1 00:00:00,120 --> 00:00:02,400 Speaker 1: Oh man, this is a doozy. This is one of 2 00:00:02,440 --> 00:00:10,040 Speaker 1: the ones that we we predicted nihilistically would become incredibly normalized. 3 00:00:10,080 --> 00:00:13,600 Speaker 1: Remember all these years ago facial recognition, Oh absolutely, when 4 00:00:13,600 --> 00:00:15,400 Speaker 1: it was just sort of like a glint in the 5 00:00:15,480 --> 00:00:18,560 Speaker 1: eye of of the Zuckster, you know, I mean like 6 00:00:18,760 --> 00:00:20,480 Speaker 1: it was about like, oh no, we're scared that they're 7 00:00:20,520 --> 00:00:23,759 Speaker 1: using facial recognition on our Facebook pictures. But now we're 8 00:00:23,800 --> 00:00:28,120 Speaker 1: actively feeding these AI apps like Lensa, you know, our 9 00:00:28,600 --> 00:00:34,320 Speaker 1: data so that they can improve facial recognition and recreation algorithms. Yeah, 10 00:00:34,360 --> 00:00:37,880 Speaker 1: and this is really an exploration of the tech that's 11 00:00:37,920 --> 00:00:41,360 Speaker 1: making it work, right. That's what we brought on Jonathan Strickland, 12 00:00:41,360 --> 00:00:44,240 Speaker 1: our buddy from Tech Stuff, to really go through all 13 00:00:44,280 --> 00:00:49,200 Speaker 1: of that minutia. And we're gonna listen back along with 14 00:00:49,240 --> 00:00:53,320 Speaker 1: you to see, uh, just how our predictions evened out. 15 00:00:53,520 --> 00:00:57,600 Speaker 1: So I hope you enjoy it and think about who 16 00:00:57,640 --> 00:01:01,520 Speaker 1: gets to look at your face on UFOs too, Psychic 17 00:01:01,560 --> 00:01:06,160 Speaker 1: powers and government conspiracies. History is riddled with unexplained events. 18 00:01:06,600 --> 00:01:09,720 Speaker 1: You can turn back now or learn the stuff they 19 00:01:09,720 --> 00:01:25,679 Speaker 1: don't want you to know. Welcome back to the show. 20 00:01:25,760 --> 00:01:28,959 Speaker 1: My name is Matt. They called me Ben, you are you, 21 00:01:29,040 --> 00:01:32,400 Speaker 1: and that makes this stuff they don't want you to know. Listeners, 22 00:01:32,440 --> 00:01:36,640 Speaker 1: you're probably wondering whatever happened to the other guy on 23 00:01:36,680 --> 00:01:40,319 Speaker 1: the show. Don't worry. Our third amigo, Nol will be 24 00:01:40,600 --> 00:01:44,759 Speaker 1: back in due time. Today, however, we've pulled a switcheroo 25 00:01:44,959 --> 00:01:48,559 Speaker 1: on you. You see, this is not the ordinary episode 26 00:01:48,760 --> 00:01:51,320 Speaker 1: of stuff they don't want you to know. Today we 27 00:01:51,360 --> 00:01:55,200 Speaker 1: are joined by a very special and returning guest, Ladies 28 00:01:55,200 --> 00:01:59,200 Speaker 1: and gentlemen, our resident tech expert, Jonathan Strickland. I'd like 29 00:01:59,240 --> 00:02:02,920 Speaker 1: to think this is an extraordinary episode of stuff they 30 00:02:02,960 --> 00:02:05,840 Speaker 1: don't want you to know. And uh, I am here 31 00:02:05,880 --> 00:02:09,400 Speaker 1: to talk about stuff they legit do not want you 32 00:02:09,440 --> 00:02:13,040 Speaker 1: to know. That's true. That's true, Jonathan, thank you for coming. 33 00:02:13,160 --> 00:02:16,440 Speaker 1: You know one thing, one thing that we can say 34 00:02:16,480 --> 00:02:19,880 Speaker 1: about Jonathan off the bat is that he has a 35 00:02:19,880 --> 00:02:23,079 Speaker 1: face that is more recognizable than most. You host a 36 00:02:23,120 --> 00:02:26,120 Speaker 1: show called tech stuff, you host a show called forward Thinking, 37 00:02:26,520 --> 00:02:30,639 Speaker 1: You've worked with Everyday Science on brain stuff, and you 38 00:02:30,880 --> 00:02:35,800 Speaker 1: regularly hit the streets on C, S and E three. Yep, 39 00:02:35,919 --> 00:02:39,560 Speaker 1: these are all true. I I often will get maybe 40 00:02:39,600 --> 00:02:44,120 Speaker 1: not often, but frequently enough, get recognized simply because my 41 00:02:44,200 --> 00:02:47,120 Speaker 1: face has been in a lot of places and associated 42 00:02:47,160 --> 00:02:50,400 Speaker 1: directly with my identity. It's not just that, hey, who's 43 00:02:50,440 --> 00:02:53,840 Speaker 1: that guy in the background of every Avengers movie or whatever. 44 00:02:54,160 --> 00:02:57,400 Speaker 1: I'm someone who my face is tied to who I 45 00:02:57,440 --> 00:03:01,800 Speaker 1: am frequently enough that people recognized me for the person 46 00:03:01,840 --> 00:03:06,560 Speaker 1: I am. Most people, I would argue, probably don't experience 47 00:03:06,600 --> 00:03:08,839 Speaker 1: that to any great degree, apart from you know, they're 48 00:03:08,840 --> 00:03:12,960 Speaker 1: they're close friends and acquaintances, that sort of thing whatever, 49 00:03:13,080 --> 00:03:15,600 Speaker 1: and their exes. Yeah, the enemies that they have ranked 50 00:03:15,680 --> 00:03:18,400 Speaker 1: up as people may or may not be aware. Ben 51 00:03:18,440 --> 00:03:22,640 Speaker 1: and I are longtime enemies. I declared you as such 52 00:03:22,800 --> 00:03:25,560 Speaker 1: within my first couple of months of working and how 53 00:03:25,600 --> 00:03:28,200 Speaker 1: stuff works well, yeah, but also to be fair, I 54 00:03:28,280 --> 00:03:33,440 Speaker 1: push people. Matt can attest to that. So most people 55 00:03:33,480 --> 00:03:36,840 Speaker 1: would think there's not a real great chance they would 56 00:03:36,840 --> 00:03:39,840 Speaker 1: be recognized whenever they go out in public. But here's 57 00:03:39,840 --> 00:03:41,840 Speaker 1: an interesting fact that you may or may not know. 58 00:03:42,000 --> 00:03:45,880 Speaker 1: In the United States, one out of every two adults 59 00:03:46,800 --> 00:03:52,120 Speaker 1: is actually their faces actually tied to their identity in 60 00:03:52,360 --> 00:03:58,600 Speaker 1: some law enforcements database. That law enforcement agency might be local, 61 00:03:58,880 --> 00:04:03,400 Speaker 1: might be federal. Most of them are linked together loosely 62 00:04:03,680 --> 00:04:06,440 Speaker 1: in the sense that if one law enforcement agency wants 63 00:04:06,480 --> 00:04:12,600 Speaker 1: to run a recognition UH test on a suspect photo, 64 00:04:13,200 --> 00:04:16,800 Speaker 1: they can get that done within other law enforcement agencies. 65 00:04:16,839 --> 00:04:21,400 Speaker 1: So you can, in theory, have half of the adult 66 00:04:21,440 --> 00:04:25,400 Speaker 1: population of the United States as your virtual lineup when 67 00:04:25,440 --> 00:04:28,960 Speaker 1: you are looking for a match on a photo. So, 68 00:04:29,080 --> 00:04:32,159 Speaker 1: if you are an adult in the United States, there's 69 00:04:32,160 --> 00:04:35,080 Speaker 1: a coin flip chance that your faces in one of 70 00:04:35,080 --> 00:04:37,840 Speaker 1: these databases. If you've committed a crime, or if you've 71 00:04:37,880 --> 00:04:41,040 Speaker 1: been accused of committing a crime in the United States, 72 00:04:41,360 --> 00:04:44,080 Speaker 1: there's essentially a chance that your face is in one 73 00:04:44,080 --> 00:04:47,440 Speaker 1: of these databases. So that's what we're talking about today. 74 00:04:47,520 --> 00:04:49,599 Speaker 1: Ladies and gentlemen. You will remember at the close of 75 00:04:49,600 --> 00:04:53,960 Speaker 1: our previous episode, we asked you to consider how many 76 00:04:54,000 --> 00:04:57,120 Speaker 1: times your face has appeared on a camera? Right, how 77 00:04:57,120 --> 00:05:00,960 Speaker 1: many times a day does a camera perceive your visage. 78 00:05:01,480 --> 00:05:04,760 Speaker 1: I would argue that a lot of people don't want 79 00:05:05,000 --> 00:05:10,520 Speaker 1: to be recognized casually for one reason or another, perfectly legitimate, 80 00:05:10,839 --> 00:05:16,800 Speaker 1: perfectly legal reasons. Yeah, if I'm if I'm walking around 81 00:05:16,960 --> 00:05:20,479 Speaker 1: a place and I just want to duck into a store, 82 00:05:20,839 --> 00:05:23,520 Speaker 1: I don't necessarily want the entire world to know that 83 00:05:23,640 --> 00:05:25,920 Speaker 1: I went into that store. Let's say that I'm shopping 84 00:05:26,040 --> 00:05:29,359 Speaker 1: for a surprise gift for my wife, totally innocent, nothing 85 00:05:29,400 --> 00:05:33,800 Speaker 1: wrong about that. Then I don't necessarily want that information 86 00:05:33,800 --> 00:05:35,960 Speaker 1: to get back to my wife and thus ruin the surprise. 87 00:05:36,600 --> 00:05:39,960 Speaker 1: Or it could be it could be anything that you 88 00:05:39,960 --> 00:05:43,080 Speaker 1: would argue is on the further end of the spectrum. 89 00:05:43,120 --> 00:05:49,200 Speaker 1: At any rate, we have a certain expectation of privacy 90 00:05:49,360 --> 00:05:51,760 Speaker 1: in here in the United States that I think is 91 00:05:51,839 --> 00:05:54,839 Speaker 1: largely based off of wishful thinking at this point. So 92 00:05:55,080 --> 00:06:03,400 Speaker 1: what is now philosophic the cool Yeah, waxing about privacy 93 00:06:03,480 --> 00:06:05,640 Speaker 1: or the nature of it or how long it's existed. 