grew up in Bainbridge Island nada party-line 303 00:17:57,740 --> 00:18:01,450 doubt the harbor the only phone line this is i think in the forties 304 00:18:01,450 --> 00:18:05,740 and everybody was listening always listen to the party line 305 00:18:05,740 --> 00:18:10,210 I telephone from Bainbridge Island there was no expectation of privacy 306 00:18:10,210 --> 00:18:13,690 it was a different different I'm 307 00:18:13,690 --> 00:18:16,920 through the 19th century I 308 00:18:16,920 --> 00:18:20,480 urban populations grew along with anonymity and 309 00:18:20,480 --> 00:18:23,620 you ended up with I arising 310 00:18:23,620 --> 00:18:27,020 a expectation of privacy something try to capture 311 00:18:27,020 --> 00:18:30,940 II in this order notional diagram where 312 00:18:30,940 --> 00:18:34,200 you had this order rockwell ara where urban density furlough 313 00:18:34,200 --> 00:18:36,250 personal and intimate with lower than you had this with the industrial 314 00:18:36,250 --> 00:18:38,450 revolution through the 19th century 315 00:18:38,450 --> 00:18:41,660 you had the sort of March higher and I in 316 00:18:41,660 --> 00:18:45,960 urban density in anonymity I public good fences are and then 317 00:18:45,960 --> 00:18:50,000 around thirty years ago you had this order when I'm gone the privacy vertigo 318 00:18:50,000 --> 00:18:50,809 our 319 00:18:50,809 --> 00:18:54,770 which is where you had on line density sorta drop really fast and personal and 320 00:18:54,770 --> 00:18:56,850 Amy drop really fast 321 00:18:56,850 --> 00:19:01,969 and these big data supplied things are just pull the rug out 322 00:19:01,970 --> 00:19:06,059 and the four out from underneath and everyone is getting their 15 minutes of 323 00:19:06,059 --> 00:19:06,800 fame 324 00:19:06,800 --> 00:19:11,480 and others are just getting completely freaked out by this level of exposure 325 00:19:11,480 --> 00:19:14,690 I there are incredible 326 00:19:14,690 --> 00:19:17,860 advances around personalized I 327 00:19:17,860 --> 00:19:22,580 medicines and communications and energy great things coming out of this 328 00:19:22,580 --> 00:19:27,330 I white one thing that I I thought is really interesting how 329 00:19:27,330 --> 00:19:30,350 I'm in space is shrinking here I 330 00:19:30,350 --> 00:19:33,699 I grew up in Miami Florida I have been 331 00:19:33,700 --> 00:19:37,190 back I'm live there hi 30 five years 332 00:19:37,190 --> 00:19:40,830 I get a facebook friend request from a buddy from high 333 00:19:40,830 --> 00:19:44,929 Adam and I'm this is a few years ago 334 00:19:44,930 --> 00:19:49,880 and so across 35 years and three thousand miles I get this friend request 335 00:19:49,880 --> 00:19:53,690 and use a good body I kinda roughneck 336 00:19:53,690 --> 00:19:56,870 I certain parts mind here are less savory 337 00:19:56,870 --> 00:20:01,000 than others and I really I my kids on their 338 00:20:01,000 --> 00:20:04,280 on the side my family all the pictures like 339 00:20:04,280 --> 00:20:07,950 really accept this friend request from this guy 340 00:20:07,950 --> 00:20:12,190 so what what what you think I did a background check 341 00:20:12,190 --> 00:20:15,940 I as I could I and he had prior 342 00:20:15,940 --> 00:20:19,660 actually but nothing violent I think 343 00:20:19,660 --> 00:20:23,010 hi so I have II said 344 00:20:23,010 --> 00:20:27,450 okay I'll get a rough time in the nineties he got married had a kid 345 00:20:27,450 --> 00:20:31,170 is act-up a little drug trafficking okay 346 00:20:31,170 --> 00:20:34,740 is the nineties I so I accept his friendship 347 00:20:34,740 --> 00:20:38,120 and it's been fun and and you know 348 00:20:38,120 --> 00:20:41,428 no ill II no cautionary tales to tell 349 00:20:41,429 --> 00:20:45,260 up but I think this level love 350 00:20:45,260 --> 00:20:48,520 information it's all public record public information 351 00:20:48,520 --> 00:20:51,850 and let me make an empathetic decision about 352 00:20:51,850 --> 00:20:55,209 someone's life and whether to include them in my 353 00:20:55,210 --> 00:20:58,630 up and I think that's it it's a new world 354 00:20:58,630 --> 00:21:01,840 there there is no there's no physical proximity there I 355 00:21:01,840 --> 00:21:05,928 return temporal proximity I so it's it's fascinating and 356 00:21:05,929 --> 00:21:09,840 I and when you look at privacy from the contact the culture and history really 357 00:21:09,840 --> 00:21:10,790 interesting and 358 00:21:10,790 --> 00:21:15,080 want it things I i it's interesting to look at is sorted this 359 00:21:15,080 --> 00:21:18,320 EU verses US privacy model 360 00:21:18,320 --> 00:21:21,899 it's deeply cultural the United States we value 361 00:21:21,900 --> 00:21:25,410 disclosure above discretion I values 362 00:21:25,410 --> 00:21:28,820 %uh speech and Prasanna sample really value 363 00:21:28,820 --> 00:21:32,419 disclosure where your opinion really value 364 00:21:32,420 --> 00:21:37,390 discretion more than disclosure they value privacy and due process and equal 365 00:21:37,390 --> 00:21:38,440 protection 366 00:21:38,440 --> 00:21:41,710 not that we'd I'll it's an issue balance 367 00:21:41,710 --> 00:21:45,290 I you know the EU views asses as as somewhat 368 00:21:45,290 --> 00:21:49,629 reckless and we you them a somewhat conservative and stodgy its 369 00:21:49,630 --> 00:21:53,290 my view that George Carlin quotes never know is there one driving faster than 370 00:21:53,290 --> 00:21:56,428 you is a maniac anyone driving slower than you is an idiot 371 00:21:56,429 --> 00:22:00,500 up I think that's what the in the Wii the EViews ice 372 00:22:00,500 --> 00:22:03,840 as maniacs and we view them as idiots and 373 00:22:03,840 --> 00:22:07,209 and I don't think anybody's right here I think it is where you sit 374 00:22:07,210 --> 00:22:11,290 where you stand everyone is not like you are doesn't think like you 375 00:22:11,290 --> 00:22:14,870 is wrong in some sense and I think that's true at all I think it's as we're 376 00:22:14,870 --> 00:22:16,739 just grappling with these 377 00:22:16,740 --> 00:22:20,040 these values that our intention and they come from 378 00:22:20,040 --> 00:22:23,070 an interesting background up 379 00:22:23,070 --> 00:22:26,629 the EU has in viable right dignity 380 00:22:26,630 --> 00:22:29,890 this write a personality I 381 00:22:29,890 --> 00:22:33,390 that based dates back in the nineteen 1902 382 00:22:33,390 --> 00:22:36,940 really solidified I after world war two 383 00:22:36,940 --> 00:22:40,700 and the Nazi abuses up the right a personality 384 00:22:40,700 --> 00:22:44,480 a really I came to embrace privacy rights 385 00:22:44,480 --> 00:22:48,140 in the US in 1890 Bryant a Brandeis Warren 386 00:22:48,140 --> 00:22:53,130 I'd created their in from this I right to be let alone doctrine 387 00:22:53,130 --> 00:22:58,110 I and so that really is all-encompassing privacy 388 00:22:58,110 --> 00:23:02,469 doctrine whereas in the US we're really about harm 389 00:23:02,470 --> 00:23:07,150 I especially with respect to I individuals and corporations everything 390 00:23:07,150 --> 00:23:07,980 except 391 00:23:07,980 --> 00:23:11,360 government and citizen which obviously have Fourth Amendment privacy 392 00:23:11,360 --> 00:23:12,250 protections 393 00:23:12,250 --> 00:23:16,130 between people and and and corporations 394 00:23:16,130 --> 00:23:20,350 it's really about harmon in 1960 up 395 00:23:20,350 --> 00:23:23,780 a a Prosser put out these these the sports 396 00:23:23,780 --> 00:23:26,889 and this forum I it's pretty straightforward 397 00:23:26,890 --> 00:23:31,120 there's the peep in court don't look at a private information are private 398 00:23:31,120 --> 00:23:34,250 I don't filing with private space don't assume 399 00:23:34,250 --> 00:23:38,160 anybody's identity don't blab 400 00:23:38,160 --> 00:23:43,860 don't blab private back I and don't flatter 401 00:23:43,860 --> 00:23:48,760 the United States privacy regime is pretty much built on those for torts 402 00:23:48,760 --> 00:23:53,170 and there's a set of sectoral privacy law that goes with those ports 403 00:23:53,170 --> 00:23:56,580 I things around like there's health care privacy 404 00:23:56,580 --> 00:24:00,720 and there is financial privacy and there's employment privacy 405 00:24:00,720 --> 00:24:03,980 there's no overarching privacy 406 00:24:03,980 --> 00:24:07,740 which really makes it difficult United States to figure out when this Privacy 407 00:24:07,740 --> 00:24:08,610 matter and wanted 408 00:24:08,610 --> 00:24:11,649 when doesn't it especially when you're designing products 409 00:24:11,650 --> 00:24:15,510 and so I came up with this thing I got on actors think 410 00:24:15,510 --> 00:24:19,400 I when I was an Intel is my guys always calm with these great ideas 411 00:24:19,400 --> 00:24:23,679 they thought were great ideas how about we let people search by picture 412 00:24:23,679 --> 00:24:27,350 you take a picture somebody and then you can really see their back on track 413 00:24:27,350 --> 00:24:31,000 and I'm like oh god awful I 414 00:24:31,000 --> 00:24:34,340 you know be all over the press and not a good way 415 00:24:34,340 --> 00:24:38,379 I but what I want to do is figure out how to tease apart 416 00:24:38,380 --> 00:24:42,420 so that this becomes actionable and 417 00:24:42,420 --> 00:24:45,929 there's sorta 3 axes that that this 418 00:24:45,929 --> 00:24:49,400 the privacy turns on I at least 419 00:24:49,400 --> 00:24:52,720 I for this formulate a formulation its 420 00:24:52,720 --> 00:24:56,370 places players and parapsychology the places are 421 00:24:56,370 --> 00:24:59,959 is a private place for a public place think that's an important access 422 00:24:59,960 --> 00:25:03,670 I there's a a story 423 00:25:03,670 --> 00:25:07,160 %uh me leaving America's game a couple years back in we're going up the ramp in 424 00:25:07,160 --> 00:25:09,200 this girl is like taking pictures 425 00:25:09,200 --> 00:25:13,610 everybody coming up the ramp and I was with somebody cuz I'm a privacy guy some 426 00:25:13,610 --> 00:25:14,270 people say 427 00:25:14,270 --> 00:25:17,559 Wiley prime and I said really 428 00:25:17,559 --> 00:25:20,889 I'm in a public place I can expect that picture would be 429 00:25:20,890 --> 00:25:24,500 I there be facial recognition %uh 430 00:25:24,500 --> 00:25:28,820 me and will be uploaded on Facebook everyone known that the game 431 00:25:28,820 --> 00:25:32,270 well how differences that from 150 years ago when they might have 432 00:25:32,270 --> 00:25:36,030 the time I decided the game not a whole lot of different 433 00:25:36,030 --> 00:25:39,559 I so it's important understand when you're in public and when your private 434 00:25:39,559 --> 00:25:42,299 sometimes it's hard to take part it out with a look at in a minute 435 00:25:42,299 --> 00:25:45,480 the second accesses players are they I these powerful players 436 00:25:45,480 --> 00:25:48,650 or is there a big power difference between the players 437 00:25:48,650 --> 00:25:52,510 like a like a citizen and a government or is it are