94 00:06:05,839 --> 00:06:09,920 Speaker 1: Is uh, it's sort of a hobby Matt and I 95 00:06:10,040 --> 00:06:14,560 Speaker 1: on the show. I guess we should start, as the 96 00:06:14,600 --> 00:06:18,560 Speaker 1: Mad Hatter said at the beginning. Sure, So, what what 97 00:06:18,600 --> 00:06:21,920 Speaker 1: do we mean when we say facial recognition? What is this? 98 00:06:22,160 --> 00:06:26,120 Speaker 1: So this was something that people were starting to work 99 00:06:26,120 --> 00:06:29,599 Speaker 1: on very seriously in the nineties and early two thousand's 100 00:06:29,680 --> 00:06:33,120 Speaker 1: because it required a great deal of processing power back 101 00:06:33,160 --> 00:06:35,920 Speaker 1: in those days to achieve what now we can do 102 00:06:36,160 --> 00:06:39,760 Speaker 1: much more simply with new types of computer architecture. But 103 00:06:39,839 --> 00:06:44,080 Speaker 1: your basic approach has been the same, which is that 104 00:06:44,680 --> 00:06:49,200 Speaker 1: you create a program and algorithm that identifies special anchor 105 00:06:49,320 --> 00:06:53,760 Speaker 1: points on an image, such as a face, and identifies 106 00:06:54,200 --> 00:06:56,800 Speaker 1: this looks like an eye. So I'm going to assume 107 00:06:56,960 --> 00:06:59,000 Speaker 1: that this is an eye. So here's the other eye, 108 00:06:59,320 --> 00:07:02,880 Speaker 1: here's the no is, here's the mouth. Based upon the 109 00:07:02,920 --> 00:07:07,000 Speaker 1: relationship of these various components within a face, I would 110 00:07:07,880 --> 00:07:11,600 Speaker 1: assign a numeric value more or less a numeric value 111 00:07:11,880 --> 00:07:14,960 Speaker 1: to this. It's called a face print, but it's essentially 112 00:07:15,320 --> 00:07:19,320 Speaker 1: the sum total of all the different features that that 113 00:07:19,360 --> 00:07:22,560 Speaker 1: particular algorithm is looking for. And just to get this 114 00:07:22,600 --> 00:07:24,080 Speaker 1: out of the way, there are a lot of different 115 00:07:24,080 --> 00:07:26,400 Speaker 1: algorithms out there, made by a lot of different companies, 116 00:07:26,400 --> 00:07:29,440 Speaker 1: so not everyone does it exactly the same way. Everyone 117 00:07:29,560 --> 00:07:32,679 Speaker 1: claims that there's the best, but they all use similar 118 00:07:32,720 --> 00:07:37,400 Speaker 1: but not identical methods. You then take this information, this 119 00:07:37,480 --> 00:07:41,880 Speaker 1: face print, this numeric value you have created, and you 120 00:07:42,040 --> 00:07:46,320 Speaker 1: run it against a database of previous face prints that 121 00:07:46,360 --> 00:07:50,520 Speaker 1: are tied to known individuals. So the picture you take 122 00:07:50,920 --> 00:07:54,160 Speaker 1: is called a probe photo. That is, the photo you 123 00:07:54,200 --> 00:07:56,280 Speaker 1: are probing the database to see if you can find 124 00:07:56,320 --> 00:07:58,920 Speaker 1: a match. You look and see if you find a 125 00:07:58,920 --> 00:08:02,520 Speaker 1: close numeric match match between the two. The probe photo 126 00:08:02,560 --> 00:08:05,200 Speaker 1: and all of your your collected images that are in 127 00:08:05,240 --> 00:08:08,200 Speaker 1: your database that you have, you know who those belong to. 128 00:08:08,920 --> 00:08:12,520 Speaker 1: So photo pro photo is of a person of interest. 129 00:08:12,600 --> 00:08:15,160 Speaker 1: Might not be a suspect, but it's certainly someone you 130 00:08:15,200 --> 00:08:16,920 Speaker 1: would like to talk to if you had a chance, 131 00:08:16,960 --> 00:08:22,160 Speaker 1: because they're involved in a investigation in some way. You 132 00:08:22,240 --> 00:08:25,680 Speaker 1: look against the database, you try and find a match numerically. 133 00:08:25,760 --> 00:08:28,840 Speaker 1: At that point, at least with most systems, but not 134 00:08:28,960 --> 00:08:32,880 Speaker 1: withal you would then have a human being, a real 135 00:08:32,960 --> 00:08:35,920 Speaker 1: life person, comparing these images to each other to see 136 00:08:35,920 --> 00:08:40,160 Speaker 1: if they actually represent a match of this photo looks 137 00:08:40,200 --> 00:08:42,120 Speaker 1: like yes, it is in fact the person who's in 138 00:08:42,160 --> 00:08:45,360 Speaker 1: the database. Because people can look very similar, people can 139 00:08:45,400 --> 00:08:49,400 Speaker 1: look very similar, and but depending upon how the algorithm 140 00:08:49,480 --> 00:08:54,439 Speaker 1: breaks down the face, it may mistakenly think that two 141 00:08:54,480 --> 00:08:56,920 Speaker 1: people are the same, even if you were to look 142 00:08:56,960 --> 00:08:59,280 Speaker 1: at them in person and say, well, that's clearly not 143 00:08:59,360 --> 00:09:01,240 Speaker 1: the same person, and I mean I can tell that's 144 00:09:01,280 --> 00:09:04,880 Speaker 1: not the same person. There are times where that happens. Uh. 145 00:09:04,960 --> 00:09:07,160 Speaker 1: It also has a lot of other factors, such as 146 00:09:07,280 --> 00:09:10,840 Speaker 1: the picture you take is the person facing face onto 147 00:09:10,840 --> 00:09:13,360 Speaker 1: the camera If it's at a slight angle, that changes 148 00:09:13,400 --> 00:09:16,920 Speaker 1: things as well. Right. Uh, there have been advances in 149 00:09:16,960 --> 00:09:20,360 Speaker 1: algorithms where it's able to take better guesses as to 150 00:09:20,400 --> 00:09:24,760 Speaker 1: the three dimensionality of a face. But previously it was 151 00:09:24,880 --> 00:09:27,800 Speaker 1: very much a two dimensional image, which meant that you 152 00:09:27,840 --> 00:09:31,280 Speaker 1: could fool it. You could fool it either on purpose 153 00:09:31,360 --> 00:09:34,280 Speaker 1: or by accident. So this doesn't have to do with 154 00:09:34,360 --> 00:09:39,439 Speaker 1: law enforcement. But in Japan they incorporated facial recognition software 155 00:09:39,600 --> 00:09:43,600 Speaker 1: in vending machines that sold cigarettes because the idea was 156 00:09:43,640 --> 00:09:47,520 Speaker 1: that if you appeared to be too young, the vending 157 00:09:47,520 --> 00:09:50,120 Speaker 1: machine wouldn't sell cigarettes to you. But if you just 158 00:09:50,240 --> 00:09:54,280 Speaker 1: held up a photo of someone who was old enough 159 00:09:54,360 --> 00:09:57,520 Speaker 1: to purchase cigarettes, it could Because it couldn't detect depth, 160 00:09:58,760 --> 00:10:02,240 Speaker 1: it was entirely just two dimensionals. Is this a face 161 00:10:02,320 --> 00:10:05,440 Speaker 1: of a person over what is eighteen or whatever? Right? 162 00:10:05,480 --> 00:10:08,280 Speaker 1: And now it's gotten to a point where because of 163 00:10:08,360 --> 00:10:10,400 Speaker 1: camera sophistication, there are a lot of cameras that can 164 00:10:10,440 --> 00:10:14,240 Speaker 1: incorporate two lenses, which gives it that that by optic 165 00:10:15,080 --> 00:10:18,400 Speaker 1: a view, a parallax view that can simulate yeah, exactly, 166 00:10:18,640 --> 00:10:21,720 Speaker 1: you can simulate that three dimensional vision. And there are 167 00:10:21,760 --> 00:10:24,520 Speaker 1: algorithms that can take advantage of that. That has been 168 00:10:24,559 --> 00:10:28,160 Speaker 1: decreased somewhat it is not as easy to fool, at 169 00:10:28,200 --> 00:10:33,240 Speaker 1: least a sophisticated machine, but it can still happen. Now. Earlier, 170 00:10:33,360 --> 00:10:41,000 Speaker 1: at the very beginning, you gave everybody a pretty harrowing statistic. Yeah, 171 00:10:41,240 --> 00:10:44,640 Speaker 1: the what did you call a coin flip chance. Yes, 172 00:10:44,679 --> 00:10:46,320 Speaker 1: there's a coin flip chance that if you are an 173 00:10:46,360 --> 00:10:48,120 Speaker 1: adult in the US, your face is in one of 174 00:10:48,160 --> 00:10:51,520 Speaker 1: these database. If you've never done anything wrong, yeah, you 175 00:10:51,559 --> 00:10:53,640 Speaker 1: don't have to do anything. But really, all you have 176 00:10:53,720 --> 00:10:57,240 Speaker 1: to have done is applied for some sort of license 177 00:10:57,880 --> 00:11:04,840 Speaker 1: or accreditation that also includes your photo passport, driver's license, 178 00:11:04,920 --> 00:11:07,480 Speaker 1: State I D. There are a lot of security clearances 179 00:11:07,520 --> 00:11:10,280 Speaker 1: that do this. So let's let's back up a little 180 00:11:10,280 --> 00:11:12,800 Speaker 1: bit and and let me explain the landscape, because it 181 00:11:12,880 --> 00:11:16,080 Speaker 1: is complicated, right. It's not as simple as there's this 182 00:11:16,160 --> 00:11:19,640 Speaker 1: one massive computer system at Langley and that's where your 183 00:11:19,640 --> 00:11:22,760 Speaker 1: face is. It's much more distributed than that, much like 184 00:11:22,800 --> 00:11:27,920 Speaker 1: the Internet itself. So the FBI has the Interstate Photo System. 185 00:11:27,960 --> 00:11:32,440 Speaker 1: This was sort of a an evolution of their national 186 00:11:32,600 --> 00:11:35,680 Speaker 1: fingerprint database. And I think you would say, all right, well, 187 00:11:35,720 --> 00:11:39,200 Speaker 1: I understand the fingerprint database. They've collected fingerprints from various 188 00:11:39,240 --> 00:11:43,439 Speaker 1: criminal investigations. This speeds things along when they are investigating 189 00:11:43,440 --> 00:11:47,720 Speaker 1: a new crime. This, however, is more about faces less 190 00:11:47,720 --> 00:11:52,080 Speaker 1: about fingerprints. So now what we're looking at is a 191 00:11:52,160 --> 00:11:56,479 Speaker 1: system where if the FBI uses their normal method of operations, 192 00:11:57,840 --> 00:12:00,600 Speaker 1: in theory, at least you probably would feel a little 193 00:12:00,679 --> 00:12:04,880 Speaker 1: less creepy about all this. And that's because if they're 194 00:12:04,920 --> 00:12:07,880 Speaker 1: using their basic method of running a search, what the 195 00:12:07,960 --> 00:12:11,560 Speaker 1: FBI would do is take a probe photo from an investigation. 196 00:12:12,040 --> 00:12:15,640 Speaker 1: They would then run it against their database, which includes 197 00:12:15,640 --> 00:12:20,560 Speaker 1: both criminal and civil photos. The civil photos some of 198 00:12:20,600 --> 00:12:23,400 Speaker 1: those are tied to some of the criminal photos. If 199 00:12:23,440 --> 00:12:25,640 Speaker 1: they have the same picture of a person from a 200 00:12:25,679 --> 00:12:29,880 Speaker 1: criminal investigation and from say a license that they had 201 00:12:29,920 --> 00:12:32,800 Speaker 1: applied for that then both of those will be tied 202 00:12:32,840 --> 00:12:35,880 Speaker 1: together in the database when they run a search. The 203 00:12:35,880 --> 00:12:39,160 Speaker 1: search only looks for matches in the criminal side. It 204 00:12:39,360 --> 00:12:42,079 Speaker 1: cannot look for matches on the civil side. So if 205 00:12:42,120 --> 00:12:47,720 Speaker 1: your ideas in the FBI's uh interstate photosystem, but it's 206 00:12:47,760 --> 00:12:51,680 Speaker 1: only in there as an example of a civil photo 207 00:12:51,760 --> 00:12:54,280 Speaker 1: and you have no criminal record to tie it to, 208 00:12:55,520 --> 00:12:59,040 Speaker 1: it should not pull you as a result. So this 209 00:12:59,160 --> 00:13:03,480 Speaker 1: leads to an immediate question. We're talking about the division 210 00:13:03,559 --> 00:13:09,360 Speaker 1: between civil and criminal images here facial images, at least 211 00:13:09,400 --> 00:13:13,240 Speaker 1: as it pertains to the FBI. Do they not automatically 212 00:13:13,320 --> 00:13:19,440 Speaker 1: correlate because of a matter of ability or because of 213 00:13:19,480 --> 00:13:24,679 Speaker 1: a matter of legislation. They do it as a policy 214 00:13:24,760 --> 00:13:27,920 Speaker 1: saying we we don't want to we don't want to 215 00:13:27,920 --> 00:13:30,360 Speaker 1: bring in the civil photos. Trust us. But this was 216 00:13:30,400 --> 00:13:33,520 Speaker 1: the basic search. I haven't gotten to their other search, 217 00:13:34,160 --> 00:13:38,280 Speaker 1: and we'll dive into that after a word from our sponsor, 218 00:13:49,000 --> 00:13:55,480 Speaker 1: and we're back. So for everybody who is just now 219 00:13:55,520 --> 00:13:58,000 Speaker 1: becoming acquainted with us, we've we've heard all of this 220 00:13:58,080 --> 00:14:01,559 Speaker 1: stuff people have seen like crime shows, c s I 221 00:14:01,760 --> 00:14:06,360 Speaker 1: or whatever, and uh, facial recognition and action films and 222 00:14:07,280 --> 00:14:11,679 Speaker 1: facial recognition and science fiction. And what we've talked about 223 00:14:11,720 --> 00:14:16,640 Speaker 1: so far is a basic search. We've explained the technology, 224 00:14:17,320 --> 00:14:20,360 Speaker 1: and we've explained a little bit about the evolution. I 225 00:14:20,360 --> 00:14:23,840 Speaker 1: think it would surprise people how quickly this has evolved. 226 00:14:26,120 --> 00:14:32,040 Speaker 1: But what's after the basic search? So all right, the 227 00:14:32,160 --> 00:14:37,080 Speaker 1: FBI S General m O is using this this criminal 228 00:14:37,160 --> 00:14:42,080 Speaker 1: database and avoiding the civil database, uh, when they're doing 229 00:14:42,080 --> 00:14:46,480 Speaker 1: their own searches. But they also have a division called 230 00:14:46,520 --> 00:14:52,880 Speaker 1: the Facial Analysis, Comparison and Evaluation Services or FACES want 231 00:14:52,880 --> 00:14:56,160 Speaker 1: to do. The acronymic FACES is kind of like there 232 00:14:56,400 --> 00:15:00,120 Speaker 1: their specialty division where they want to run a more 233 00:15:00,600 --> 00:15:05,720 Speaker 1: UH more extensive search. So the FACES operators can run 234 00:15:05,720 --> 00:15:09,040 Speaker 1: a search about that that incorporates both criminal and civil 235 00:15:09,720 --> 00:15:14,840 Speaker 1: UH sources, and they can work with local state law 236 00:15:14,920 --> 00:15:18,240 Speaker 1: enforcement agencies, will not just state, but tribal law enforcement 237 00:15:18,280 --> 00:15:22,520 Speaker 1: agencies as well. In the United States, they have a 238 00:15:23,960 --> 00:15:26,320 Speaker 1: they have this ability to work with all of them, 239 00:15:26,320 --> 00:15:28,520 Speaker 1: to partner with them. Um, it's kind of a tip 240 00:15:28,560 --> 00:15:30,800 Speaker 1: for tat sort of approach. Whenever anyone needs to run 241 00:15:30,800 --> 00:15:33,040 Speaker 1: one of these, they can run it up the chain 242 00:15:33,120 --> 00:15:36,120 Speaker 1: and see if they can get a bigger net to 243 00:15:36,120 --> 00:15:40,200 Speaker 1: to pull from. Right, So what will happen is you 244 00:15:40,240 --> 00:15:43,800 Speaker 1: would send a request to the FACES department. They could 245 00:15:43,840 --> 00:15:46,400 Speaker 1: run it against not only the FBI's database, which is 246 00:15:46,880 --> 00:15:50,720 Speaker 1: it's extensive, but that is not the one in every 247 00:15:50,760 --> 00:15:55,880 Speaker 1: two Americans databases. That that statistic comes from the collection 248 00:15:56,040 --> 00:15:59,400 Speaker 1: of databases across the United States at all different levels 249 00:15:59,400 --> 00:16:04,800 Speaker 1: of law enforcements. So again, state, tribal, city, UH, federal, 250 00:16:04,920 --> 00:16:07,680 Speaker 1: all of those different levels combined, that's where you get 251 00:16:07,720 --> 00:16:11,720 Speaker 1: the one in two. But the FBI essentially has access 252 00:16:11,760 --> 00:16:14,920 Speaker 1: to most of that because of these relationships they've developed 253 00:16:14,920 --> 00:16:18,080 Speaker 1: with various other entities. So you would send a request 254 00:16:18,200 --> 00:16:24,520 Speaker 1: to FACES. FACES would then handle the request sent out 255 00:16:24,560 --> 00:16:28,120 Speaker 1: to other agencies, right so within they would do the 256 00:16:27,920 --> 00:16:30,440 Speaker 1: all the FBI search, but they would also say, Hey, 257 00:16:30,680 --> 00:16:34,080 Speaker 1: State of Utah, can you do this search? City of 258 00:16:34,160 --> 00:16:36,320 Speaker 1: Las Vegas can you do this search? You know what 259 00:16:36,320 --> 00:16:39,000 Speaker 1: we're gonna expand this, California can you do this search? 260 00:16:39,320 --> 00:16:44,160 Speaker 1: And you start again casting a wider net. Well, you 261 00:16:44,200 --> 00:16:49,160 Speaker 1: start running into problems very very quickly with this approach. Well, 262 00:16:49,720 --> 00:16:53,680 Speaker 1: multiple multiple avenues. Let's start with technology. You remember earlier 263 00:16:53,720 --> 00:16:58,800 Speaker 1: I mentioned that no, no facial recognition algorithm does the 264 00:16:58,960 --> 00:17:01,760 Speaker 1: approach the exact same way as a different one, right like, 265 00:17:02,600 --> 00:17:06,439 Speaker 1: So it's not like there's just one way to achieve 266 00:17:06,440 --> 00:17:10,639 Speaker 1: facial recognition technology and some may be better than others. 267 00:17:10,760 --> 00:17:14,000 Speaker 1: And the real problem is that we don't really know 268 00:17:14,600 --> 00:17:17,919 Speaker 1: because most of these companies haven't had a third party 269 00:17:17,960 --> 00:17:21,480 Speaker 1: come in and evaluate the software for accuracy. So there's 270 00:17:21,520 --> 00:17:26,080 Speaker 1: no UM, there's no quality assurance, there's no accountability at all. 271 00:17:26,400 --> 00:17:30,960 Speaker 1: So one from a technology standpoint, you cannot be certain 272 00:17:31,000 --> 00:17:34,359 Speaker 1: that everyone is on the same playing ground, like they're 273 00:17:34,400 --> 00:17:37,760 Speaker 1: on the same level. They may all be using different technologies, 274 00:17:37,880 --> 00:17:41,080 Speaker 1: some of which may be more effective than others. That's 275 00:17:41,080 --> 00:17:43,320 Speaker 1: problem number one. Just to add on to that, one 276 00:17:43,359 --> 00:17:51,520 Speaker 1: of the larger problematic trends in many uh many government 277 00:17:51,520 --> 00:17:58,320 Speaker 1: agencies attempt to implement technology is the arrival and persistence 278 00:17:58,840 --> 00:18:04,919 Speaker 1: of outdated legacy stuff because they're locked into some service 279 00:18:04,960 --> 00:18:08,040 Speaker 1: agreement or contract for or they just don't have budget 280 00:18:08,280 --> 00:18:11,800 Speaker 1: for it. Like if you're running if you're running old servers, 281 00:18:11,880 --> 00:18:16,359 Speaker 1: old computers, and they aren't capable of running the most 282 00:18:16,440 --> 00:18:21,720 Speaker 1: up to date versions of software that's out there, then yeah, 283 00:18:21,760 --> 00:18:27,200 Speaker 1: that's yet another layer of technology issues. Yes, ladies and gentlemen, 284 00:18:27,280 --> 00:18:31,160 Speaker 1: if this sounds kind of weird to you, um, one 285 00:18:31,320 --> 00:18:35,680 Speaker 1: of the one of the moments in American pop culture 286 00:18:35,880 --> 00:18:42,920 Speaker 1: when the average non military, non law enforcement officer person 287 00:18:43,040 --> 00:18:46,040 Speaker 1: learned about this was actually watching the wire when they 288 00:18:46,040 --> 00:18:49,480 Speaker 1: had to use typewriters. Do you remember that? And this 289 00:18:49,600 --> 00:18:53,720 Speaker 1: is just another example of this. So not only are 290 00:18:55,280 --> 00:19:01,000 Speaker 1: their various forms of facial recognition algorithms and technology, some 291 00:19:01,080 --> 00:19:04,720 Speaker 1: of which may be proprietary, all of it is proprietary. 292 00:19:04,840 --> 00:19:09,879 Speaker 1: And not only do these not all's play nice together. 293 00:19:09,920 --> 00:19:15,400 Speaker 1: I guess or agree at times. But some are I'm 294 00:19:15,480 --> 00:19:20,639 Speaker 1: overwhelmingly certain. Some are pretty outmoded, Yeah, effectively obsolete. Some 295 00:19:20,680 --> 00:19:24,800 Speaker 1: of them are. And again you don't necessarily know unless 296 00:19:24,840 --> 00:19:28,639 Speaker 1: you are getting regular updates from all the different agencies 297 00:19:28,680 --> 00:19:31,200 Speaker 1: out there. There are thousands of these law enforcement agencies 298 00:19:31,320 --> 00:19:33,840 Speaker 1: right Like, it's not we're not talking about others. There's 299 00:19:33,880 --> 00:19:36,160 Speaker 1: fifty of them, because there's fifty states. No, there's any 300 00:19:36,240 --> 00:19:38,880 Speaker 1: more than that. And you're talking about databases that are 301 00:19:38,880 --> 00:19:41,760 Speaker 1: not necessarily directly connected to one another. So each one 302 00:19:41,880 --> 00:19:44,600 Speaker 1: is its own little island with its own little like. 303 00:19:44,680 --> 00:19:47,120 Speaker 1: Some of them may be using the same facial recognition software, 304 00:19:47,240 --> 00:19:49,560 Speaker 1: some of them may be using different versions of the 305 00:19:49,640 --> 00:19:54,680 Speaker 1: same software. Uh Now, if if they all worked, if 306 00:19:54,720 --> 00:19:57,960 Speaker 1: all the facial recognition software worked flawlessly, this would not 307 00:19:58,000 --> 00:19:59,639 Speaker 1: be an issue. It wouldn't matter if you went and 308 00:19:59,680 --> 00:20:02,480 Speaker 1: bought from Company A versus Company B, or if you 309 00:20:02,520 --> 00:20:05,400 Speaker 1: had Company a's one point over version versus company as 310 00:20:05,440 --> 00:20:08,440 Speaker 1: one point five version. If they all worked perfectly, that's 311 00:20:08,480 --> 00:20:11,280 Speaker 1: not an issue. But we don't know because we haven't 312 00:20:11,280 --> 00:20:15,040 Speaker 1: had these third party audits for a lot of those technologies. 313 00:20:15,080 --> 00:20:17,280 Speaker 1: You get a lot of claims like You'll see a 314 00:20:17,280 --> 00:20:21,080 Speaker 1: company say we're accurate, and then you say, well, well, 315 00:20:21,119 --> 00:20:24,640 Speaker 1: where's the proof of that? And they say trust us, 316 00:20:24,680 --> 00:20:26,720 Speaker 1: And then you say, well, can we hold you to 317 00:20:26,800 --> 00:20:28,760 Speaker 1: that in court? And like, oh no, our terms of 318 00:20:28,800 --> 00:20:32,520 Speaker 1: service say you can't. We can't be legally held accountable 319 00:20:32,560 --> 00:20:35,360 Speaker 1: for the claim that we make that it is accurate. 