they peers 438 00:25:52,510 --> 00:25:56,870 fundamentally what can go wrong I think that is a big issue 439 00:25:56,870 --> 00:26:00,809 what are the perils the players 440 00:26:00,809 --> 00:26:04,190 I disparity actually Bruce Schneier came up with that 441 00:26:04,190 --> 00:26:07,980 a few years ago I that privacy is really about 442 00:26:07,980 --> 00:26:11,130 a power equalizing power I think that that 443 00:26:11,130 --> 00:26:15,390 was a brilliant insight looking this a couple examples around this if you look 444 00:26:15,390 --> 00:26:15,670 at 445 00:26:15,670 --> 00:26:19,690 the places pay sometimes there in a public place when you think you're in a 446 00:26:19,690 --> 00:26:21,290 private place through the nineteen 447 00:26:21,290 --> 00:26:24,780 I 2009 study by MIT 448 00:26:24,780 --> 00:26:27,960 gaidar project where they were able to I 449 00:26:27,960 --> 00:26:32,970 discern someone sexual orientation by just their friends on Facebook 450 00:26:32,970 --> 00:26:36,630 even if their privacy saying a clampdown and how able friend who's a professor 451 00:26:36,630 --> 00:26:37,320 there 452 00:26:37,320 --> 00:26:40,639 hates no point in having controlled information is all out there anyway 453 00:26:40,640 --> 00:26:44,780 I and I think he makes appoint I 454 00:26:44,780 --> 00:26:48,080 the RP there are things that people are going to derive I 455 00:26:48,080 --> 00:26:51,419 from public data that is private 456 00:26:51,420 --> 00:26:54,870 and I think that is it really important I 457 00:26:54,870 --> 00:26:58,610 issue that we're going to talk about it a little bit 458 00:26:58,610 --> 00:27:01,799 here's one that I i really I really 459 00:27:01,799 --> 00:27:05,929 behavioral targeting me prolly and we have 460 00:27:05,929 --> 00:27:09,890 decide I 461 00:27:09,890 --> 00:27:14,100 if you look at up social media 462 00:27:14,100 --> 00:27:17,570 services on the web what we're paying with our money 463 00:27:17,570 --> 00:27:23,030 or paying with their data I I did this post a couple years ago called 464 00:27:23,030 --> 00:27:27,660 your privacy is worth five bucks a month and the way I came up with five bucks a 465 00:27:27,660 --> 00:27:28,410 month 466 00:27:28,410 --> 00:27:32,000 Facebook market cap with the user base is what the investor expectation on 467 00:27:32,000 --> 00:27:34,850 return would be and I kinda backed up by bottom I'm 468 00:27:34,850 --> 00:27:38,168 it's actually much less than that because market valuing 469 00:27:38,169 --> 00:27:42,320 Facebook's revenue quite a bit higher I maybe a pokémon 470 00:27:42,320 --> 00:27:45,389 I by the fact is I 471 00:27:45,390 --> 00:27:49,890 you can pay with your dating to pay with your money and I proposed at that time 472 00:27:49,890 --> 00:27:54,120 I two-and-a-half years ago the Facebook should have a tracker charge of 473 00:27:54,120 --> 00:27:57,790 you either pay them there but you're you pay them a pokémon 474 00:27:57,790 --> 00:28:00,918 or they can use your data however they like 475 00:28:00,919 --> 00:28:05,590 I someone gotta pay for the service is no free lunch here 476 00:28:05,590 --> 00:28:09,860 and actually based on winner came up with a similar 477 00:28:09,860 --> 00:28:13,120 propose a few months back I do think that the end the day we gotta figure 478 00:28:13,120 --> 00:28:15,159 have charged for these products 479 00:28:15,160 --> 00:28:19,600 I sure we love the service by if we're not careful 480 00:28:19,600 --> 00:28:23,428 we're gonna be on the menu 481 00:28:23,429 --> 00:28:26,549 talk about power a little bit I would add to this low bat I 482 00:28:26,549 --> 00:28:30,290 this is this notion of graph privacy rights for our disparity 483 00:28:30,290 --> 00:28:34,760 so peer to peer there's not much privacy they're not not have any privacy rights 484 00:28:34,760 --> 00:28:38,210 incorporation into and a customer and employee there are certainly 485 00:28:38,210 --> 00:28:43,850 statutory protections second five the FTC Act says a corporation's 486 00:28:43,850 --> 00:28:47,820 certainly I can be unfair unfairly or deceive 487 00:28:47,820 --> 00:28:51,189 the Fair Credit Reporting Act says that there's certain 488 00:28:51,190 --> 00:28:54,620 procedure the company must go through when they're doing a background check on 489 00:28:54,620 --> 00:28:58,639 a prospective employee I between government and citizen 490 00:28:58,640 --> 00:29:02,200 I there's constitutional protection 491 00:29:02,200 --> 00:29:06,200 between you and your God 9 492 00:29:06,200 --> 00:29:11,010 up between an example of this was like Target 493 00:29:11,010 --> 00:29:14,040 remember the target anybody remember the target now you're pregnant 494 00:29:14,040 --> 00:29:19,340 I this is pretty interesting it's a very simple machine learning problem 495 00:29:19,340 --> 00:29:23,809 because up women had a %uh baby registries 496 00:29:23,809 --> 00:29:27,820 so they would say hey I'm pregnant I and people to register to get the baby stuff 497 00:29:27,820 --> 00:29:29,928 they would see what those customers 498 00:29:29,929 --> 00:29:33,110 who were registered I for Babies R 499 00:29:33,110 --> 00:29:37,010 what they bought so they wanted had buildup in training data 500 00:29:37,010 --> 00:29:40,990 around what pregnant women purchase and then they extrapolated that to the 501 00:29:40,990 --> 00:29:42,890 restive they're 502 00:29:42,890 --> 00:29:46,750 their their customer base I 503 00:29:46,750 --> 00:29:50,230 the challenges they really think about power applications about because there 504 00:29:50,230 --> 00:29:50,980 was a 505 00:29:50,980 --> 00:29:55,070 I Teen Mom you start to get coupons in the mail 506 00:29:55,070 --> 00:29:58,780 saying hey we know you're pregnant combine unscented lotion 507 00:29:58,780 --> 00:30:03,250 server now betting and I and the Father 508 00:30:03,250 --> 00:30:06,260 this teen call park inside 509 00:30:06,260 --> 00:30:11,280 is not pregnant until the dotted 510 00:30:11,280 --> 00:30:15,260 and that was that there's a huge there's a power disparity between parents and 511 00:30:15,260 --> 00:30:16,580 children 512 00:30:16,580 --> 00:30:20,149 I and that were really was a respected by Park 513 00:30:20,150 --> 00:30:23,559 and and that only one 514 00:30:23,559 --> 00:30:27,809 abuse of power consider that if an employee employer new 515 00:30:27,809 --> 00:30:31,639 headed time that their employees with pride 516 00:30:31,640 --> 00:30:35,440 might they not give them I certain assignments 517 00:30:35,440 --> 00:30:39,290 I might be not promote I insurance 518 00:30:39,290 --> 00:30:43,330 I might be delivered hand more expensive great 519 00:30:43,330 --> 00:30:47,620 so people in powerful positions and exploit 520 00:30:47,620 --> 00:30:51,719 II my can exploit this information 521 00:30:51,720 --> 00:30:56,049 and is only one up to equation back 522 00:30:56,049 --> 00:30:59,780 by I there is a a correlation between 523 00:30:59,780 --> 00:31:05,470 chain idea the difference and in power between players and the secrets the 524 00:31:05,470 --> 00:31:09,440 and actually uses when I look at teams I and whether they 525 00:31:09,440 --> 00:31:12,780 how much they trust each other and trust me your trust their leaders 526 00:31:12,780 --> 00:31:16,299 I very if they have a lot of secrets there now are trust 527 00:31:16,299 --> 00:31:19,530 and you can use this as a signal I get the sense there's a %uh 528 00:31:19,530 --> 00:31:23,740 secrets running around that the team that means is not a lot of trust and 529 00:31:23,740 --> 00:31:27,350 and that's not a good thing for prop productivity team 530 00:31:27,350 --> 00:31:30,689 I wrote this post while back why employees why 531 00:31:30,690 --> 00:31:34,840 name they lied because as the boss of all the boss is in the position I love 532 00:31:34,840 --> 00:31:35,820 some power 533 00:31:35,820 --> 00:31:39,559 if they can engender trust with the team the team is going to do whatever they 534 00:31:39,559 --> 00:31:41,620 can typically live 535 00:31:41,620 --> 00:31:46,199 because lie I levels the playing the PowerPoint feel 536 00:31:46,200 --> 00:31:50,309 that's what they happen and there's no there's no you know it's a man that 537 00:31:50,309 --> 00:31:54,870 that's what weaker players need to do that to level things out 538 00:31:54,870 --> 00:31:58,399 putting this all together this place is players pareles 539 00:31:58,400 --> 00:32:03,169 cases up if you look on the vertical more powerful players on the vertical 540 00:32:03,169 --> 00:32:06,350 more private places on the horizontal and 541 00:32:06,350 --> 00:32:09,480 I on the on the right are are are 542 00:32:09,480 --> 00:32:13,030 on the you're on the right or the most private up places and on the laughter 543 00:32:13,030 --> 00:32:16,250 the more public place and powerful players on the top and weaker players on 544 00:32:16,250 --> 00:32:17,090 the bottom 545 00:32:17,090 --> 00:32:20,500 and then you could do things like take 546 00:32:20,500 --> 00:32:25,039 paid cases like soap in the press and plopped a monograph like this 547 00:32:25,039 --> 00:32:29,549 and the the sphere the side the spheres how perilous 548 00:32:29,549 --> 00:32:33,389 I the outcome could be so I took a few 549 00:32:33,390 --> 00:32:37,419 actually I sorta collect fees as a hobby is only like have higher 550 00:32:37,419 --> 00:32:40,440 fifty or sixty other by would make their 551 00:32:40,440 --> 00:32:43,960 the chart even noisier than it is i right now so let's walk through this a 552 00:32:43,960 --> 00:32:44,980 couple 553 00:32:44,980 --> 00:32:48,220 up so there was a case where the US supported 554 00:32:48,220 --> 00:32:52,429 a buncha breadth for because they tweeted that they're going to destroy 555 00:32:52,429 --> 00:32:53,570 America 556 00:32:53,570 --> 00:32:57,120 and really what the man was just gonna party really hard and so the TSA saw 557 00:32:57,120 --> 00:32:57,729 that 558 00:32:57,730 --> 00:33:01,919 and the board I and and so 559 00:33:01,919 --> 00:33:06,520 certainly public data but in the a very powerful player 560 00:33:06,520 --> 00:33:10,320 can really exploit that and the point deportation 561 00:33:10,320 --> 00:33:14,299 is a major deal I mean incarceration execution deportation 562 00:33:14,299 --> 00:33:17,870 those are things that only governments can do up 563 00:33:17,870 --> 00:33:22,580 there was a case %uh actually pain in Georgia where'd she posted pictures 564 00:33:22,580 --> 00:33:26,350 a a of her vacation for a fuse and Ireland 565 00:33:26,350 --> 00:33:32,049 drinking something I she shared those on Facebook and somehow the 566 00:33:32,049 --> 00:33:35,320 school board got when those pictures and the 567 00:33:35,320 --> 00:33:39,909 her for drinking I and I 568 00:33:39,910 --> 00:33:43,760 this is anything case because Ashley Payne said I felt violated like someone 569 00:33:43,760 --> 00:33:46,490 stole the pictures at my house 570 00:33:46,490 --> 00:33:50,070 I and I think she got that far wrong it's not that 571 00:33:50,070 --> 00:33:55,330 no one stole the pictures share the pictures and and sharing is this store 572 00:33:55,330 --> 00:33:59,120 gateway drug between private and public its order the slippery slope is there 573 00:33:59,120 --> 00:34:00,510 something in the office and 574 00:34:00,510 --> 00:34:03,950 out there the abuse happened 575 00:34:03,950 --> 00:34:08,069 by the school board muster contract said you cannot drink at all 576 00:34:08,069 --> 00:34:11,668 while you're under contract whether it's on our time or yours 577 00:34:11,668 --> 00:34:16,259 the school board abuse that information and I know she had filed suit on 578 00:34:16,260 --> 00:34:20,770 Havasu came out but it's clear that was a clear abusive information 579 00:34:20,770 --> 00:34:24,060 not inappropriate accessible information 580 00:34:24,060 --> 00:34:29,859 up I like this one too I the FBI 581 00:34:29,859 --> 00:34:34,060 put a GPS transmitter on and fun 582 00:34:34,060 --> 00:34:37,090 car you drug kingpin in maryland 583 00:34:37,090 --> 00:34:40,409 I hadam that right he was a drug 584 00:34:40,409 --> 00:34:43,530 and he was convicted and by 585 00:34:43,530 --> 00:34:48,230 the FBI put the GPS transmitter on his car without a search warrant 586 00:34:48,230 --> 00:34:52,159 tell them from and obviously all that information 587 00:34:52,159 --> 00:34:56,480 clearly showed he was guilty the supreme want this case when the Supreme Court 588 00:34:56,480 --> 00:34:59,640 this this supreme court voted nine to nothing 589 00:34:59,640 --> 00:35:02,790 in Aug 09 nothing anything that 590 00:35:02,790 --> 00:35:05,950 that this was inappropriate search violation 591 00:35:05,950 --> 00:35:10,089 %uh and punt on 4th amendment vacated vacated the conviction 592 00:35:10,090 --> 00:35:13,460 I clearly a very proud the 593 00:35:13,460 --> 00:35:17,180 clearly a private place but before this is 594 00:35:17,180 --> 00:35:21,000 this case not so clearly because the your house is clearly private but is 595 00:35:21,000 --> 00:35:22,140 your car private 596 00:35:22,140 --> 00:35:27,000 and the judges rule that your car actually is a private place 597 00:35:27,000 --> 00:35:31,060 which I think was a a victory for privacy advocates 598 00:35:31,060 --> 00:35:34,670 I did something on the up the pier side 599 00:35:34,670 --> 00:35:38,849 love the spectrum I if you member Tyler Clementi 600 00:35:38,849 --> 00:35:41,960 I a web cam 601 00:35:41,960 --> 00:35:45,260 captured I'm having sacked II and 602 00:35:45,260 --> 00:35:48,490 and suicide public same 603 00:35:48,490 --> 00:35:52,330 and so I its isn't just 604 00:35:52,330 --> 00:35:56,020 large our disparities that and in Para without sometimes 605 00:35:56,020 --> 00:35:59,869 violation of privacy even by peers and in parallel outcomes 606 00:35:59,869 --> 00:36:02,970 so its important got understand 607 00:36:02,970 --> 00:36:07,368 I this this order framework and figure out 608 00:36:07,369 --> 00:36:11,770 if you're designing products certainly where 609 00:36:11,770 --> 00:36:15,190 I'll where does your product I lawn on 610 00:36:15,190 --> 00:36:18,490 and if you're doing it now what's important who are the players 611 00:36:18,490 --> 00:36:21,799 yeah what is the space for your gathering the data and what potentially 612 00:36:21,800 --> 00:36:22,210 could go 613 00:36:22,210 --> 00:36:26,119 wrong so now I i switch to a way 614 00:36:26,119 --> 00:36:30,660 I from policy for a second and talk about technology a little bit 615 00:36:30,660 --> 00:36:35,970 I there's a lot to date out there 616 00:36:35,970 --> 00:36:39,049 this is a %uh study I 617 00:36:39,050 --> 00:36:42,220 fourth track cockeyed 618 00:36:42,220 --> 00:36:45,509 earlier this year I actually did last fall 619 00:36:45,510 --> 00:36:48,690 and I wanted to 620 00:36:48,690 --> 00:36:52,760 or understand what if we want to take the capacity of the human brain 621 00:36:52,760 --> 00:36:56,020 rented at Amazon what would that cost 622 00:36:56,020 --> 00:36:59,030 and so Chris westbury universe Alberta 623 00:36:59,030 --> 00:37:02,590 a prop third gas and that 624 00:37:02,590 --> 00:37:07,599 human brain has 20,000 teraflops compute capacity and 625 00:37:07,599 --> 00:37:11,190 2.5 had a bite storage in in 626 00:37:11,190 --> 00:37:15,810 estimate I O network but if you're just the by that Amazon will cost you twenty 627 00:37:15,810 --> 00:37:17,230 seven million dollars a month 628 00:37:17,230 --> 00:37:22,950 today however the good news is in 20 years at a cost you less and 629 00:37:22,950 --> 00:37:27,779 seven thousand dollars a year about fifty three thousand dollars 630 00:37:27,780 --> 00:37:31,680 which in the reason I kinda did this analysis 631 00:37:31,680 --> 00:37:35,098 was we're doing all these inferences 632 00:37:35,099 --> 00:37:39,930 outside over on brain I were decide we're making inferences 633 00:37:39,930 --> 00:37:43,210 outside have being 634 00:37:43,210 --> 00:37:47,359 in the same room as someone I for example some comes in an interview for a 635 00:37:47,359 --> 00:37:48,040 job 636 00:37:48,040 --> 00:37:51,970 you know you can guess their age guess it and there you can guess the race 637 00:37:51,970 --> 00:37:55,450 I see your making your you're thinking all this stuff 638 00:37:55,450 --> 00:37:59,000 and use it that's okay course you know you know what if they're showing you 639 00:37:59,000 --> 00:38:00,420 know they're pregnant 640 00:38:00,420 --> 00:38:04,410 I if you act on those sorts of things you have a problem 641 00:38:04,410 --> 00:38:08,420 and so now the next question is well if a big data engine somewhere is making 642 00:38:08,420 --> 00:38:09,869 some inferences 643 00:38:09,869 --> 00:38:14,390 should that be allowed even if those inferences are like way creepy 644 00:38:14,390 --> 00:38:19,098 and very objectionable should that big data engine still be able to 645 00:38:19,099 --> 00:38:23,900 the ride those inferences so this was motivated by 646 00:38:23,900 --> 00:38:27,140 reminding a conga line the I 647 00:38:27,140 --> 00:38:30,868 he's the Hadapsar Chatwing and and we were doing I was on a panel with some 648 00:38:30,869 --> 00:38:32,330 privacy advocates and 649 00:38:32,330 --> 00:38:35,670 we're talking about the riving up public 650 00:38:35,670 --> 00:38:38,800 I or private back from public data and Daniel 651 00:38:38,800 --> 00:38:43,220 raise Haney said you now if you can derive private fat from public that it 652 00:38:43,220 --> 00:38:45,040 sounds like you guys are trying to decide 653 00:38:45,040 --> 00:38:49,859 fine and you category of thought crime not just seems wrong 654 00:38:49,859 --> 00:38:53,560 and I was really good question the prime I we'll look at ourselves 655 00:38:53,560 --> 00:38:57,130 like yeah that's right question I gotta go think about 656 00:38:57,130 --> 00:39:00,570 and so I did a and so I started I went back to the source 657 00:39:00,570 --> 00:39:04,109 thought prime George Orwell and said the promise thought crime 658 00:39:04,109 --> 00:39:07,330 and the reason if feels up bad it's because 659 00:39:07,330 --> 00:39:12,200 all everything else leave from thought and point six hundred years ago louts 660 00:39:12,200 --> 00:39:13,589 who said the same thing 661 00:39:13,590 --> 00:39:16,730 that's become words words become action back to become habits habits become 662 00:39:16,730 --> 00:39:17,800 character 663 00:39:17,800 --> 00:39:20,760 your character becomes your destiny everything starts with aw if you could 664 00:39:20,760 --> 00:39:21,230 stop 665 00:39:21,230 --> 00:39:24,520 why you can stop all these have things might stop 666 00:39:24,520 --> 00:39:27,960 good things up 667 00:39:27,960 --> 00:39:31,230 so if you look at the thought 668 00:39:31,230 --> 00:39:35,170 and speech and access we know that actions of in our criminal 669 00:39:35,170 --> 00:39:38,530 sometimes speeches actual criminal I in the case 670 00:39:38,530 --> 00:39:42,010 libel or slander like we talked about with process sports 671 00:39:42,010 --> 00:39:45,260 conflict with a legitimate interest I think it snowed in 672 00:39:45,260 --> 00:39:49,900 is having trouble with publishing I think he's in a published 673 00:39:49,900 --> 00:39:53,020 I but got down we are not crime 674 00:39:53,020 --> 00:39:57,330 except in the case really weird cabron case conspiracy 675 00:39:57,330 --> 00:40:02,180 theory or conspiracy a crime and using that comes outta 676 00:40:02,180 --> 00:40:05,950 societies sorry existential response to it rat 677 00:40:05,950 --> 00:40:10,470 that is for to be on NEC conspiracy crimes around 678 00:40:10,470 --> 00:40:13,910 organized crime and terrorism typically 679 00:40:13,910 --> 00:40:18,330 but I anything out are every other time 680 00:40:18,330 --> 00:40:23,529 conspiracy is not a crime so this is for the ideology around a stock prime 681 00:40:23,530 --> 00:40:27,940 up a real crime or not and big data engines do inferences 682 00:40:27,940 --> 00:40:31,400 that the conclusion is yes of course big day 683 00:40:31,400 --> 00:40:34,400 can do can perform I 684 00:40:34,400 --> 00:40:37,780 so I so that was the theory of it so then I decided to 685 00:40:37,780 --> 00:40:42,790 really test this theory and if I design a felon classifier 686 00:40:42,790 --> 00:40:46,660 because we had all this great criminal dated lesi said 687 00:40:46,660 --> 00:40:50,460 well let's how to put on this little bit and say well okay thought isn't a crime 688 00:40:50,460 --> 00:40:54,080 what if I try to predict third or figure out if someone's a felon or not based on 689 00:40:54,080 --> 00:40:57,880 from there so is actually a piece in today's Bloomberg 690 00:40:57,880 --> 00:41:02,050 news on this town class are there is the link I the talks to this 691 00:41:02,050 --> 00:41:05,880 in in some detail by I'll go through it a little bit with you 692 00:41:05,880 --> 00:41:08,930 I I can tell you is we have criminal data for everybody 693 00:41:08,930 --> 00:41:12,509 United States 150 million defended I 694 00:41:12,510 --> 00:41:16,400 which if you think about that number for many you know that's gotta be wrong 695 00:41:16,400 --> 00:41:19,970 not $250 million people with a criminal record United States 696 00:41:19,970 --> 00:41:23,890 there's a lot of duplication that they usually 697 00:41:23,890 --> 00:41:27,220 that they describe down to about forty million dependence 698 00:41:27,220 --> 00:41:30,419 in the United States and then what we did 699 00:41:30,420 --> 00:41:34,550 during this exercises week Wednesday we weaklings the data we link today and we 700 00:41:34,550 --> 00:41:35,680 sampled state 701 00:41:35,680 --> 00:41:39,460 Kentucky not to become Kentucky by it was this there was this 702 00:41:39,460 --> 00:41:42,599 it was a small enough data so that we can release it 703 00:41:42,599 --> 00:41:46,849 this state is up now get a I under I know my known as the Big Data engine 704 00:41:46,849 --> 00:41:47,230 that 705 00:41:47,230 --> 00:41:50,270 that hours and Talia so you gotta get having I got 706 00:41:50,270 --> 00:41:54,470 well down the data up