320 00:20:35,960 --> 00:20:38,440 Speaker 1: That's an issue, right, So that's all technology. There's another 321 00:20:38,480 --> 00:20:44,240 Speaker 1: technology issue that gets super uncomfortable. What's that most of 322 00:20:44,240 --> 00:20:48,639 Speaker 1: the time, facial recognition software is not equally accurate across 323 00:20:48,720 --> 00:20:56,040 Speaker 1: all ethnicities because the way people choose these algorithms and 324 00:20:56,080 --> 00:20:58,320 Speaker 1: the way they designed the faces, you know, the facial 325 00:20:58,359 --> 00:21:01,920 Speaker 1: recognition that the element that they're looking for. It may 326 00:21:01,920 --> 00:21:04,880 Speaker 1: be that in one ethnicity you see a greater variety 327 00:21:04,920 --> 00:21:09,560 Speaker 1: of the features that were chosen to be the elements 328 00:21:09,600 --> 00:21:13,679 Speaker 1: that you're looking for versus others. So within one ethnicity, 329 00:21:13,720 --> 00:21:17,040 Speaker 1: you might see a greater variation in something like cheekbones. 330 00:21:17,240 --> 00:21:22,320 Speaker 1: Let's say the cheekbone size, the cheekbone like, the prominence. Yeah, 331 00:21:23,000 --> 00:21:25,480 Speaker 1: that kind of stuff might be really important with or 332 00:21:25,520 --> 00:21:27,880 Speaker 1: at least there might be a greater variety within one 333 00:21:27,920 --> 00:21:31,639 Speaker 1: ethnicity and less in another. Well, if your technology is 334 00:21:31,680 --> 00:21:35,439 Speaker 1: focusing on that, and it doesn't focus so much on 335 00:21:35,480 --> 00:21:38,760 Speaker 1: the other qualities that have a greater variety within a 336 00:21:38,800 --> 00:21:41,760 Speaker 1: different ethnicity. You're gonna get a lot of false positives 337 00:21:42,359 --> 00:21:44,639 Speaker 1: because the technology you're looking at, if it's looking at 338 00:21:45,080 --> 00:21:48,240 Speaker 1: a subject from that ethnicity that doesn't have great representation 339 00:21:48,280 --> 00:21:51,240 Speaker 1: in your technology, you're gonna get a lot of false 340 00:21:51,520 --> 00:21:56,199 Speaker 1: positive identifications because the tech can It's like, it's like 341 00:21:56,440 --> 00:21:59,480 Speaker 1: that that statement about someone being racist. They cannot tell 342 00:21:59,520 --> 00:22:02,760 Speaker 1: the different between two different individuals from the same race. 343 00:22:02,800 --> 00:22:05,640 Speaker 1: They don't see the difference because they have become so 344 00:22:07,119 --> 00:22:12,159 Speaker 1: ethnically minded of their own ethnicity. They only recognize the 345 00:22:12,160 --> 00:22:16,800 Speaker 1: differences that are representative of their respect by ethnicity. And 346 00:22:16,880 --> 00:22:22,240 Speaker 1: this is what's what's fascinating and terrifying about this. I 347 00:22:22,800 --> 00:22:26,040 Speaker 1: think the most immediate fascinating and terrifying thing is that 348 00:22:26,119 --> 00:22:31,879 Speaker 1: we see again that software is limited by the creator. 349 00:22:32,040 --> 00:22:38,360 Speaker 1: Technology is limited by we all make these amazing tools 350 00:22:39,440 --> 00:22:42,479 Speaker 1: based on our own parameters. We very quickly make an 351 00:22:42,520 --> 00:22:46,040 Speaker 1: assumption that something that's that's working on what appears to 352 00:22:46,080 --> 00:22:49,439 Speaker 1: be cold hard logic is infallible, right like that that 353 00:22:49,560 --> 00:22:54,840 Speaker 1: emotion that things like uh a rationality don't play a 354 00:22:54,880 --> 00:22:56,680 Speaker 1: factor except for the fact that they played a factor 355 00:22:56,720 --> 00:23:00,400 Speaker 1: when we made this stuff, and it didn't necessarily from 356 00:23:00,400 --> 00:23:03,719 Speaker 1: a conscious decision. It may have been something that just 357 00:23:03,800 --> 00:23:06,440 Speaker 1: in the design elements people were trying to figure out, 358 00:23:06,440 --> 00:23:08,800 Speaker 1: all right, well, what components are the ones we need 359 00:23:08,840 --> 00:23:11,200 Speaker 1: to really focus on in order for our technology to work. 360 00:23:11,200 --> 00:23:13,720 Speaker 1: And some guy leaned in, took off his clan hood 361 00:23:13,720 --> 00:23:16,600 Speaker 1: and said, cheek bones. Yeah, that's a very I mean, 362 00:23:16,840 --> 00:23:20,200 Speaker 1: probably not quite as a as extreme as that, but 363 00:23:20,560 --> 00:23:24,920 Speaker 1: the point being that it there have been some some claims, 364 00:23:24,960 --> 00:23:27,760 Speaker 1: at least I don't know about any deep studies, but 365 00:23:27,880 --> 00:23:34,240 Speaker 1: claims that facial recognition technology disproportionately affects people of UH 366 00:23:34,720 --> 00:23:40,399 Speaker 1: specific ethnicities, particularly black people. They have been uh, disproportionately 367 00:23:40,440 --> 00:23:44,720 Speaker 1: affected by this false positives. Yeah, absolutely, yeah. And and 368 00:23:44,800 --> 00:23:48,919 Speaker 1: also there it's that goes down a road that is 369 00:23:50,000 --> 00:23:52,480 Speaker 1: far beyond the scope of this episode, because we'd have 370 00:23:52,480 --> 00:23:57,720 Speaker 1: to start talking about the vast gap uh in the 371 00:23:57,760 --> 00:24:00,359 Speaker 1: experience of being say, a white person in the United 372 00:24:00,440 --> 00:24:04,080 Speaker 1: States and being the subject of a law enforcement investigation 373 00:24:04,080 --> 00:24:07,480 Speaker 1: and being a black person. There's huge issues there that 374 00:24:07,560 --> 00:24:10,280 Speaker 1: really go beyond what I'm talking about here. Uh, And 375 00:24:10,320 --> 00:24:12,520 Speaker 1: I'm not gonna jump into that because I'm sure you've 376 00:24:12,560 --> 00:24:14,960 Speaker 1: covered those sort of things in previous episodes. If you're 377 00:24:15,000 --> 00:24:18,160 Speaker 1: interested in learning more about that, you can look at 378 00:24:18,200 --> 00:24:21,560 Speaker 1: the Center on Privacy and Technology at Georgetown Law. They 379 00:24:21,600 --> 00:24:25,080 Speaker 1: have a website called Perpetual Lineup dot org and it 380 00:24:25,160 --> 00:24:28,280 Speaker 1: goes through a lot of what you've been talking about. Yeah, 381 00:24:28,320 --> 00:24:30,840 Speaker 1: it's it is a real thing. I want to stress 382 00:24:30,920 --> 00:24:34,520 Speaker 1: that it is a real thing. People sometimes will think 383 00:24:34,560 --> 00:24:38,280 Speaker 1: that I've encountered people with resistance to it because they 384 00:24:38,280 --> 00:24:42,359 Speaker 1: don't like the idea of an inherently unfair system. It 385 00:24:42,440 --> 00:24:46,600 Speaker 1: goes against some of their um their own personal values, 386 00:24:46,600 --> 00:24:51,240 Speaker 1: which I completely understand. But it is a real thing. Yeah. Yeah. Well, also, 387 00:24:51,480 --> 00:24:57,320 Speaker 1: every this is gonna be Ben's blanket statement for the episode. 388 00:24:57,880 --> 00:25:00,879 Speaker 1: Always end up having one, and please write to me 389 00:25:00,920 --> 00:25:06,640 Speaker 1: if you disagree. All systems are inherently, on some level unfair. Yes, yeah, 390 00:25:06,680 --> 00:25:10,320 Speaker 1: because systems parse and we don't have we don't have 391 00:25:10,359 --> 00:25:15,040 Speaker 1: a perfect way of incorporating every possible life experience and 392 00:25:15,080 --> 00:25:18,960 Speaker 1: perspective when developing a system. Systems are meant to sort 393 00:25:19,040 --> 00:25:22,919 Speaker 1: things into different categories. Typically it's like a bureaucracy. If 394 00:25:22,920 --> 00:25:25,600 Speaker 1: you know how the bureaucracy works. You can very quickly 395 00:25:25,680 --> 00:25:28,640 Speaker 1: move through a bureaucracy, but if you don't, your experience 396 00:25:28,680 --> 00:25:33,560 Speaker 1: is going to be painfully laborious. And also, I mean, 397 00:25:33,800 --> 00:25:37,199 Speaker 1: what else can we expect if if every system is 398 00:25:37,359 --> 00:25:46,560 Speaker 1: inherently imperfect. We can't even as a species get our 399 00:25:46,720 --> 00:25:53,120 Speaker 1: collection of people's face pictures correct. And we're pouring millions 400 00:25:53,160 --> 00:25:57,200 Speaker 1: into this. Let me let me throw another issue here. Yes, 401 00:25:57,280 --> 00:26:00,680 Speaker 1: so we had technology, it's it's all propried terran doesn't 402 00:26:00,720 --> 00:26:06,960 Speaker 1: necessarily work well together. It's disproportionately discriminatory, leading to not 403 00:26:07,119 --> 00:26:10,400 Speaker 1: only false positives, but in some cases wrongful of rest 404 00:26:10,440 --> 00:26:14,159 Speaker 1: and imprisoned. And uh and Bennet's about to get a 405 00:26:14,200 --> 00:26:18,399 Speaker 1: whole lot worse. Yes, you're right, it's time for a 406 00:26:18,400 --> 00:26:32,359 Speaker 1: word from our sponsor. And we returned. It wasn't that bad. 407 00:26:32,480 --> 00:26:36,399 Speaker 1: I I actually chuckled a little bit, but it was 408 00:26:36,560 --> 00:26:41,120 Speaker 1: a tease. We weren't talking, weren't talking about the ad break. 409 00:26:41,440 --> 00:26:46,639 Speaker 1: We were talking about yet another disturbing aspect of the 410 00:26:46,680 --> 00:26:49,560 Speaker 1: emergence of facial recognition. Yeah, I wish, I wish I 411 00:26:49,560 --> 00:26:52,200 Speaker 1: could say there's only maybe one or two layers left, 412 00:26:52,240 --> 00:26:56,800 Speaker 1: But honestly, this, this is this is a mire of issues. 413 00:26:57,080 --> 00:27:01,320 Speaker 1: So remember I said, we have thousands different agencies, all 414 00:27:01,359 --> 00:27:06,320 Speaker 1: with various databases, all using various types of facial recognition software. 415 00:27:06,400 --> 00:27:09,600 Speaker 1: The FBI can tap many of them when doing one 416 00:27:09,680 --> 00:27:13,480 Speaker 1: of these searches. Um on top of this, you don't 417 00:27:13,560 --> 00:27:16,920 Speaker 1: have you don't have a lot of regulations in place 418 00:27:17,000 --> 00:27:21,280 Speaker 1: for most of these agencies on how they use this technology. 419 00:27:21,320 --> 00:27:26,240 Speaker 1: So there are no rules, which means there's no accountability 420 00:27:26,280 --> 00:27:28,440 Speaker 1: there either. It means that your face could be pulled 421 00:27:28,480 --> 00:27:32,360 Speaker 1: into a virtual lineup. Essentially, it could be pulled up 422 00:27:32,400 --> 00:27:35,880 Speaker 1: as a hit on one of these facial recognition searches. 