and we to find some features 707 00:41:54,470 --> 00:41:58,209 and train the learner and and so here's how the 708 00:41:58,210 --> 00:42:01,440 the learner work is an alternating decision tree 709 00:42:01,440 --> 00:42:05,750 I here's a few of the features gender I colored how many tattoos 710 00:42:05,750 --> 00:42:09,300 up whether I'm a criminal offenses they had with 711 00:42:09,300 --> 00:42:12,750 I they weren't so as you walk down the stairs to records here 712 00:42:12,750 --> 00:42:16,660 one male one female and this is all trained on this Kentucky data 713 00:42:16,660 --> 00:42:20,580 and so I if you're female with hazel eyes 714 00:42:20,580 --> 00:42:23,950 and I to more or more tattoos and have 715 00:42:23,950 --> 00:42:27,118 for your misdemeanors you start adding up these 716 00:42:27,119 --> 00:42:30,180 score three feature and you end up with 4.4 717 00:42:30,180 --> 00:42:34,690 and its above a threshold 3.5 with with makes a binary decision 718 00:42:34,690 --> 00:42:38,619 whether you're found or not the top record I 719 00:42:38,619 --> 00:42:42,700 doesn't have a high enough ordered to be classified as a now I'm 720 00:42:42,700 --> 00:42:46,689 so so you can actually check this out I have 721 00:42:46,690 --> 00:42:50,290 widget actually I might on my blog where you can actually go 722 00:42:50,290 --> 00:42:54,190 up putting you can actually answer the questions at all decide whether you look 723 00:42:54,190 --> 00:42:55,240 like 724 00:42:55,240 --> 00:42:58,348 early like a pallet you would think 725 00:42:58,349 --> 00:43:01,480 so up so 726 00:43:01,480 --> 00:43:04,990 at some point the technology ends and the policy begins 727 00:43:04,990 --> 00:43:08,700 and this is why I wanted to how to take that departed for technology then come 728 00:43:08,700 --> 00:43:09,299 back to 729 00:43:09,300 --> 00:43:12,950 the policy so this is a a a false negative 730 00:43:12,950 --> 00:43:16,210 fall versus false-positive her that I 731 00:43:16,210 --> 00:43:19,570 defines the performance at the the classifier 732 00:43:19,570 --> 00:43:24,400 and so that line is that the performance of the classifier based on what wrestle 733 00:43:24,400 --> 00:43:27,810 you want to choose and the best classifier 734 00:43:27,810 --> 00:43:31,400 would be that line would go through the origin is that where you would have no 735 00:43:31,400 --> 00:43:33,080 false positives 736 00:43:33,080 --> 00:43:36,270 you have no false negatives are evidently and 737 00:43:36,270 --> 00:43:40,050 obviously our classifier stock much were 738 00:43:40,050 --> 00:43:45,080 in the ideal classifier but if you look at the extreme fear 739 00:43:45,080 --> 00:43:48,109 they tell you lock because you have to pick a threshold 740 00:43:48,109 --> 00:43:51,650 in the previous why we picked the pressure 3.5 741 00:43:51,650 --> 00:43:56,890 up in this country your you're innocent until proven guilty so you want 742 00:43:56,890 --> 00:44:00,980 the class started not be that aggressive because if you're a very aggressive 743 00:44:00,980 --> 00:44:04,570 aggressive at all you start to bring in a classified people as 744 00:44:04,570 --> 00:44:08,599 sounds that are not found for example if you look at the far right there 745 00:44:08,599 --> 00:44:11,770 the the false positive rate is nineteen percent 746 00:44:11,770 --> 00:44:15,920 yeah I don't miss any Fallon's your false negative rate is zero 747 00:44:15,920 --> 00:44:19,930 by nineteen percent the PV classifies are not now 748 00:44:19,930 --> 00:44:23,618 not cool although if you're hiring its cool 749 00:44:23,619 --> 00:44:27,260 so the resort is anarchy Terry trade-off 750 00:44:27,260 --> 00:44:32,099 you doing a policy determination on where the threshold 751 00:44:32,099 --> 00:44:36,470 be if you if you wanna going for the anarchy side 752 00:44:36,470 --> 00:44:40,399 I I think the most lenient I 753 00:44:40,400 --> 00:44:44,060 spot on this her is ninety-one percent false negatives 754 00:44:44,060 --> 00:44:47,480 the only catch nine-percent the actual Fallon's 755 00:44:47,480 --> 00:44:52,060 but you don't miss identify anybody so it's Thunder Down right no one is about 756 00:44:52,060 --> 00:44:57,020 I and and so policymakers have to choose where you want to be 757 00:44:57,020 --> 00:45:00,280 on this car because this is 758 00:45:00,280 --> 00:45:03,420 effectively and other technologies like it 759 00:45:03,420 --> 00:45:06,760 our computerized informants or computerized 760 00:45:06,760 --> 00:45:11,930 tipster and if you don't couple good policy with 761 00:45:11,930 --> 00:45:15,230 is kinda technology now problem 762 00:45:15,230 --> 00:45:19,440 for placing a problem is in court I because reason suspicion or probable 763 00:45:19,440 --> 00:45:20,330 cause 764 00:45:20,330 --> 00:45:24,680 is usually our mechanism for vetting out that police work 765 00:45:24,680 --> 00:45:29,348 I however States courts have upheld 766 00:45:29,349 --> 00:45:33,980 profile it was the case with USB Sokolow where this guy comes in from 767 00:45:33,980 --> 00:45:37,420 a wide Tampa two days 768 00:45:37,420 --> 00:45:41,180 today ticket paid ass used an assumed name 769 00:45:41,180 --> 00:45:44,879 I they profiled him %ah this guy looks like drug trafficker and sure enough he 770 00:45:44,880 --> 00:45:46,760 was then 771 00:45:46,760 --> 00:45:49,910 the ruling was upheld but 772 00:45:49,910 --> 00:45:54,730 this predictive information had live all alone by itself 773 00:45:54,730 --> 00:45:58,369 ask actually be corroborated with other information 774 00:45:58,369 --> 00:46:01,609 I as the particularized be efficient to get 775 00:46:01,609 --> 00:46:05,759 I think most importantly needs to be detailed enough that it doesn't identify 776 00:46:05,760 --> 00:46:08,869 I everybody identifies 777 00:46:08,869 --> 00:46:13,080 just the target and I think that's important based on 778 00:46:13,080 --> 00:46:16,509 and there's more you can look at more references there 779 00:46:16,510 --> 00:46:19,869 I just on Monday I V I saw the news 780 00:46:19,869 --> 00:46:24,210 i New York City's stop and frisk ruled unconstitutional 781 00:46:24,210 --> 00:46:27,359 they've been doing it for about 10 years I 782 00:46:27,359 --> 00:46:30,690 six hundred percent increase I 783 00:46:30,690 --> 00:46:35,940 and stop and frisk over the last decade doesn't 2011 684 thousand people 784 00:46:35,940 --> 00:46:39,839 we're stopping rest I 785 00:46:39,839 --> 00:46:43,369 so I in this is a a quote from the judge 786 00:46:43,369 --> 00:46:47,790 I and she said the city police the blackness back to be stopped the same 787 00:46:47,790 --> 00:46:48,940 rate as a proportion 788 00:46:48,940 --> 00:46:52,550 the local criminal suspect population 789 00:46:52,550 --> 00:46:56,320 that sound like a baseball problem to me I 790 00:46:56,320 --> 00:46:59,960 and I said okay well let me put on this a little bit and see 791 00:46:59,960 --> 00:47:04,170 and I don't know for those their unfamiliar with 792 00:47:04,170 --> 00:47:07,530 basically says that I can you adjust 793 00:47:07,530 --> 00:47:12,630 a a probability based on a a general statistics to something that 794 00:47:12,630 --> 00:47:16,780 uses prize some even additional information 795 00:47:16,780 --> 00:47:20,130 so in this case we want in our the cop wants to know 796 00:47:20,130 --> 00:47:24,580 given that someone's a minority what's the chance that that we should arrest 797 00:47:24,580 --> 00:47:28,250 that's the art that's the money on on on the left 798 00:47:28,250 --> 00:47:31,500 and the arrest raid is the PSA on the right 799 00:47:31,500 --> 00:47:35,130 and that the the 800 00:47:35,130 --> 00:47:38,890 ratio I in the middle there is 801 00:47:38,890 --> 00:47:42,368 given that someone is arrested 802 00:47:42,369 --> 00:47:46,660 where they might are they a minority and what's the minority population 803 00:47:46,660 --> 00:47:50,089 and so the plug in some numbers very rough numbers and I don't think furtive 804 00:47:50,089 --> 00:47:50,970 movements 805 00:47:50,970 --> 00:47:54,828 with for it movements are I I'm assuming that just based on 806 00:47:54,829 --> 00:47:58,160 race alone what what's going on here 807 00:47:58,160 --> 00:48:01,450 so if you say that New York City assumes that 808 00:48:01,450 --> 00:48:05,328 87 percent I love people that arrested 809 00:48:05,329 --> 00:48:09,329 are minorities and 50 percent as a minority population in New York City 810 00:48:09,329 --> 00:48:09,599 with 811 00:48:09,599 --> 00:48:13,640 boxes pack I and the the general 812 00:48:13,640 --> 00:48:17,118 arrest rate is five percent 813 00:48:17,119 --> 00:48:21,190 so given that you're minority should you be arrested 814 00:48:21,190 --> 00:48:24,810 is about 10 percent which is twice the amount 815 00:48:24,810 --> 00:48:28,920 I love the general population and so if you're on the from the city 816 00:48:28,920 --> 00:48:32,030 or conservative blogger within a few I'm saying 817 00:48:32,030 --> 00:48:35,829 it is a hundred percent greater chance that a minority stop is justified 818 00:48:35,829 --> 00:48:38,220 overran 819 00:48:38,220 --> 00:48:42,339 like 100 percent but then if you look at 820 00:48:42,339 --> 00:48:45,540 this probability what's the problem someone's 821 00:48:45,540 --> 00:48:49,079 a minority and should not be arrested 822 00:48:49,079 --> 00:48:53,329 that number is ninety percent and the general population 823 00:48:53,329 --> 00:48:56,599 it's 95 percent so my question then is 824 00:48:56,599 --> 00:48:59,270 if it's not okay to stop and frisk ninety-five percent of the general 825 00:48:59,270 --> 00:48:59,970 population 826 00:48:59,970 --> 00:49:03,680 for nothing then why is it okay to stop and frisk 827 00:49:03,680 --> 00:49:07,339 ninety percent minority population for nothing 828 00:49:07,339 --> 00:49:10,849 it's almost as if like it at the airport 829 00:49:10,849 --> 00:49:15,630 you walk through and certain people get to go through without getting to act and 830 00:49:15,630 --> 00:49:17,180 other people get to go after 831 00:49:17,180 --> 00:49:21,669 get checked that's a problem 832 00:49:21,670 --> 00:49:25,980 so going to the national stage this is back to our 833 00:49:25,980 --> 00:49:29,480 are places players Parel framework and and 834 00:49:29,480 --> 00:49:32,619 if you notice before I nor the NSA I 835 00:49:32,619 --> 00:49:37,150 citizen surveillance up 836 00:49:37,150 --> 00:49:40,530 this is an important one because we are really grappling as the country 837 00:49:40,530 --> 00:49:44,650 where the trade-off between security and privacy 838 00:49:44,650 --> 00:49:47,910 and security and there's a lot of 30 people here right now 839 00:49:47,910 --> 00:49:52,750 I security is probably at least 10 years ahead a prime 840 00:49:52,750 --> 00:49:56,140 I I think generally security 841 00:49:56,140 --> 00:49:59,220 easy to formally it's really hard to execute well 842 00:49:59,220 --> 00:50:02,810 but privacy is this it's hard to even formula 843 00:50:02,810 --> 00:50:06,540 and so typically security