423 00:27:36,760 --> 00:27:40,240 Speaker 1: And because many of these places have no rules or regulations, 424 00:27:41,080 --> 00:27:43,879 Speaker 1: you are not necessarily going to be alerted to this fact. 425 00:27:43,960 --> 00:27:46,240 Speaker 1: You may not be aware that your face is even 426 00:27:46,280 --> 00:27:49,639 Speaker 1: there in the first place. The first you may hear 427 00:27:49,680 --> 00:27:52,439 Speaker 1: of it is if it goes far enough for for 428 00:27:52,560 --> 00:27:56,160 Speaker 1: an agent to come and come knocking and ask like, hey, 429 00:27:56,720 --> 00:27:59,080 Speaker 1: your face popped up when we ran this search. We 430 00:27:59,119 --> 00:28:00,480 Speaker 1: need to talk to you about the thing that you 431 00:28:00,520 --> 00:28:05,960 Speaker 1: may or may not have any connection to an alibi better. So, 432 00:28:06,320 --> 00:28:11,199 Speaker 1: this is incredibly disturbing that because there are no rules 433 00:28:11,720 --> 00:28:16,480 Speaker 1: and no accountability, you can't as a citizen take any 434 00:28:16,520 --> 00:28:20,240 Speaker 1: steps because they didn't break rules. The rules didn't exist 435 00:28:20,400 --> 00:28:22,199 Speaker 1: before we started recording. I said, it's kind of like 436 00:28:22,200 --> 00:28:24,239 Speaker 1: if there were no laws, I could walk into your 437 00:28:24,280 --> 00:28:27,280 Speaker 1: house and take whatever I wanted because there's no law 438 00:28:27,320 --> 00:28:30,240 Speaker 1: against stealing as long as I have. If might makes right, 439 00:28:30,400 --> 00:28:33,680 Speaker 1: then this is your time because there's only policy. Yeah, 440 00:28:33,760 --> 00:28:37,400 Speaker 1: there's there's policy internally for these agencies, and even that 441 00:28:37,440 --> 00:28:39,840 Speaker 1: doesn't match up from agency the agency. Yeah, there are 442 00:28:39,920 --> 00:28:45,120 Speaker 1: eighteen states that have memorandums of agreement on this with 443 00:28:45,200 --> 00:28:48,360 Speaker 1: the FBI specifically, and and there are more states that 444 00:28:48,400 --> 00:28:51,840 Speaker 1: they're looking at adding to this all the time. But again, 445 00:28:52,400 --> 00:28:55,680 Speaker 1: within each individual agency, they may have very different policies. 446 00:28:55,800 --> 00:28:58,920 Speaker 1: Some of them have been trying to be responsible, right, 447 00:28:59,160 --> 00:29:02,080 Speaker 1: some of them have of uh, done third party audits 448 00:29:02,080 --> 00:29:04,520 Speaker 1: of their systems that sort of thing. That's great, that's 449 00:29:04,520 --> 00:29:07,400 Speaker 1: at least a step in the right direction, But until 450 00:29:07,480 --> 00:29:11,840 Speaker 1: we actually make rules, either at the state or federal level, 451 00:29:12,680 --> 00:29:16,240 Speaker 1: we don't have any again, any accountability of something were 452 00:29:16,280 --> 00:29:18,480 Speaker 1: to go wrong, Like, how do you complain if there's 453 00:29:18,520 --> 00:29:23,040 Speaker 1: no rule against it? Right? Well, one of the policies 454 00:29:23,040 --> 00:29:27,720 Speaker 1: that some of these agencies have is that before you 455 00:29:27,760 --> 00:29:32,160 Speaker 1: can return results to the FBI or to the investigative agent, 456 00:29:33,080 --> 00:29:36,200 Speaker 1: it has to first pass through humans who look at 457 00:29:36,200 --> 00:29:39,480 Speaker 1: the pictures and determine whether or not they make a 458 00:29:39,520 --> 00:29:42,800 Speaker 1: good match. So each of these agencies also have different 459 00:29:43,240 --> 00:29:47,040 Speaker 1: agreements as to how many results they will return to 460 00:29:47,760 --> 00:29:51,960 Speaker 1: whatever agent right like, uh, In the FBI, it's I 461 00:29:52,000 --> 00:29:56,280 Speaker 1: believe between two and fifty, with the default being twenty results. 462 00:29:56,760 --> 00:29:59,160 Speaker 1: But some places they'll be like no, we we will 463 00:29:59,200 --> 00:30:01,120 Speaker 1: send you one, or will send you too, or we'll 464 00:30:01,120 --> 00:30:04,400 Speaker 1: send you up to eighty results. And then human beings 465 00:30:04,440 --> 00:30:07,680 Speaker 1: are supposed to call through these and decide which ones 466 00:30:07,720 --> 00:30:12,440 Speaker 1: are actually the more uh likely matches to the one 467 00:30:12,480 --> 00:30:14,920 Speaker 1: that the computer found. You're ready to have your mind 468 00:30:14,920 --> 00:30:21,080 Speaker 1: blown here, Ben, Let's say, okay, all right, Let's say 469 00:30:21,120 --> 00:30:26,160 Speaker 1: that your picture has come up in one of these investigations. 470 00:30:27,000 --> 00:30:30,320 Speaker 1: Let's say that the agency that did it does have 471 00:30:30,440 --> 00:30:33,400 Speaker 1: this policy that it must be reviewed by a human 472 00:30:33,440 --> 00:30:37,760 Speaker 1: being before it goes any further. Fifty of the time 473 00:30:37,800 --> 00:30:41,480 Speaker 1: they're going to make the wrong call the human being 474 00:30:41,880 --> 00:30:46,760 Speaker 1: because this is hard. It's not just hard for computers, 475 00:30:46,800 --> 00:30:49,560 Speaker 1: it's hard for people to take. If you're not looking 476 00:30:49,600 --> 00:30:53,080 Speaker 1: at the exact same photos side by side. If you're 477 00:30:53,080 --> 00:30:55,000 Speaker 1: looking at a person on a different day, from a 478 00:30:55,040 --> 00:30:57,840 Speaker 1: different angle, from a different distance, under different lighting conditions, 479 00:30:57,840 --> 00:31:02,200 Speaker 1: wearing different clothing, wearing a different hairstyle, wearing whatever, it's 480 00:31:02,280 --> 00:31:05,719 Speaker 1: hard like unless you are really familiar with that person. 481 00:31:06,120 --> 00:31:08,360 Speaker 1: I mean, I've seen pictures of people who look like 482 00:31:08,480 --> 00:31:11,040 Speaker 1: me where I looked and I thought, I don't remember 483 00:31:11,080 --> 00:31:13,240 Speaker 1: ever being there, and then as I looked, like, oh wait, 484 00:31:13,240 --> 00:31:17,960 Speaker 1: that's not me. That has happened just to me, don't 485 00:31:17,960 --> 00:31:20,320 Speaker 1: beat yourself up. I went home with the wrong girl, Like, yeah, 486 00:31:20,400 --> 00:31:23,720 Speaker 1: I mean, I've definitely walked into the wrong house. But 487 00:31:23,760 --> 00:31:26,120 Speaker 1: that's a totally different issue. That has nothing to do 488 00:31:26,160 --> 00:31:27,760 Speaker 1: with my recognition. That just has to do with me 489 00:31:27,800 --> 00:31:30,000 Speaker 1: not paying attention. That was back before you realize that 490 00:31:30,080 --> 00:31:32,640 Speaker 1: Steely was against the law. Yeah, you know, people need 491 00:31:32,680 --> 00:31:34,200 Speaker 1: to make these things a little more clear to me. 492 00:31:34,800 --> 00:31:39,840 Speaker 1: Challenging stuff. So it turns out that you people really 493 00:31:39,880 --> 00:31:45,280 Speaker 1: need to go through a very thorough, challenging, difficult training 494 00:31:45,440 --> 00:31:51,080 Speaker 1: process to become experts in matching photos. If you have 495 00:31:51,160 --> 00:31:53,840 Speaker 1: not done that, and a lot of agencies don't require 496 00:31:53,960 --> 00:31:59,360 Speaker 1: that level of training. There's a fifty error rate so 497 00:31:59,800 --> 00:32:05,720 Speaker 1: not only could the machines misidentify somebody, the human beings 498 00:32:05,720 --> 00:32:10,000 Speaker 1: who are meant to be the the checks and balances 499 00:32:10,200 --> 00:32:13,280 Speaker 1: against such things could make the same mistake. So too 500 00:32:13,320 --> 00:32:16,280 Speaker 1: bad coin flips. You're in a lineup, you might be 501 00:32:16,400 --> 00:32:21,959 Speaker 1: matched as someone of interest. Now you might say, let's say, 502 00:32:22,000 --> 00:32:23,680 Speaker 1: let's say that you're the kind of person who says, 503 00:32:24,080 --> 00:32:26,640 Speaker 1: I didn't do anything wrong to worry. What do I 504 00:32:26,680 --> 00:32:31,400 Speaker 1: have to worry about? Well, I mean, you're talking about 505 00:32:31,480 --> 00:32:35,920 Speaker 1: a violation of a basic amendment here, potentially the protection 506 00:32:35,920 --> 00:32:40,480 Speaker 1: against unreasonable search and seizure. You're being searched with no justification. 507 00:32:40,640 --> 00:32:43,600 Speaker 1: There's no reason for them to search you, but they're 508 00:32:43,640 --> 00:32:46,600 Speaker 1: going to be looking into you and your background. Let's 509 00:32:46,600 --> 00:32:49,960 Speaker 1: say that you did do something that was of minor 510 00:32:50,840 --> 00:32:54,520 Speaker 1: significance in the past that maybe you know you're sorry 511 00:32:54,520 --> 00:32:57,880 Speaker 1: for you did something dumb. You you are sorry for it, 512 00:32:57,920 --> 00:33:00,840 Speaker 1: you have reformed, you're not you know, you luckily weren't 513 00:33:00,880 --> 00:33:03,880 Speaker 1: taken to task for it. I guarantee you if there's 514 00:33:03,880 --> 00:33:06,360 Speaker 1: a federal investigation, that's gonna be one of the things 515 00:33:06,440 --> 00:33:09,160 Speaker 1: that pops up. I mean, it's it's things like your 516 00:33:09,360 --> 00:33:13,600 Speaker 1: Your life could be ruined or at least significantly impacted 517 00:33:13,600 --> 00:33:17,120 Speaker 1: in a negative way for something of minor importance that 518 00:33:17,160 --> 00:33:19,400 Speaker 1: may have happened in the past that has nothing at 519 00:33:19,440 --> 00:33:22,760 Speaker 1: all to do with whatever the investigation is. And also, 520 00:33:23,520 --> 00:33:29,040 Speaker 1: let's let's go a step further here. Sure, so the 521 00:33:29,120 --> 00:33:35,320 Speaker 1: practice of expunging a criminal conviction, right, that happens a 522 00:33:35,320 --> 00:33:39,320 Speaker 1: lot with especially with people who commit a crime when 523 00:33:39,320 --> 00:33:44,920 Speaker 1: they're teenagers. Right. Sure that whether whether or not that's 524 00:33:44,960 --> 00:33:51,680 Speaker 1: expunged doesn't really matter in today's investigatory climate. It's kind 525 00:33:51,720 --> 00:33:54,120 Speaker 1: of like that game where people pretend the floor is 526 00:33:54,240 --> 00:33:59,200 Speaker 1: lava and no one steps on the floor. I'm sorry 527 00:33:59,240 --> 00:34:01,560 Speaker 1: you have to learn about at this the editorial board, 528 00:34:01,640 --> 00:34:05,520 Speaker 1: the floor is lava over at the editorial department. We Um, 529 00:34:05,600 --> 00:34:08,040 Speaker 1: we started that joke. Um, that's why we have the 530 00:34:08,080 --> 00:34:10,879 Speaker 1: carpet tiles. Ben, about two weeks before you got here. 531 00:34:11,000 --> 00:34:14,960 Speaker 1: Shut up, just kept going. You don't have they're not 532 00:34:15,239 --> 00:34:21,200 Speaker 1: special shoes. Man. On the flip side, My hamstrings are amazing. 533 00:34:21,320 --> 00:34:26,200 Speaker 1: That's true. That's true. Alright, So, um, we're adding some levity. 534 00:34:26,239 --> 00:34:32,360 Speaker 1: But we're adding levity. Yeah, you have to because and 535 00:34:32,440 --> 00:34:35,040 Speaker 1: I am not you know me, Ben, I am not 536 00:34:35,320 --> 00:34:40,320 Speaker 1: a They're all out to get you kind of guy. No, no, you. 537 00:34:40,320 --> 00:34:43,839 Speaker 1: You endeavor to be as objective as possible when you're 538 00:34:44,040 --> 00:34:48,640 Speaker 1: discussing anything tech related. I just feel very strongly that 539 00:34:48,800 --> 00:34:56,120 Speaker 1: this particular practice is dangerous and irresponsible on multiple levels. 540 00:34:56,120 --> 00:34:59,920 Speaker 1: And I'm not the only one. Congress called the FBI 541 00:35:00,000 --> 00:35:02,959 Speaker 1: I in front of the principal's office back in March 542 00:35:03,440 --> 00:35:06,080 Speaker 1: seventeen and essentially was saying the same thing they were saying, 543 00:35:06,440 --> 00:35:11,080 Speaker 1: there's no accountability here. Uh. This is horrifying. It is 544 00:35:11,120 --> 00:35:17,840 Speaker 1: a gross uh over overstepping of your authority. It Uh, 545 00:35:18,200 --> 00:35:23,280 Speaker 1: it is violating people's privacy. People should have an expectation 546 00:35:23,640 --> 00:35:27,000 Speaker 1: to not be called into these virtual lineups, even if 547 00:35:27,040 --> 00:35:29,200 Speaker 1: you never know about it. The fact that it's happening 548 00:35:29,920 --> 00:35:33,799 Speaker 1: is disturbing, right. I wonder if it qualifies on your 549 00:35:33,920 --> 00:35:37,000 Speaker 1: unreasonable search that and that's one of the arguments. That's 550 00:35:37,120 --> 00:35:40,680 Speaker 1: Georgetown University has has a great report on this that 551 00:35:40,800 --> 00:35:43,600 Speaker 1: is very easy to read. Uh. There's also the Government 552 00:35:43,600 --> 00:35:47,360 Speaker 1: Accountability Office that has an amazing report specifically about the FBI. 553 00:35:47,400 --> 00:35:52,120 Speaker 1: Georgetown looks at the broader picture. Government Accountability Office looks 554 00:35:52,160 --> 00:35:55,160 Speaker 1: specifically at the FBI's use because it's a federal agency 555 00:35:55,200 --> 00:35:57,760 Speaker 1: in the Government Accountability Offices is also a federal agency. 556 00:35:58,160 --> 00:36:01,360 Speaker 1: The Georgetown one looks at the broader use of facial 557 00:36:01,360 --> 00:36:05,080 Speaker 1: recognition and law enforcement in general, and they there are 558 00:36:05,120 --> 00:36:08,880 Speaker 1: some very strong points saying this seems like it violates 559 00:36:08,880 --> 00:36:12,719 Speaker 1: our protection against unreasonable search and seizure, and that you 560 00:36:12,760 --> 00:36:15,640 Speaker 1: know that that's a huge deal. That's that that's a 561 00:36:15,840 --> 00:36:19,440 Speaker 1: protection granted to us by the Constitution of the United States. 562 00:36:19,520 --> 00:36:25,080 Speaker 1: So there's definitely a movement in government to push these 563 00:36:25,120 --> 00:36:30,759 Speaker 1: agencies to develop regulations. And I suspect that if the 564 00:36:30,800 --> 00:36:34,360 Speaker 1: agencies failed to do so in a timely manner, we 565 00:36:34,400 --> 00:36:39,320 Speaker 1: will see actual legislative movement in various state and federal 566 00:36:39,800 --> 00:36:44,960 Speaker 1: governing bodies. Two codify it at that level, if not 567 00:36:45,120 --> 00:36:48,680 Speaker 1: at the actual agency level. Uh. And I would prefer 568 00:36:48,719 --> 00:36:51,319 Speaker 1: it to be codified at the state and federal level 569 00:36:51,520 --> 00:36:54,799 Speaker 1: because I always worry about any group that's allowed to 570 00:36:54,800 --> 00:36:58,240 Speaker 1: make up its own rules. Yeah. And and I find 571 00:36:58,280 --> 00:37:03,040 Speaker 1: it optimistic to to say that someone would limit this 572 00:37:03,280 --> 00:37:11,160 Speaker 1: ability because we know, for instance, that historically when aggregation 573 00:37:11,320 --> 00:37:18,040 Speaker 1: of powerful data bases occurs, it almost never just due 574 00:37:18,040 --> 00:37:22,400 Speaker 1: to the different paces of technological evolution and legislative evolution, 575 00:37:23,239 --> 00:37:28,239 Speaker 1: the It almost never is reigned back efficiently. I mean 576 00:37:28,320 --> 00:37:34,000 Speaker 1: like the like the right now. If, for instance, I 577 00:37:34,040 --> 00:37:37,680 Speaker 1: was that I decided to pull it Edwards Snowden or 578 00:37:37,760 --> 00:37:42,520 Speaker 1: Julian Assang or like what's a good bond villain? Doctor okay, 579 00:37:42,560 --> 00:37:44,480 Speaker 1: doctor no, I was like, I'm gonna do a doctor 580 00:37:44,520 --> 00:37:49,160 Speaker 1: no thing or rs Goldfinger, Yes, yes, I'm gonna do 581 00:37:49,239 --> 00:37:54,080 Speaker 1: a doctor no goldfinger thing. And I'm like, Matt Jonathan 582 00:37:54,120 --> 00:37:58,520 Speaker 1: and I are going to do some some terrible attack 583 00:37:58,640 --> 00:38:02,600 Speaker 1: on the US from our secret underwater base. Of course, 584 00:38:02,960 --> 00:38:06,160 Speaker 1: of course that's where we would do that. The volcano 585 00:38:06,200 --> 00:38:09,240 Speaker 1: base is nice. But if you're doing like a world 586 00:38:09,320 --> 00:38:13,040 Speaker 1: level attack where you're you're legitimately causing vast destruction, you 587 00:38:13,040 --> 00:38:15,440 Speaker 1: want to be in that underwater base. You genuinely have 588 00:38:15,520 --> 00:38:17,879 Speaker 1: to be afraid of floor lava in the volcano base, 589 00:38:17,960 --> 00:38:21,640 Speaker 1: just you really do. Those carpet tiles are a lifesaver. 590 00:38:22,600 --> 00:38:25,880 Speaker 1: Your ham strings do look great. So what What What 591 00:38:26,080 --> 00:38:33,200 Speaker 1: happened though, is uh, I would quickly learn that whatever 592 00:38:33,840 --> 00:38:41,000 Speaker 1: legislation exists will not in practice stop uh stop the 593 00:38:41,120 --> 00:38:46,080 Speaker 1: n essay from knowing everything I've done online and from 594 00:38:46,200 --> 00:38:51,480 Speaker 1: extrapolating in the interest of national security, any other information. 595 00:38:52,040 --> 00:38:56,759 Speaker 1: The FISA courts are for the birds. Yeah let me 596 00:38:56,960 --> 00:38:58,960 Speaker 1: let me let me throw one more. I would make 597 00:38:59,000 --> 00:39:02,560 Speaker 1: it worse. Okay, you wanted to be worse. So you 598 00:39:02,600 --> 00:39:06,279 Speaker 1: know I mentioned earlier that that idea that if I 599 00:39:06,320 --> 00:39:09,680 Speaker 1: haven't done anything wrong, I'm all right. Here in the 600 00:39:09,760 --> 00:39:13,080 Speaker 1: United States, one of the things that we are allowed 601 00:39:13,080 --> 00:39:16,960 Speaker 1: to do is protected by the Constitution, is to assemble 602 00:39:17,040 --> 00:39:23,040 Speaker 1: peacefully in public and express free speech. So there has 603 00:39:23,120 --> 00:39:28,080 Speaker 1: been a fear, justifiably in my mind, that an agency 604 00:39:28,080 --> 00:39:34,279 Speaker 1: that perhaps has a specific political inclination could use such 605 00:39:34,320 --> 00:39:38,880 Speaker 1: facial recognition software to identify participants in a peaceful protest 606 00:39:39,680 --> 00:39:47,040 Speaker 1: for later let us say discussions. Right, you don't think 607 00:39:47,040 --> 00:39:50,400 Speaker 1: that's already been done. What I'm here to say is 608 00:39:50,440 --> 00:39:55,160 Speaker 1: that Georgetown, when they did their report, they asked various 609 00:39:55,200 --> 00:39:58,560 Speaker 1: agencies about their policies as to the use of facial 610 00:39:58,600 --> 00:40:02,440 Speaker 1: recognition software. They found out of the fifty two agencies 611 00:40:02,520 --> 00:40:07,080 Speaker 1: that they were able to contact and get a response from, 612 00:40:07,239 --> 00:40:12,160 Speaker 1: only one has a specific rule against using facial recognition 613 00:40:12,160 --> 00:40:15,960 Speaker 1: to identify people participating in public demonstrations or free speech 614 00:40:16,040 --> 00:40:19,040 Speaker 1: in general. Only one out of fifty two that they 615 00:40:19,080 --> 00:40:23,720 Speaker 1: asked specifically has a rule against that. So, in other words, 616 00:40:23,880 --> 00:40:27,360 Speaker 1: fifty one out of fifty two have no rule. It 617 00:40:27,360 --> 00:40:29,520 Speaker 1: means that if the police were to use it, or 618 00:40:29,560 --> 00:40:33,719 Speaker 1: whatever other law enforcement agency were to use facial recognition, 619 00:40:33,880 --> 00:40:35,640 Speaker 1: perhaps even in a live feed, because you can do 620 00:40:35,640 --> 00:40:37,239 Speaker 1: it in video just as well as you could do 621 00:40:37,280 --> 00:40:41,520 Speaker 1: it in photographs, to start identifying to populate a crowd 622 00:40:41,560 --> 00:40:45,560 Speaker 1: with names and identities who are protesting a particular thing. 