loses out 844 00:50:06,540 --> 00:50:10,190 our Privacy loses out to security it's sort of like in the 845 00:50:10,190 --> 00:50:14,230 you know the old adage where with effective breakfast the pics are 846 00:50:14,230 --> 00:50:16,530 committed but the chickens are involved 847 00:50:16,530 --> 00:50:21,430 security is about commitment and privacy is really about involved 848 00:50:21,430 --> 00:50:25,220 and that's why privacy typical loses out to security 849 00:50:25,220 --> 00:50:29,399 however I think that's that's potentially long-term disaster 850 00:50:29,400 --> 00:50:32,920 and is a a quote by Alex Howard 851 00:50:32,920 --> 00:50:36,420 a a talked about their quite a bit and 852 00:50:36,420 --> 00:50:39,660 I think cafe pretty well ones that were being affect ross 853 00:50:39,660 --> 00:50:43,299 without being able to verify and I think that's it a big issue 854 00:50:43,300 --> 00:50:48,390 around as a democracy how do we find the right balance between security and 855 00:50:48,390 --> 00:50:49,348 privacy 856 00:50:49,349 --> 00:50:53,230 and soap President Obama is actually calling for more transparency 857 00:50:53,230 --> 00:50:56,890 in five record them for balance laws to be reviewed 858 00:50:56,890 --> 00:51:00,839 I which i think is great I think we need more of that as in SF to be the day that 859 00:51:00,839 --> 00:51:04,119 the fis the core rules but soon after we should be in the know 860 00:51:04,119 --> 00:51:08,280 what actually was decided and and why 861 00:51:08,280 --> 00:51:12,510 this was really though kinda disturbing the NSA chief announced plans to replace 862 00:51:12,510 --> 00:51:15,560 a thousand system administrators with machines I 863 00:51:15,560 --> 00:51:19,480 you know as snowden and what's known is a with a system administrator 864 00:51:19,480 --> 00:51:22,950 I and so I guess since the leak is stopped 865 00:51:22,950 --> 00:51:26,069 the obviously need your is get rid of everybody who 866 00:51:26,070 --> 00:51:30,869 as active that and replace them with machines 867 00:51:30,869 --> 00:51:35,540 I think that's a problem and and I think that if you look at sort of his pyramid 868 00:51:35,540 --> 00:51:39,359 update a least information information now now 869 00:51:39,359 --> 00:51:43,000 least the wisdom we usually don't get to the wisdom part 870 00:51:43,000 --> 00:51:46,890 we're very good at being in amber with our tools in our toys 871 00:51:46,890 --> 00:51:50,460 I we usually don't think about how they and 872 00:51:50,460 --> 00:51:54,520 I impact our society substantially profound way 873 00:51:54,520 --> 00:51:57,710 as they got these things like government doesn't trust people they trust me 874 00:51:57,710 --> 00:52:00,980 you know much transparency know what they're doing and wisdom is come hard to 875 00:52:00,980 --> 00:52:01,700 come by 876 00:52:01,700 --> 00:52:05,598 and then you have the sentient Brian the cloud in less than 20 years possibly 877 00:52:05,599 --> 00:52:09,270 I I would be alarmist but break 878 00:52:09,270 --> 00:52:12,300 you know i mean are what are we walking ourselves into 879 00:52:12,300 --> 00:52:17,430 I that might be hard to walk ourselves out a up 880 00:52:17,430 --> 00:52:21,868 Henry Mason as is you're amazing II chief scientist at Bentley I think 881 00:52:21,869 --> 00:52:25,680 meredith is great at it I think I I really like witches 882 00:52:25,680 --> 00:52:29,259 map with code is awesome and yes it is really awesome 883 00:52:29,260 --> 00:52:33,790 but it's not always awesome I sometimes it's not so awesome 884 00:52:33,790 --> 00:52:37,550 I think we'll realize in 2008 that map as was not so awesome 885 00:52:37,550 --> 00:52:41,520 I where we had financial weapons of mass destruction that almost brought down the 886 00:52:41,520 --> 00:52:42,220 World 887 00:52:42,220 --> 00:52:45,430 world economy I so 888 00:52:45,430 --> 00:52:49,190 like any good equation be improved by a well 889 00:52:49,190 --> 00:52:52,900 place television I why we 890 00:52:52,900 --> 00:52:56,030 try to look at what are the values that 891 00:52:56,030 --> 00:52:59,740 we in dow with our great technology in pools 892 00:52:59,740 --> 00:53:03,410 I to make map with code awesome the values are 893 00:53:03,410 --> 00:53:07,859 negative their its powerfully negatively awesome 894 00:53:07,859 --> 00:53:11,598 but if they're good values then maybe they can really help things the problem 895 00:53:11,599 --> 00:53:14,260 is that values are not very exact 896 00:53:14,260 --> 00:53:19,220 I they are often in conflict like security and privacy 897 00:53:19,220 --> 00:53:22,759 and security and liberty and and restraint and openness and things like 898 00:53:22,760 --> 00:53:23,820 that so 899 00:53:23,820 --> 00:53:26,950 we have to figure out how to get there from here 900 00:53:26,950 --> 00:53:30,169 I one last story 901 00:53:30,170 --> 00:53:33,690 I when I got out of school 902 00:53:33,690 --> 00:53:36,950 and now I I'm ready 5144 lock 903 00:53:36,950 --> 00:53:41,240 what now Lockheed Martin at the time with dynamic spaces systems in San Diego 904 00:53:41,240 --> 00:53:44,669 and I was gonna work on the space shuttle and we're gonna put our 905 00:53:44,670 --> 00:53:48,070 upper stage in the shuttle so I saw for work 906 00:53:48,070 --> 00:53:51,650 January 27 1986 the next day 907 00:53:51,650 --> 00:53:55,420 the next day the shuttle blows up 908 00:53:55,420 --> 00:53:59,190 and so we were like we on this government contract to work on our heels 909 00:53:59,190 --> 00:54:01,230 for about six months while they did 910 00:54:01,230 --> 00:54:04,720 the the investigation why this happened 911 00:54:04,720 --> 00:54:08,899 and it turned out as we all know that the Rogers commission was 912 00:54:08,900 --> 00:54:12,000 form and richard fineman I you 913 00:54:12,000 --> 00:54:15,050 I i think I love you should now I 914 00:54:15,050 --> 00:54:19,810 Nobel laureate in physics actually came out with the reason why 915 00:54:19,810 --> 00:54:23,990 this happen and it turns out it's very cold for the morning and the O-rings 916 00:54:23,990 --> 00:54:25,629 lost their last the city and 917 00:54:25,630 --> 00:54:28,770 and gas II ghastly I 918 00:54:28,770 --> 00:54:32,119 ghastly and and and punctured 919 00:54:32,119 --> 00:54:35,500 the me and so 920 00:54:35,500 --> 00:54:38,880 funny thing is fine really was able to be on the Rogers Commission 921 00:54:38,880 --> 00:54:41,970 right his wife actually convince them to be on the commission she said you know 922 00:54:41,970 --> 00:54:43,230 this and eleven guys 923 00:54:43,230 --> 00:54:46,420 and and women I think I was on active up 924 00:54:46,420 --> 00:54:50,010 me eleven people in this not running around from meeting to meeting 925 00:54:50,010 --> 00:54:51,570 presentation presentation 926 00:54:51,570 --> 00:54:55,500 me one mosquito buzzing around looking for the answer now is 927 00:54:55,500 --> 00:54:59,690 Richard and she cannot be right because William Rogers secretary say during the 928 00:54:59,690 --> 00:55:01,140 commission at one point was 929 00:55:01,140 --> 00:55:06,509 quarters you know saying its fine guys really got me crazy 930 00:55:06,510 --> 00:55:10,010 and and and yet he found the answer and he wasn't 931 00:55:10,010 --> 00:55:14,099 in aviation expert he was in the space industry is why I 932 00:55:14,099 --> 00:55:17,609 I like to think I've as an eclectic general 933 00:55:17,609 --> 00:55:22,140 and aquatic generalist are these technologies that actually can endow the 934 00:55:22,140 --> 00:55:24,660 technology with the value the culture 935 00:55:24,660 --> 00:55:28,299 and Steve Jobs is famous for this obviously I love is quote I 936 00:55:28,300 --> 00:55:31,470 1996 he says I 937 00:55:31,470 --> 00:55:34,649 nonprofit Microsoft is there just have no taste 938 00:55:34,650 --> 00:55:39,380 and not in a small have no case in a big way 939 00:55:39,380 --> 00:55:42,890 I in that they don't think originally and they don't in dow 940 00:55:42,890 --> 00:55:46,490 their their products with the diver the culture 941 00:55:46,490 --> 00:55:49,819 and I think that's really important if you look at I 942 00:55:49,819 --> 00:55:55,980 Norio guy up who is president of Sony in the seventies 943 00:55:55,980 --> 00:56:00,609 he oversaw the development %uh compact disc player 944 00:56:00,609 --> 00:56:04,819 and he required that it had 74 minutes 945 00:56:04,819 --> 00:56:10,380 a body y7 for me 946 00:56:10,380 --> 00:56:13,930 there you know 53 beethoven's ninth 947 00:56:13,930 --> 00:56:18,290 74 minutes and he didn't want to change the CD 948 00:56:18,290 --> 00:56:21,740 and so he drove the teams to get seventy four minutes 949 00:56:21,740 --> 00:56:25,098 %uh audio on that CD and 950 00:56:25,099 --> 00:56:29,400 I just love that story because it really shows that this cultural value 951 00:56:29,400 --> 00:56:32,560 music and art was in doubt into 952 00:56:32,560 --> 00:56:36,759 I something that was critical for the aging something rely upon our 953 00:56:36,760 --> 00:56:41,930 built on I into what what we take for granted now in 954 00:56:41,930 --> 00:56:45,759 music on demand and you see that I 955 00:56:45,760 --> 00:56:49,640 important dynamic over and over and over again 956 00:56:49,640 --> 00:56:53,328 because quite honestly I would argue we are making this stuff up as we go 957 00:56:53,329 --> 00:56:57,310 anyone who tells you differently to the following you were following themselves 958 00:56:57,310 --> 00:57:00,900 and I think they're sorta four stages that that we go through 959 00:57:00,900 --> 00:57:04,400 I friend in my eye sign through this of an element 960 00:57:04,400 --> 00:57:08,069 by your bike out beer they will do anything I 961 00:57:08,069 --> 00:57:11,400 and so he and and I think there's evolution of technology you go from 962 00:57:11,400 --> 00:57:14,890 in a sense which is sorta deny like there's no problem and then you get this 963 00:57:14,890 --> 00:57:16,290 frontier stage 964 00:57:16,290 --> 00:57:19,819 in privacy certainly we r we're in this for a while with this yr 965 00:57:19,819 --> 00:57:24,970 just get over it privacy dad promulgated by a lotta technology leaders 966 00:57:24,970 --> 00:57:29,740 I meant we often have the sort of I over rotation for regulation like there's all 967 00:57:29,740 --> 00:57:30,459 these ongoing 968 00:57:30,460 --> 00:57:35,119 every delays in and I think regulations a very blunt instrument and its 969 00:57:35,119 --> 00:57:38,210 sometimes it's like scratching your back with a chainsaw 970 00:57:38,210 --> 00:57:41,990 yes you get yes but there's a lot well collateral damage 971 00:57:41,990 --> 00:57:45,439 and then you get to this innovation stage right thank you have 972 00:57:45,440 --> 00:57:48,790 smart regulation or smart policy connected 973 00:57:48,790 --> 00:57:52,150 to innovation and so the question is really 974 00:57:52,150 --> 00:57:55,420 how you get there from here and our targets not 975 00:57:55,420 --> 00:57:59,730 how but it's more who is important 976 00:57:59,730 --> 00:58:03,329 and I talked about geeks suits and long 977 00:58:03,329 --> 00:58:07,680 for a while now and I think those are the pillars responsible innovation 978 00:58:07,680 --> 00:58:11,339 I I think if you have only two of the three 979 00:58:11,339 --> 00:58:14,828 you don't really have what's necessary and sufficient to 980 00:58:14,829 --> 00:58:18,880 I get to the right answer I 981 00:58:18,880 --> 00:58:22,740 I think if if you look at because raw making this up as we go 982 00:58:22,740 --> 00:58:26,149 there's a sort of Overton Window let you must you must have tension on either 983 00:58:26,150 --> 00:58:27,069 side 984 00:58:27,069 --> 00:58:31,849 I in order to find the right answer so you have to have this push poll 985 00:58:31,849 --> 00:58:35,109 tug-of-war to find the right answer cuz honestly we're just kicking around in 986 00:58:35,109 --> 00:58:35,560 the dark 987 00:58:35,560 --> 00:58:39,089 really don't understand where we're going and we need this dynamic 988 00:58:39,089 --> 00:58:44,029 a a debate and experimentation to get there 989 00:58:44,030 --> 00:58:47,660 because I think the cycles really important I think we invent things and 990 00:58:47,660 --> 00:58:48,390 then we 991 00:58:48,390 --> 00:58:51,629 question them in scrutinizing criticize them I 992 00:58:51,630 --> 00:58:56,030 this crew clearly does that and embraces that I think that's really important 993 00:58:56,030 --> 00:58:59,579 and then you need to do it again life lesson learned do it again 994 00:58:59,579 --> 00:59:02,700 I wanna the objective I'll 995 00:59:02,700 --> 00:59:07,049 not it's the new companies really understand data flows in like 996 00:59:07,050 --> 00:59:11,020 companies understand the data they have been I compete with that efficiently and 997 00:59:11,020 --> 00:59:14,410 easily and I'm really excited about that because I think there's a lot 998 00:59:14,410 --> 00:59:18,790 knowledge and wisdom out there that's that needs to be understood 999 00:59:18,790 --> 00:59:22,069 so that's all I have I really appreciate the attention 1000 00:59:22,069 --> 00:59:25,470 I i think we have I'm question to 1001 00:59:25,470 --> 00:59:32,470 with the bay thanks 1002 00:59:38,619 --> 00:59:41,910 I wanted to make a comment about slide 1003 00:59:41,910 --> 00:59:46,190 you know how much people your privacy 1004 00:59:46,190 --> 00:59:49,520 back in Iraq where you known 1005 00:59:49,520 --> 00:59:54,200 very small number of people and you could move somewhere else and no one 1006 00:59:54,200 --> 00:59:55,828 would know anything about you 1007 00:59:55,829 --> 01:00:00,079 there's a difference between someone seen your license plate 1008 01:00:00,079 --> 01:00:03,520 in public and having databases 1009 01:00:03,520 --> 01:00:07,940 have your license plate with GPS coordinates an times 1010 01:00:07,940 --> 01:00:11,950 info in so there's there's a 1011 01:00:11,950 --> 01:00:16,480 it's not quite just which public lot more 1012 01:00:16,480 --> 01:00:19,520 i would i would agree I but I the world is 1013 01:00:19,520 --> 01:00:23,280 my point is the world is shrinking and these datasets are shrinking it 1014 01:00:23,280 --> 01:00:26,730 and there is we are going my point is not that 1015 01:00:26,730 --> 01:00:30,319 it's all its I'm not saying get over it i'm saying that 1016 01:00:30,319 --> 01:00:34,450 these technology are shrinking the world and we need to use the values that we 1017 01:00:34,450 --> 01:00:35,310 had then 1018 01:00:35,310 --> 01:00:40,109 love respect and empathy in order to live in a world where you just can't 1019 01:00:40,109 --> 01:00:41,910 hide behind anonymity 1020 01:00:41,910 --> 01:00:46,368 and so yes there are certain abuses around access to data that we 1021 01:00:46,369 --> 01:00:50,050 we should curtail I certainly health data and other 1022 01:00:50,050 --> 01:00:53,400 other kinds of data need to be protected but when those 1023 01:00:53,400 --> 01:00:56,920 when that they'd invariably gets out how we gonna be 1024 01:00:56,920 --> 01:01:01,160 with it are we gonna II behave responsibly and respectfully 1025 01:01:01,160 --> 01:01:04,279 and not abuse that that information I think 1026 01:01:04,280 --> 01:01:10,000 work we're headed down that that road in this can turn off all the data 1027 01:01:10,000 --> 01:01:14,040 you have to act responsibly back 1028 01:01:14,040 --> 01:01:17,060 problem ike for a regular person 1029 01:01:17,060 --> 01:01:20,250 how would you even know who's collecting information on you 1030 01:01:20,250 --> 01:01:23,260 let alone can you go there opt out correct 1031 01:01:23,260 --> 01:01:26,869 sector how do you even know I think I think it's coming one reason we put out 1032 01:01:26,869 --> 01:01:28,970 that through our products the least you could now 1033 01:01:28,970 --> 01:01:32,109 at least you can see what's out there nears there's actually 1034 01:01:32,109 --> 01:01:35,420 legislation in Congress now to give you access to the information 1035 01:01:35,420 --> 01:01:38,970 about that's publicly available and even private L 1036 01:01:38,970 --> 01:01:42,419 and I'm I'm a big supporter that because I think without transparency 1037 01:01:42,420 --> 01:01:45,720 we don't know what's going on so I think once we know then we could take 1038 01:01:45,720 --> 01:01:47,189 appropriate steps 1039 01:01:47,190 --> 01:01:50,460 are you control it how you corrected you comment on it 1040 01:01:50,460 --> 01:01:53,630 make sure the record's clear 1041 01:01:53,630 --> 01:01:56,839 sir I I'm moving this with as a ride 1042 01:01:56,839 --> 01:01:59,869 really like this talking thing was great lots of questions but I'll try to stick 1043 01:01:59,869 --> 01:02:00,150 to 1044 01:02:00,150 --> 01:02:04,349 to just two so one is comment about the power disparity 1045 01:02:04,349 --> 01:02:08,040 sort of hinted at in beginning when you talk about 1046 01:02:08,040 --> 01:02:12,390 stalkers and victims can I think I think isn't and recognize that power disparity 1047 01:02:12,390 --> 01:02:13,700 because as an 1048 01:02:13,700 --> 01:02:17,180 law-abiding individual you're kinda powerless just any crazy person out 1049 01:02:17,180 --> 01:02:17,828 there 1050 01:02:17,829 --> 01:02:21,490 and actually because my work in security I had a mentally unstable stopper for a 1051 01:02:21,490 --> 01:02:22,529 while and was 1052 01:02:22,530 --> 01:02:25,680 by talking to realize how easy they can find out where well it 1053 01:02:25,680 --> 01:02:29,930 so that's one thing I think you should add to that a question 1054 01:02:29,930 --> 01:02:33,680 the other thing is whether consumers really have a choice today 1055 01:02:33,680 --> 01:02:36,819 is there a 20-year-old who can choose not to be on Facebook 1056 01:02:36,819 --> 01:02:40,410 not have a cell phone be off the grid and still have a social life 1057 01:02:40,410 --> 01:02:43,799 that really a choice I agree I don't think it's a choice and that's why I 1058 01:02:43,800 --> 01:02:45,859 think we are in this round 1059 01:02:45,859 --> 01:02:48,859 new policies to deal with that inescapable reality the 1060 01:02:48,859 --> 01:02:52,869 just can't get off the grid in so how do you 1061 01:02:52,869 --> 01:02:56,569 to the ladies previous question is how do you know when information is out 1062 01:02:56,569 --> 01:02:57,509 there 1063 01:02:57,510 --> 01:03:01,300 how you make sure the people act responsibly with 1064 01:03:01,300 --> 01:03:05,220 that's the error that we ran and its gonna i think is gonna be this 1065 01:03:05,220 --> 01:03:08,259 reemergence love a of manners 1066 01:03:08,260 --> 01:03:12,020 and and empathy and better behavior in order to 1067 01:03:12,020 --> 01:03:15,460 deal with right the previous lady's question now that 1068 01:03:15,460 --> 01:03:19,230 your village is actually the world that's right 1069 01:03:19,230 --> 01:03:23,339 thank you I'm keep 1070 01:03:23,339 --> 01:03:27,750 Park so I just one question for you 1071 01:03:27,750 --> 01:03:33,000 you belief that uses have right to view the profiles that you have created for 1072 01:03:33,000 --> 01:03:33,930 them 1073 01:03:33,930 --> 01:03:36,960 and should they be able to change 1074 01:03:36,960 --> 01:03:40,309 what the thing might be wrong about them 1075 01:03:40,309 --> 01:03:45,109 yeah absolutely i II that's one reason why talk about their product 1076 01:03:45,109 --> 01:03:49,230 if you that transparency I so there's always 1077 01:03:49,230 --> 01:03:52,250 to questions and and you you now both 1078 01:03:52,250 --> 01:03:55,730 which is what information's out there and then can I correct 1079 01:03:55,730 --> 01:03:58,920 and the challenge with correction that we have today which i think is a 1080 01:03:58,920 --> 01:04:03,790 same is that a company like & telly it sits at the very end to the data 1081 01:04:03,790 --> 01:04:04,509 supplied thing 1082 01:04:04,510 --> 01:04:07,650 their retail and there's this tremendous 1083 01:04:07,650 --> 01:04:11,180 data sucking machine that is they supply chain 1084 01:04:11,180 --> 01:04:14,359 upstream that retail and so 1085 01:04:14,359 --> 01:04:17,930 what you really want to do if you really want to correct it at the headwaters 1086 01:04:17,930 --> 01:04:21,089 that supply chain at all possible I 1087 01:04:21,089 --> 01:04:25,040 and that's not possible today unfortunately I think it's changing 1088 01:04:25,040 --> 01:04:28,630 and I think more more public pressure around 1089 01:04:28,630 --> 01:04:31,930 that will will take a obvious ultimately 1090 01:04:31,930 --> 01:04:34,940 if you look at the Fair Credit Reporting Act which i think is important to the 1091 01:04:34,940 --> 01:04:36,369 fair card bring act was 1092 01:04:36,369 --> 01:04:39,589 established in 1970 and 1093 01:04:39,589 --> 01:04:44,040 it basically says that someone's going to use information for your lively 1094 01:04:44,040 --> 01:04:47,240 where you live when you get a job and I 1095 01:04:47,240 --> 01:04:50,700 no matter where they get the data they have to share with you whether there's 1096 01:04:50,700 --> 01:04:51,270 any 1097 01:04:51,270 --> 01:04:54,940 adverse information and what that says is 1098 01:04:54,940 --> 01:04:58,530 and resided at it because if your name is John Smith the chance that you're a 1099 01:04:58,530 --> 01:05:00,050 victim of mistaken identity 1100 01:05:00,050 --> 01:05:03,780 issue and there's a lot of John Smith's out there that are criminals and their 1101 01:05:03,780 --> 01:05:06,599 son Ubuntu Johnson is that are law-abiding citizens 1102 01:05:06,599 --> 01:05:10,690 and if someone pulls wrong back on 2nd decides not to hire you because it 1103 01:05:10,690 --> 01:05:15,200 that's not fair you're not that johnson et and so 1104 01:05:15,200 --> 01:05:18,720 access information I to the 1105 01:05:18,720 --> 01:05:23,200 previous element point is the world is shrinking and 1106 01:05:23,200 --> 01:05:26,410 date is available we can't abuse it and 1107 01:05:26,410 --> 01:05:29,450 FCRA says release important 1108 01:05:29,450 --> 01:05:32,750 survival I'm Contax housing why men 1109 01:05:32,750 --> 01:05:35,990 concern you have to abide by these rules 1110 01:05:35,990 --> 01:05:39,290 how do we get there with these 1111 01:05:39,290 --> 01:05:42,910 less vital uses the day that's the question I think that 1112 01:05:42,910 --> 01:05:47,578 we as a society figure out 1113 01:05:47,579 --> 01:05:51,710 sorry tyson he said now works for Facebook so I definitely enjoyed the 1114 01:05:51,710 --> 01:05:52,910 stock 1115 01:05:52,910 --> 01:05:56,009 I had to kind related questions first would be 1116 01:05:56,010 --> 01:06:00,380 do you see as we move forward as technology gets better and we 1117 01:06:00,380 --> 01:06:03,819 have more and more information that is 1118 01:06:03,819 --> 01:06:07,710 being gathered created about people even indirectly 1119 01:06:07,710 --> 01:06:11,470 tells the story morning agreeing on innovation tax where 1120 01:06:11,470 --> 01:06:16,040 you things can be inferred about you will the value transparency 1121 01:06:16,040 --> 01:06:19,569 go down as the information about you out there 1122 01:06:19,569 --> 01:06:23,759 just grows and grows with time yeah I 1123 01:06:23,760 --> 01:06:27,450 really good question that's why I ultimately have great faith 1124 01:06:27,450 --> 01:06:30,490 and regulation are 1125 01:06:30,490 --> 01:06:34,569 or at least policy around abusive the data itself because 1126 01:06:34,569 --> 01:06:38,430 at the end of the day doesn't matter where I got the day weather was inferred 1127 01:06:38,430 --> 01:06:40,220 or not or private or public 1128 01:06:40,220 --> 01:06:43,578 am I gonna use that data against somebody 1129 01:06:43,579 --> 01:06:46,869 I and for an abusive use wetherbee 1130 01:06:46,869 --> 01:06:51,309 completely criminal like stocking and and other kinds offenses that are 1131 01:06:51,309 --> 01:06:55,950 harassment those are the things or my to use it in a very subtle way 1132 01:06:55,950 --> 01:07:00,680 I to discriminate I know that 1133 01:07:00,680 --> 01:07:05,700 insurance companies know who are back there now they know who are bad risks 1134 01:07:05,700 --> 01:07:09,220 are they using that to like based on social the 1135 01:07:09,220 --> 01:07:12,808 media a no is a bad wrist 1136 01:07:12,809 --> 01:07:15,890 I don't think they're at the point using that data to 1137 01:07:15,890 --> 01:07:19,348 allow or disallow policies when you call to cancel 1138 01:07:19,349 --> 01:07:22,750 in a fight you less if you're a bad wrist 1139 01:07:22,750 --> 01:07:26,640 and so you start to see the sort of insidious use of this information 1140 01:07:26,640 --> 01:07:30,598 and are those uses all appropriate or not I think youth based policy 1141 01:07:30,599 --> 01:07:34,609 is where we're going because access based policy 1142 01:07:34,609 --> 01:07:38,828 is you really can't you don't know where they came from 1143 01:07:38,829 --> 01:07:42,130 so there's its it's gonna be 1144 01:07:42,130 --> 01:07:46,220 a growing problem where you space is very 1145 01:07:46,220 --> 01:07:50,020 although difficult at least track 1146 01:07:50,020 --> 01:07:53,849 yeah I I think you can answer yeah addressing was there which was that 1147 01:07:53,849 --> 01:07:57,920 UN Security we've said for a long time that security through obscurity is 1148 01:07:57,920 --> 01:08:01,170 is not a practical approach I've been seeing a lot of people in the press 1149 01:08:01,170 --> 01:08:04,869 more recently dealings rights issues anti-media 1150 01:08:04,869 --> 01:08:08,119 privacy through security but again it's for those things were as the technology 1151 01:08:08,119 --> 01:08:08,859 grows 1152 01:08:08,859 --> 01:08:12,640 that to seems to become licensed practical rights yeah i i agree 1153 01:08:12,640 --> 01:08:16,068 I don't think that I don't think that there 1154 01:08:16,069 --> 01:08:19,250 i think is a short term thing with this Floyd they're they're worth companies 1155 01:08:19,250 --> 01:08:20,738 out there that which is floyd 1156 01:08:20,738 --> 01:08:24,269 the net with stuff about in order to cover-up 1157 01:08:24,270 --> 01:08:28,150 the other stuff I don't think like an arms race to me 1158 01:08:28,149 --> 01:08:31,460 thanks I for 1159 01:08:31,460 --> 01:08:35,310 such from you this authentic neatly so 1160 01:08:35,310 --> 01:08:39,850 I like your comparison between europe and the US so there was something 1161 01:08:39,850 --> 01:08:43,109 probably usually know there's a big debate in Europe on the right 1162 01:08:43,109 --> 01:08:47,160 to be forgotten so I would really like to hear your opinion on this 1163 01:08:47,160 --> 01:08:50,870 yeah that's it that's a good 1i actually I hate them like my post but I will post 1164 01:08:50,870 --> 01:08:51,330 on that 1165 01:08:51,330 --> 01:08:55,100 I harm that talks about the right to be forgotten 1166 01:08:55,100 --> 01:08:58,969 and the problem that I have with the right to be forgotten 1167 01:08:58,969 --> 01:09:02,429 it is is really at the crux of this disclosure discretion 1168 01:09:02,429 --> 01:09:07,520 I debate if you look at disclosure close on the heels of disclosure is the right 1169 01:09:07,520 --> 01:09:09,279 to an accurate history 1170 01:09:09,279 --> 01:09:13,440 and if you up if you have a right to be forgotten 1171 01:09:13,439 --> 01:09:18,160 where is history gonna go from there and in the exam I like to use is on 1172 01:09:18,160 --> 01:09:21,858 the public Twitter fight so I've been involved in some Twitter fights 1173 01:09:21,859 --> 01:09:25,480 you know are and there's been a lot of water fights 1174 01:09:25,479 --> 01:09:28,968 so the thought experiment I like that provide 1175 01:09:28,969 --> 01:09:34,529 is I'm in for a fight with you we both say things that we regret 1176 01:09:34,529 --> 01:09:37,529 and we both agree that we just like 1177 01:09:37,529 --> 01:09:40,900 a racist because we both embarrass ourselves 1178 01:09:40,899 --> 01:09:43,910 knows unfortunate got it would only say 1179 01:09:43,910 --> 01:09:48,679 a a racetrack I don't think that's right 1180 01:09:48,679 --> 01:09:51,359 and the reason I don't think it's right it's because other people were watching 1181 01:09:51,359 --> 01:09:53,190 the Twitter fight 1182 01:09:53,189 --> 01:09:57,219 and there's a public value to that and they're gonna take some lessons from 1183 01:09:57,219 --> 01:09:57,940 that 1184 01:09:57,940 --> 01:10:02,989 may be good lessons may be bad but there's a public record their 1185 01:10:02,989 --> 01:10:06,000 up and the fact that he should be able to 1186 01:10:06,000 --> 01:10:10,179 erase public history i think is really dangerous 1187 01:10:10,179 --> 01:10:14,710 up I think on the flip side or the argument I often hear is 1188 01:10:14,710 --> 01:10:18,480 yeah but the internet never forgets can't there be 1189 01:10:18,480 --> 01:10:23,049 some tailing off I love 1190 01:10:23,050 --> 01:10:26,250 up the historical record and every heart I 1191 01:10:26,250 --> 01:10:29,670 yes I drop you want I think the problem is not really 1192 01:10:29,670 --> 01:10:32,770 tweets but things is more 1193 01:10:32,770 --> 01:10:36,699 it's a serious offense like your criminal record for essence 1194 01:10:36,699 --> 01:10:39,730 in particular for his been acquitted I think 1195 01:10:39,730 --> 01:10:43,928 a certain point the problem with Google in Italy one guy was 1196 01:10:43,929 --> 01:10:47,840 first inflicted or persecuted for something he was completely acquitted 1197 01:10:47,840 --> 01:10:51,380 right expired so expungement of records should be possible 1198 01:10:51,380 --> 01:10:54,679 I agree I agree with that and there are certain circumstances but I think 1199 01:10:54,679 --> 01:10:55,070 they're 1200 01:10:55,070 --> 01:10:59,120 the key word there are certain circumstances that 1201 01:10:59,120 --> 01:11:02,780 warrant that kinda that country we 1202 01:11:02,780 --> 01:11:06,550 Intel is expunged record all the time to a criminal records and I think that's 1203 01:11:06,550 --> 01:11:07,320 important 1204 01:11:07,320 --> 01:11:11,980 and but I think they are very particular rise assert certain classes 1205 01:11:11,980 --> 01:11:15,530 that are the most egregious that really should have a shelf life 1206 01:11:15,530 --> 01:11:19,739 up but I think when you say right to be forgotten 1207 01:11:19,739 --> 01:11:23,759 its could be a interpret very broadly now all the sudden 1208 01:11:23,760 --> 01:11:27,449 you know the internet is for getting every seven days and obviously 1209 01:11:27,449 --> 01:11:34,449 you know that's hyperbole but if it's hard to find that balance 1210 01:11:34,600 --> 01:11:38,460 Scott Hensley with the science and tech for DHS 1211 01:11:38,460 --> 01:11:41,630 during the talk it seems very western centric you talked about 1212 01:11:41,630 --> 01:11:46,210 implicitly things here in the US and Europe it seems to me that 1213 01:11:46,210 --> 01:11:49,760 we're headed down a slippery slope where this is you know social control 1214 01:11:49,760 --> 01:11:51,449 capability 1215 01:11:51,449 --> 01:11:54,719 thought about what men are dicks in your Chinese here you can be doing with your 1216 01:11:54,719 --> 01:11:57,179 technology in say 10 years to 1217 01:11:57,179 --> 01:12:01,170 facilitate things and make people like Joseph Goebbels proud in terms of 1218 01:12:01,170 --> 01:12:03,520 controlling everything that's going on in our society 1219 01:12:03,520 --> 01:12:06,719 thought about that side equation sure I mean 1220 01:12:06,719 --> 01:12:10,150 I think that I think I mean is your point that 1221 01:12:10,150 --> 01:12:14,549 if you understand more the information I that is therefore more easily 1222 01:12:14,550 --> 01:12:19,460 you can more easily exploited that affair 1223 01:12:19,460 --> 01:12:22,750 I was struck by the movie some years ago the lives of others which is an east 1224 01:12:22,750 --> 01:12:25,239 european documentary that looked at 1225 01:12:25,239 --> 01:12:29,230 citizen spying on themselves all across Eastern Germany which was one of the 1226 01:12:29,230 --> 01:12:31,799 ways that the country was so tightly controlled 1227 01:12:31,800 --> 01:12:36,070 finally came apart but somebody's technology some 1228 01:12:36,070 --> 01:12:39,380 this stuff when you look at it from the perspective 1229 01:12:39,380 --> 01:12:43,350 state-run societies that are interested in controlling 1230 01:12:43,350 --> 01:12:46,660 all their society the implications are not particularly good 1231 01:12:46,660 --> 01:12:49,809 yeah I i'm not I think when para disparities 1232 01:12:49,810 --> 01:12:54,570 large the law must 1233 01:12:54,570 --> 01:12:58,190 curtail those kinds of abuses now yeah on certain 1234 01:12:58,190 --> 01:13:01,530 per pallet Erin societies there's no law 1235 01:13:01,530 --> 01:13:05,670 and so therefore people need to keep their secrets in order to 1236 01:13:05,670 --> 01:13:08,820 to balance the power disparity I'm not 1237 01:13:08,820 --> 01:13:11,880 I'm ap I absolutely believe that of course I mean that 1238 01:13:11,880 --> 01:13:15,679 as the only to you hack