623 00:40:45,640 --> 00:40:49,080 Speaker 1: Let's say it's police brutality, which would be very relevant, 624 00:40:49,160 --> 00:40:53,520 Speaker 1: especially in recent years. That is a legitimate concern that 625 00:40:53,520 --> 00:40:57,719 Speaker 1: that people's free speech could be, uh could be could 626 00:40:57,719 --> 00:41:00,640 Speaker 1: be violated. If you feel like you can't go out 627 00:41:00,640 --> 00:41:05,000 Speaker 1: there and express your thoughts because you might be unfairly 628 00:41:05,040 --> 00:41:08,400 Speaker 1: targeted as a result of that, that is a violation 629 00:41:08,440 --> 00:41:10,759 Speaker 1: of your free speech, not to mention, you know, your 630 00:41:10,760 --> 00:41:14,640 Speaker 1: personal safety, depending upon how legit those fears are. So 631 00:41:16,040 --> 00:41:19,520 Speaker 1: that is incredibly troubling. I don't know if you noticed this, 632 00:41:19,680 --> 00:41:23,880 Speaker 1: Ben Matt. Maybe you guys don't notice it. I've noticed 633 00:41:23,920 --> 00:41:29,640 Speaker 1: that there have been a couple of protests recently, in 634 00:41:29,680 --> 00:41:31,680 Speaker 1: the last twelve months or so. I've seen a few, 635 00:41:32,040 --> 00:41:35,239 Speaker 1: like some of some of which have had a significant 636 00:41:35,280 --> 00:41:39,640 Speaker 1: number of participants, both on both on different sides of 637 00:41:39,640 --> 00:41:42,239 Speaker 1: political issues, right, not just I'm not saying that this 638 00:41:42,360 --> 00:41:45,600 Speaker 1: is one group targeting one other group. I'm saying there 639 00:41:45,600 --> 00:41:48,319 Speaker 1: are a lot of different parties involved here, and no 640 00:41:48,360 --> 00:41:52,239 Speaker 1: matter how this technology is used against whichever group it 641 00:41:52,360 --> 00:41:55,080 Speaker 1: is used, the fact that there are no rules about 642 00:41:55,080 --> 00:41:58,680 Speaker 1: how to use it is deeply troubling, and I would 643 00:41:58,800 --> 00:42:03,040 Speaker 1: argue fundament to Lee un American, I would agree, what 644 00:42:03,080 --> 00:42:07,520 Speaker 1: do you think, Matt, Yeah, definitely. I've seen the photographs 645 00:42:07,520 --> 00:42:11,920 Speaker 1: and videos of police filming using video cameras to film 646 00:42:12,040 --> 00:42:18,160 Speaker 1: protests and protesters, and that makes me very nervous. And honestly, 647 00:42:18,280 --> 00:42:21,440 Speaker 1: if you're able to get at those images, then you 648 00:42:21,440 --> 00:42:24,920 Speaker 1: can run that search, even if you weren't the one 649 00:42:24,960 --> 00:42:28,440 Speaker 1: filming it. Right. Like, one of the elements we didn't 650 00:42:28,440 --> 00:42:31,239 Speaker 1: really touch on in this in this episode, but that 651 00:42:31,560 --> 00:42:33,839 Speaker 1: is related, is the fact that we have lots of 652 00:42:33,880 --> 00:42:38,239 Speaker 1: private companies out there that incorporate facial recognition technology as 653 00:42:38,280 --> 00:42:41,719 Speaker 1: a feature for you, for consumers. Right. So Facebook is 654 00:42:41,760 --> 00:42:43,719 Speaker 1: a great example. You take a picture of someone on 655 00:42:43,760 --> 00:42:47,080 Speaker 1: Facebook that you're friends with. Facebook will often say, hey, 656 00:42:47,239 --> 00:42:49,120 Speaker 1: do you want to tag this photo? And they'll even 657 00:42:49,160 --> 00:42:51,200 Speaker 1: give you the name of the person because their facial 658 00:42:51,200 --> 00:42:54,160 Speaker 1: recognition software is good enough to let you know, like, oh, well, 659 00:42:54,160 --> 00:42:56,200 Speaker 1: we're pretty sure this is your buddy, you know, Bob, 660 00:42:56,520 --> 00:42:58,000 Speaker 1: So you want us to go ahead and tag Bob 661 00:42:58,080 --> 00:43:02,240 Speaker 1: for you, Well, you know, on its own, as a feature, 662 00:43:02,520 --> 00:43:06,040 Speaker 1: there's nothing necessarily wrong with that. But you could imagine 663 00:43:06,840 --> 00:43:11,839 Speaker 1: that same sort of of deep, deep wealth of data 664 00:43:12,080 --> 00:43:15,520 Speaker 1: being put to nefarious use under the right circumstances, maybe 665 00:43:15,520 --> 00:43:18,560 Speaker 1: not by Facebook itself, but perhaps by an entity that 666 00:43:18,600 --> 00:43:21,560 Speaker 1: forces Facebook to hand that information over exactly. And I'm 667 00:43:21,560 --> 00:43:27,520 Speaker 1: glad you said that, because more more frightening than law 668 00:43:27,640 --> 00:43:35,080 Speaker 1: enforcement unfairly targeting people is the idea of the increasingly 669 00:43:35,200 --> 00:43:40,680 Speaker 1: gray area and the increasingly black box interactions between large 670 00:43:40,800 --> 00:43:44,600 Speaker 1: data aggregators like Facebook, Google, your local I s P 671 00:43:45,320 --> 00:43:50,239 Speaker 1: and or Internet service provider, and the rest of the 672 00:43:50,280 --> 00:43:53,520 Speaker 1: alphabet soup of the federal government, you know. And it 673 00:43:53,560 --> 00:43:58,319 Speaker 1: could be anything from the FBI to the n s A, 674 00:43:58,520 --> 00:44:02,760 Speaker 1: to local police to Hartman writes in and to Facebook 675 00:44:02,800 --> 00:44:05,520 Speaker 1: and says, hey, we need to We think there are 676 00:44:05,560 --> 00:44:08,640 Speaker 1: these three guys who went to this underwater base off 677 00:44:08,680 --> 00:44:12,759 Speaker 1: the coast of redacted, and they've been posting photos in 678 00:44:12,800 --> 00:44:16,480 Speaker 1: their timeline, and we just want to make sure that 679 00:44:16,560 --> 00:44:18,960 Speaker 1: these three guys are the same three guys who did 680 00:44:19,040 --> 00:44:23,359 Speaker 1: that amazing diamond heist a few years ago for their 681 00:44:23,480 --> 00:44:27,200 Speaker 1: for their freeze ray. Yes, yes, yeah, Hey do you 682 00:44:27,200 --> 00:44:32,040 Speaker 1: guys remember b O b Y Yeah yeah yeah yeah 683 00:44:32,080 --> 00:44:36,279 Speaker 1: bob uh he he came at us a little while 684 00:44:36,320 --> 00:44:39,880 Speaker 1: ago us being the world through social social media and 685 00:44:39,920 --> 00:44:43,160 Speaker 1: talked about the Earth being flat, and we talked about that. Well, 686 00:44:43,239 --> 00:44:48,880 Speaker 1: he thinks that Snapchat is really just collecting blackmail photos 687 00:44:48,920 --> 00:44:53,240 Speaker 1: of our silly faces doing things. I will tell you this. Uh, 688 00:44:53,360 --> 00:44:57,160 Speaker 1: if you think that the images and video that you 689 00:44:57,440 --> 00:45:02,040 Speaker 1: take on Snapchat that disappear after hours has truly disappeared, 690 00:45:02,080 --> 00:45:06,240 Speaker 1: you're absolutely wrong. Yeah. No, that exists on a server somewhere. 691 00:45:06,840 --> 00:45:09,239 Speaker 1: It has to. And then furthermore, and I love that 692 00:45:09,280 --> 00:45:13,479 Speaker 1: we're exploring this part too, because okay, Facebook, people pile 693 00:45:13,560 --> 00:45:17,759 Speaker 1: onto Facebook often because it is, you know, diabolical. It's 694 00:45:17,840 --> 00:45:23,520 Speaker 1: it's an enormous corporation that is completely dependent upon users 695 00:45:23,840 --> 00:45:28,400 Speaker 1: surrendering information about themselves. So due to its very nature 696 00:45:28,480 --> 00:45:30,480 Speaker 1: as to what it is and how it makes money, 697 00:45:30,520 --> 00:45:33,080 Speaker 1: it makes money through the fact that people use it 698 00:45:33,120 --> 00:45:36,480 Speaker 1: and reveal their likes, their dislikes there you know the 699 00:45:36,560 --> 00:45:39,560 Speaker 1: things that they want. Thus you can serve ads against that. 700 00:45:40,200 --> 00:45:43,719 Speaker 1: It's very easy for that business model to take a 701 00:45:43,840 --> 00:45:48,440 Speaker 1: very dark turn without without a lot of discipline and 702 00:45:48,520 --> 00:45:54,560 Speaker 1: policies and and the will to actually uh perform business 703 00:45:54,560 --> 00:45:57,600 Speaker 1: in an ethical manner. And to be completely fair, I 704 00:45:57,680 --> 00:46:00,640 Speaker 1: don't think Facebook is run by a bunch of vampire demons, 705 00:46:01,040 --> 00:46:03,960 Speaker 1: but I do think that it can be very easy 706 00:46:04,080 --> 00:46:08,520 Speaker 1: to make unethical decisions even without being aware of it, 707 00:46:08,600 --> 00:46:11,600 Speaker 1: when you have that massive amount of data. Absolutely, and 708 00:46:11,600 --> 00:46:15,880 Speaker 1: it also extends to as you said, snapchat. And I 709 00:46:16,360 --> 00:46:19,640 Speaker 1: had an issue when um one of one of our 710 00:46:19,760 --> 00:46:23,120 Speaker 1: co host here knowl was going through a Pokemon Go 711 00:46:23,360 --> 00:46:29,399 Speaker 1: phase and it don't put put Pokemon Go camera on 712 00:46:29,480 --> 00:46:32,920 Speaker 1: me or hanging out, so there's always you know, a 713 00:46:32,960 --> 00:46:36,960 Speaker 1: different Pokemon that's on you. And of course being a 714 00:46:37,000 --> 00:46:40,440 Speaker 1: little bit on the paranoid side, I'm kidding, I'm way 715 00:46:40,480 --> 00:46:43,600 Speaker 1: deep into that. Well, Ben, Ben, I mean you're I 716 00:46:43,640 --> 00:46:46,040 Speaker 1: hate to tell you this, but you are a pig post. 717 00:46:46,880 --> 00:46:49,480 Speaker 1: You just got piggies all over you. What do you 718 00:46:49,480 --> 00:46:53,799 Speaker 1: think they mean when they say catch them all talking 719 00:46:53,840 --> 00:46:56,960 Speaker 1: about all the faces? Yeah, one of the first things, 720 00:46:57,000 --> 00:46:59,239 Speaker 1: I turned off on Pokemon Go, although it was not 721 00:46:59,400 --> 00:47:03,160 Speaker 1: for fear. Well, there's some concern for privacy because you know, 722 00:47:03,239 --> 00:47:05,480 Speaker 1: you're using it in public and there are people all 723 00:47:05,520 --> 00:47:07,600 Speaker 1: around you. And while you can take photos of people 724 00:47:07,640 --> 00:47:10,200 Speaker 1: in public, there's nothing illegal against that. I just think 725 00:47:10,239 --> 00:47:13,880 Speaker 1: it's you know, rude. So I turned off the camera 726 00:47:14,080 --> 00:47:17,680 Speaker 1: slash augmented reality part, so it just has the animated 727 00:47:17,719 --> 00:47:22,440 Speaker 1: background as opposed to an actual camera view. Um and uh. 728 00:47:22,480 --> 00:47:25,840 Speaker 1: But again I did that not really out of concern 729 00:47:25,880 --> 00:47:28,640 Speaker 1: of other people, but rather because it doesn't drain. My 730 00:47:28,680 --> 00:47:31,319 Speaker 1: battery is fast, so it's a selfish reason. I wish 731 00:47:31,360 --> 00:47:33,479 Speaker 1: I could say I was being a really decent human being, 732 00:47:33,520 --> 00:47:35,439 Speaker 1: but really I just didn't want my phone to die. 733 00:47:35,600 --> 00:47:38,160 Speaker 1: But is it really turning your camera off? Well, I 734 00:47:38,160 --> 00:47:40,640 Speaker 1: mean it's not showing it on my screen, whether it's 735 00:47:40,680 --> 00:47:43,520 Speaker 1: showing it on someone screen. Oh gosh, I guess. I guess. 736 00:47:43,520 --> 00:47:47,480 Speaker 1: What I am doing is consistently giving Nintendo and the 737 00:47:47,480 --> 00:47:50,840 Speaker 1: Pokemon Corporation up to date information about who is on 738 00:47:50,880 --> 00:47:54,719 Speaker 1: the belt line. That's good, that's what they need, and 739 00:47:54,760 --> 00:47:56,920 Speaker 1: it's a shame that they don't pay you. So there 740 00:47:56,920 --> 00:47:59,040 Speaker 1: there are a couple of things here that we should 741 00:47:59,160 --> 00:48:03,920 Speaker 1: we should explore. So sure. Another another fact that we 742 00:48:04,000 --> 00:48:08,040 Speaker 1: must add is, yes, your face is out there. We 743 00:48:08,040 --> 00:48:10,520 Speaker 1: were increasingly, you know, I think it was Andy Warhol 744 00:48:10,560 --> 00:48:13,680 Speaker 1: who said, now everybody will have fifteen minutes of fame. 745 00:48:14,000 --> 00:48:19,080 Speaker 1: We're increasingly arriving at a world where everyone in some 746 00:48:19,200 --> 00:48:23,360 Speaker 1: way is going to be recognizable in one of these databases. 