I I don't necessarily think that 1239 01:13:15,679 --> 01:13:18,980 the way forward is to up stop 1240 01:13:18,980 --> 01:13:22,349 turn of the day to turn up tools I because 1241 01:13:22,350 --> 01:13:26,570 up the abuses I'd I think it's important to deal head on with the abuses 1242 01:13:26,570 --> 01:13:31,580 I think white I mean I'm surprised that's the at work known question 1243 01:13:31,580 --> 01:13:36,239 away for I by I I think yes 1244 01:13:36,239 --> 01:13:39,449 you know every cool can be used for good or for ill 1245 01:13:39,449 --> 01:13:42,669 and I think the law is often white 1246 01:13:42,670 --> 01:13:46,500 prepare what we're prepared to society to get the most out of a technology 1247 01:13:46,500 --> 01:13:51,300 yeah and it could be a view hi I think that we are running an experiment but I 1248 01:13:51,300 --> 01:13:53,350 don't think the way forward is back 1249 01:13:53,350 --> 01:13:58,890 name we disagree agreed 1250 01:13:58,890 --> 01:14:03,040 thank you very much excellent talk about a dozen years ago there was a 1251 01:14:03,040 --> 01:14:07,440 guy promoting big data before it became very popular but basically 1252 01:14:07,440 --> 01:14:11,620 he said that om the worst thing you can do 1253 01:14:11,620 --> 01:14:15,199 at the time now was in terms of giving up 1254 01:14:15,199 --> 01:14:20,480 who you are was basically go to the same supermarket at that time ANP supermarket 1255 01:14:20,480 --> 01:14:21,400 was very big 1256 01:14:21,400 --> 01:14:24,460 and basically over a month go in there 1257 01:14:24,460 --> 01:14:28,440 him pay with your credit card because from their data and they were getting 1258 01:14:28,440 --> 01:14:31,049 about a terabyte of data from the cash registers 1259 01:14:31,050 --> 01:14:34,890 every day so basically they knew how many toothpicks you 1260 01:14:34,890 --> 01:14:38,980 wat if you bought pampers all kinds medical 1261 01:14:38,980 --> 01:14:43,299 so basically they had a profile that was obviously was internal 1262 01:14:43,300 --> 01:14:46,800 to their own data move servers but my question is 1263 01:14:46,800 --> 01:14:51,739 is that information shared and is that legal and it was when I go into a credit 1264 01:14:51,739 --> 01:14:53,650 card and i buy something 1265 01:14:53,650 --> 01:14:57,400 is that legally doable today 1266 01:14:57,400 --> 01:15:01,120 terms of what I bought that's not true is it legally terrible are 1267 01:15:01,120 --> 01:15:05,420 arts can someone go by that date is that correct no words if I bought 1268 01:15:05,420 --> 01:15:09,370 a hundred dollars with the groceries and I bought everything from soup to nuts 1269 01:15:09,370 --> 01:15:12,929 my question is is that information currently 1270 01:15:12,929 --> 01:15:16,210 on the market on the market sold ob no it's not 1271 01:15:16,210 --> 01:15:20,449 at least I mean we were races buyers information and tell you 1272 01:15:20,449 --> 01:15:24,379 I we never saw those datasets I think those datasets that they were made 1273 01:15:24,380 --> 01:15:25,219 available 1274 01:15:25,219 --> 01:15:30,469 I would there be a huge try again that's what the guy I mentioned the point that 1275 01:15:30,469 --> 01:15:33,130 he made from it with Anissa going back a dozen years 1276 01:15:33,130 --> 01:15:36,679 is that he was saying is that if people knew how much information 1277 01:15:36,679 --> 01:15:40,449 because from it you okay what books did you buy at the supermarket 1278 01:15:40,449 --> 01:15:43,589 you know arwood political leanings you have any kind of you know 1279 01:15:43,590 --> 01:15:47,670 social interest you have or don't have and that is that they said well if 1280 01:15:47,670 --> 01:15:52,219 koko's down by 10 cents you buy pepsi right miserable this type of thing 1281 01:15:52,219 --> 01:15:56,690 which was me scared everybody but that sorta follows up with 1282 01:15:56,690 --> 01:16:02,349 do you think that we're going to go to legislation and I think with the prior 1283 01:16:02,350 --> 01:16:05,420 questioner asked was that I am the owner 1284 01:16:05,420 --> 01:16:09,260 up some pieces of digital data whether it's my face 1285 01:16:09,260 --> 01:16:14,780 my search a log my public record my non public records that really don't need to 1286 01:16:14,780 --> 01:16:16,040 be public 1287 01:16:16,040 --> 01:16:20,340 arm do you see that really we need to define that because really at some point 1288 01:16:20,340 --> 01:16:22,000 people could say well my 1289 01:16:22,000 --> 01:16:25,210 DNA sequencing is my property is in it 1290 01:16:25,210 --> 01:16:29,810 yeah I thank you yeah II greats a question I think that 1291 01:16:29,810 --> 01:16:33,810 the question the the number the question is what is private and what is public 1292 01:16:33,810 --> 01:16:37,350 and I think there needs to be in there is right now 1293 01:16:37,350 --> 01:16:40,719 they did that private is generally private except when we 1294 01:16:40,719 --> 01:16:44,429 give it away and so supermarkets don't have 1295 01:16:44,429 --> 01:16:47,600 re are the the pain 1296 01:16:47,600 --> 01:16:52,219 affair in that information fireplaces is much much greater than the game 1297 01:16:52,219 --> 01:16:56,420 I online services the quaint little different 1298 01:16:56,420 --> 01:17:00,540 I would argue that your private 1299 01:17:00,540 --> 01:17:04,199 where online behavior is pretty pride 1300 01:17:04,199 --> 01:17:09,169 and yet we're giving away in exchange for services all the time 1301 01:17:09,170 --> 01:17:12,390 and you know that that worries 1302 01:17:12,390 --> 01:17:17,000 because I think that the supermarkets have and the change have 1303 01:17:17,000 --> 01:17:21,620 a much bigger vested interest in keeping your private information private 1304 01:17:21,620 --> 01:17:25,910 because it is so a revealing band 1305 01:17:25,910 --> 01:17:30,230 summer the online service and and the problem that I have with the online 1306 01:17:30,230 --> 01:17:31,910 service is really the 1307 01:17:31,910 --> 01:17:35,420 we you know we as users other because we like it we love 1308 01:17:35,420 --> 01:17:38,739 of free stuff great but it's not really free 1309 01:17:38,739 --> 01:17:43,839 and there's really no business model yet that summers at Wal 1310 01:17:43,840 --> 01:17:47,080 allow me to pay for services with money 1311 01:17:47,080 --> 01:17:50,910 instead of data and I think that's a problem 1312 01:17:50,910 --> 01:17:53,929 II center for last three years Facebook should at 1313 01:17:53,929 --> 01:17:56,929 a charge me option I 1314 01:17:56,929 --> 01:18:01,370 I think it's at odds with with I think the current business model online with 1315 01:18:01,370 --> 01:18:02,280 other online services 1316 01:18:02,280 --> 01:18:06,949 coming to pick on Facebook at all I but many online services are free 1317 01:18:06,949 --> 01:18:10,969 but they're not free and I think 1318 01:18:10,969 --> 01:18:14,960 unfortunately I people like to 1319 01:18:14,960 --> 01:18:18,170 up they don't like the privacy violations by they hate 1320 01:18:18,170 --> 01:18:22,620 paying with their money more and as I alluded to earlier 1321 01:18:22,620 --> 01:18:27,460 if you look at sort of the the present cost for the future privacy disaster 1322 01:18:27,460 --> 01:18:32,090 its 0 for most people equation they do is like well 1323 01:18:32,090 --> 01:18:35,320 twenty years ago block my faith but I really wanna 1324 01:18:35,320 --> 01:18:38,340 check out this new service online and I'm gonna do it 1325 01:18:38,340 --> 01:18:42,770 and so their that I think that dynamic has grab hold the bus 1326 01:18:42,770 --> 01:18:46,060 and there hasn't been anything that is shaking 1327 01:18:46,060 --> 01:18:50,070 I think the the latest I 1328 01:18:50,070 --> 01:18:54,349 government trespasses have helps advertisers that 1329 01:18:54,350 --> 01:18:57,730 to that a little bit I and maybe they'll be more drum beat around 1330 01:18:57,730 --> 01:19:01,830 backers may be over rotated for openness we're gonna guys iraq to back 1331 01:19:01,830 --> 01:19:05,910 for defining private spaces and a more comprehensive way 1332 01:19:05,910 --> 01:19:09,050 I do think we're going to get there privacy is not dead 1333 01:19:09,050 --> 01:19:12,199 this is a new technology I anyway 1334 01:19:12,199 --> 01:19:15,919 you've read II Jeff Jarvis writes about this the right about gutenberg in the 1335 01:19:15,920 --> 01:19:17,670 present burden to figure out 1336 01:19:17,670 --> 01:19:21,739 you know half millennia ago then even know what the printing presses point for 1337 01:19:21,739 --> 01:19:23,928 fifty years later after was invented 1338 01:19:23,929 --> 01:19:26,960 we are white and year maybe 1339 01:19:26,960 --> 01:19:32,150 20 up the Internet and so we're figuring this out as we go 1340 01:19:32,150 --> 01:19:35,739 we need to have you know ask tough questions 1341 01:19:35,739 --> 01:19:40,830 a the services that we use in an hour governor 1342 01:19:40,830 --> 01:19:47,670 and the of so if there are no more questions 1343 01:19:47,670 --> 01:19:52,300 at I about snow I after 1344 01:19:52,300 --> 01:19:55,780 after up so 1345 01:19:55,780 --> 01:19:59,259 I as I alluded to i you know 1346 01:19:59,260 --> 01:20:03,489 so I think we're all II don't speak for anyone I'm a little next month 1347 01:20:03,489 --> 01:20:08,330 on the one hand I feel like he clearly broke the law 1348 01:20:08,330 --> 01:20:11,489 I but he clearly 1349 01:20:11,489 --> 01:20:14,730 gave us some information we need and I'm 1350 01:20:14,730 --> 01:20:19,379 is a great quote from Martin Luther King 1351 01:20:19,380 --> 01:20:22,520 the cold front the jet just a bit is 1352 01:20:22,520 --> 01:20:27,179 I highest form respect for the law is to break the law banning face 1353 01:20:27,179 --> 01:20:30,239 your accuser and and he did 1354 01:20:30,239 --> 01:20:33,309 the first part but he didn't do the second part and 1355 01:20:33,310 --> 01:20:37,670 he would be much more hero its second part faces accuser became 1356 01:20:37,670 --> 01:20:41,710 I because I but I I do feel like we do need to have this national debate 1357 01:20:41,710 --> 01:20:44,969 about the security privacy tradeoff for making 1358 01:20:44,969 --> 01:20:48,370 I and where our society is going to be long term 1359 01:20:48,370 --> 01:20:52,800 because it's important that for democracy for work we need to know why 1360 01:20:52,800 --> 01:20:57,100 the people that are acting on our behalf are doing and so he gave us a glimpse 1361 01:20:57,100 --> 01:20:57,920 into what 1362 01:20:57,920 --> 01:21:01,219 they're doing and its I think its 1363 01:21:01,219 --> 01:21:04,790 its their hearts are in the right place to try to protect us 1364 01:21:04,790 --> 01:21:07,870 but it's a slippery slope Italian karen is Mike 1365 01:21:07,870 --> 01:21:12,659 here brought up I and the only way democracy I think operates 1366 01:21:12,659 --> 01:21:16,309 efficiently is education and transparency and 1367 01:21:16,310 --> 01:21:20,380 debate thanks everybody I think the speaker again please