747 00:48:23,400 --> 00:48:27,600 Speaker 1: But living in an anonymous life is almost impossible if 748 00:48:27,640 --> 00:48:32,319 Speaker 1: you are living in modern society, right, And to me 749 00:48:32,719 --> 00:48:35,520 Speaker 1: a little twisting the knife with this, this is being 750 00:48:35,640 --> 00:48:39,359 Speaker 1: sold at a profit and none of the no one 751 00:48:39,520 --> 00:48:44,440 Speaker 1: is being compensated. So, for instance, every time Matt's UH 752 00:48:44,600 --> 00:48:48,440 Speaker 1: image shows up, it's a false positive. Every time that 753 00:48:48,440 --> 00:48:52,680 Speaker 1: that one resource that is entirely created by him shows 754 00:48:52,760 --> 00:48:56,840 Speaker 1: up in you know, a p D, FBI, etcetera, he 755 00:48:57,280 --> 00:49:01,480 Speaker 1: doesn't get any royalty from that, and I don't I 756 00:49:02,600 --> 00:49:06,399 Speaker 1: for the record, I completely think that will never actually happen. Now, 757 00:49:06,640 --> 00:49:09,640 Speaker 1: if we ever see a time where the FBI opens 758 00:49:09,719 --> 00:49:13,160 Speaker 1: up a fun and zany FBI T shirts store where 759 00:49:13,200 --> 00:49:17,000 Speaker 1: you can get your photo stored that was stored on 760 00:49:17,080 --> 00:49:20,160 Speaker 1: the law enforcement database onto a T shirt and maybe 761 00:49:20,160 --> 00:49:22,360 Speaker 1: as a mug shot or something that's a great idea. 762 00:49:22,760 --> 00:49:26,320 Speaker 1: Oh on a coffee mic mug shot. Coffee mug shot. 763 00:49:26,440 --> 00:49:31,120 Speaker 1: So let's uh, well, let's as as they like to say, 764 00:49:31,200 --> 00:49:34,440 Speaker 1: let's in big in this all right. I think that's 765 00:49:34,440 --> 00:49:40,399 Speaker 1: a cromulent decision. Thank you. I agree. So we're we're 766 00:49:40,440 --> 00:49:44,919 Speaker 1: talking right now about the United States, which it does 767 00:49:45,000 --> 00:49:48,760 Speaker 1: have an extensive network, as you said, an extensive array 768 00:49:48,800 --> 00:49:54,040 Speaker 1: of networks. But my question is when it runs into 769 00:49:54,440 --> 00:49:59,480 Speaker 1: like multinational private industry, does it also run into international 770 00:49:59,560 --> 00:50:06,520 Speaker 1: law enforce I mean, really, if you don't think agencies 771 00:50:06,560 --> 00:50:09,000 Speaker 1: like the Central Intelligence Agency have their own versions of 772 00:50:09,080 --> 00:50:12,240 Speaker 1: the database that incorporate people from all over the world, 773 00:50:12,239 --> 00:50:15,120 Speaker 1: you're absolutely wrong. I mean it has to write like 774 00:50:16,160 --> 00:50:19,520 Speaker 1: and at some point in some aspects you understand, right, 775 00:50:19,520 --> 00:50:25,840 Speaker 1: you're talking about trying to protect national interests from very flexible, 776 00:50:26,040 --> 00:50:32,520 Speaker 1: very mobile, very agile enemies essentially, or or agents if 777 00:50:32,560 --> 00:50:35,040 Speaker 1: you want to call them that, whether they're state agents 778 00:50:35,160 --> 00:50:39,799 Speaker 1: or otherwise. So you understand, you understand the necessity. The 779 00:50:39,800 --> 00:50:45,000 Speaker 1: problem is, again, unless you have a clear set of rules, 780 00:50:45,480 --> 00:50:48,000 Speaker 1: you know, you have some transparency there on how it operates, 781 00:50:48,040 --> 00:50:51,600 Speaker 1: and you have the ability to actually put your technology 782 00:50:51,640 --> 00:50:56,719 Speaker 1: to the test to show that it is in fact accurate. Ultimately, 783 00:50:56,760 --> 00:50:58,719 Speaker 1: you're left with a question of do we did we 784 00:50:58,760 --> 00:51:01,719 Speaker 1: get the right person? This leads to deaths if we 785 00:51:01,760 --> 00:51:04,600 Speaker 1: If we get the wrong person, two things happen. The 786 00:51:04,640 --> 00:51:07,279 Speaker 1: bad guy got away, right, because you didn't get the 787 00:51:07,280 --> 00:51:09,480 Speaker 1: person you wanted to get, and you've given the bad 788 00:51:09,520 --> 00:51:12,640 Speaker 1: guys more reason to hate you. Because you are targeting 789 00:51:12,680 --> 00:51:16,719 Speaker 1: people who are innocent. You are creating more bad guys. Right. 790 00:51:16,840 --> 00:51:18,839 Speaker 1: If you go out there and you are targeting people 791 00:51:18,880 --> 00:51:21,960 Speaker 1: who are are innocent of the crimes that you think 792 00:51:21,960 --> 00:51:25,480 Speaker 1: they committed because you're following a wrong lead due to 793 00:51:25,560 --> 00:51:31,719 Speaker 1: a technological glitch, then you are justifying that anti sentiment 794 00:51:32,600 --> 00:51:35,520 Speaker 1: because you are doing what you have been accused of doing, 795 00:51:35,560 --> 00:51:39,840 Speaker 1: which is coming into another culture and disrupting it. So 796 00:51:40,120 --> 00:51:43,320 Speaker 1: this is true, whether it's domestically and we're talking about 797 00:51:43,360 --> 00:51:47,759 Speaker 1: people who are disproportionately targeted by law enforcement because of 798 00:51:47,800 --> 00:51:50,040 Speaker 1: this sort of same sort of issue, or on a 799 00:51:50,040 --> 00:51:54,320 Speaker 1: global scale. Moreover, there are other nations that have even 800 00:51:54,400 --> 00:51:57,839 Speaker 1: more pervasive camera technologies built around them than the United 801 00:51:57,880 --> 00:52:01,160 Speaker 1: States does. Hello England, I love you, but I can't 802 00:52:01,160 --> 00:52:04,000 Speaker 1: go anywhere without being on a camera in England, Right, 803 00:52:04,040 --> 00:52:08,640 Speaker 1: don't they have the highest density of CCTV. They certainly did, 804 00:52:08,760 --> 00:52:10,440 Speaker 1: at least for a long time. I don't know if 805 00:52:10,440 --> 00:52:12,719 Speaker 1: they currently hold that record, but they for for a 806 00:52:12,760 --> 00:52:16,120 Speaker 1: long time, like especially in London, because they had so 807 00:52:16,160 --> 00:52:19,800 Speaker 1: many different CCTV systems that were most of them were 808 00:52:19,840 --> 00:52:23,239 Speaker 1: self contained, right, so when a shop owner has a 809 00:52:23,680 --> 00:52:27,120 Speaker 1: security camera, yeah, essentially that's what it was. So it 810 00:52:27,160 --> 00:52:29,719 Speaker 1: wasn't like it was just this vast network apart from 811 00:52:29,719 --> 00:52:31,399 Speaker 1: what you know you would see like in in um 812 00:52:31,680 --> 00:52:35,480 Speaker 1: Sherlock where microft is able to zoom in on Watson 813 00:52:35,520 --> 00:52:37,640 Speaker 1: wherever he goes because he can tap into every camera 814 00:52:37,680 --> 00:52:40,160 Speaker 1: that's on every street corner. Well, that's one thing because 815 00:52:40,200 --> 00:52:43,799 Speaker 1: those were those were streets, so those would presumably be 816 00:52:43,840 --> 00:52:47,640 Speaker 1: operated by by the government itself or at least some 817 00:52:47,680 --> 00:52:49,480 Speaker 1: branch of the government. But you wouldn't be able to 818 00:52:49,520 --> 00:52:53,319 Speaker 1: necessarily tap into like every every stores feed because it's 819 00:52:53,320 --> 00:52:55,800 Speaker 1: a self contained system. It's not connected to a network 820 00:52:55,840 --> 00:53:00,960 Speaker 1: for now, which brings us to, of course the future. Yes, okay, 821 00:53:01,200 --> 00:53:09,240 Speaker 1: very last part uh speculation on speculation, educated guesses where 822 00:53:09,320 --> 00:53:13,399 Speaker 1: is this going? So? On the one hand, I am 823 00:53:13,520 --> 00:53:16,319 Speaker 1: somewhat encouraged in the United States that a lot of 824 00:53:16,360 --> 00:53:20,080 Speaker 1: focus has been brought to this matter along with the 825 00:53:20,120 --> 00:53:24,440 Speaker 1: Government Accountability Office and Georgetown University. You have Congress actually 826 00:53:24,480 --> 00:53:28,799 Speaker 1: taking the FBI to task about this. Obviously, there would 827 00:53:28,840 --> 00:53:30,600 Speaker 1: need to be a lot more of those sort of 828 00:53:30,640 --> 00:53:34,120 Speaker 1: discussions at all levels of government for that to actually 829 00:53:34,480 --> 00:53:39,600 Speaker 1: lead to regulations and transparency. I think it's a good start. 830 00:53:39,880 --> 00:53:41,440 Speaker 1: I do not know if it's going to have enough 831 00:53:41,480 --> 00:53:47,120 Speaker 1: momentum to carry it forward to prevent massive abuse of 832 00:53:47,160 --> 00:53:50,319 Speaker 1: this kind of technology on multiple levels, because I mean 833 00:53:50,360 --> 00:53:52,640 Speaker 1: that's happening right now. I don't know if it'll curtail 834 00:53:52,680 --> 00:53:57,160 Speaker 1: it um. I hope it will. That's one possible future 835 00:53:57,200 --> 00:54:00,200 Speaker 1: is that we'll see over time the development of these 836 00:54:00,280 --> 00:54:03,560 Speaker 1: rules and policies that various agencies have to follow so 837 00:54:03,640 --> 00:54:07,120 Speaker 1: that the use of this technology is done in an 838 00:54:07,120 --> 00:54:11,800 Speaker 1: accountable and responsible way, and that uh, you know. The 839 00:54:11,880 --> 00:54:15,680 Speaker 1: FBI they state that when they're running these searches, any 840 00:54:15,719 --> 00:54:18,640 Speaker 1: search result they get back, that does not mean the 841 00:54:18,680 --> 00:54:23,000 Speaker 1: person that they get has become a suspect. It means 842 00:54:23,080 --> 00:54:25,720 Speaker 1: that they have a match that a potential lead to follow. 843 00:54:25,719 --> 00:54:28,560 Speaker 1: They're very careful in their wording to say that they 844 00:54:28,560 --> 00:54:31,880 Speaker 1: haven't identified to suspect that's why they use probe photo 845 00:54:32,040 --> 00:54:35,799 Speaker 1: rather than suspect photo. However, words only go so far. 846 00:54:36,320 --> 00:54:38,279 Speaker 1: How where do the actions go? And that's the big 847 00:54:38,320 --> 00:54:43,239 Speaker 1: fear is that we don't really know yet. So optimist 848 00:54:44,160 --> 00:54:48,480 Speaker 1: Jonathan would say, over time, we developed these rules, we 849 00:54:48,520 --> 00:54:51,799 Speaker 1: hold these agencies accountable for them, We make certain that 850 00:54:51,880 --> 00:54:55,200 Speaker 1: no one is relying too heavily on a technology that 851 00:54:55,440 --> 00:55:00,880 Speaker 1: is not perfect, and it just becomes another another tool 852 00:55:01,160 --> 00:55:04,719 Speaker 1: of agencies to use for specific situations in order to 853 00:55:04,719 --> 00:55:07,040 Speaker 1: investigate leads. And that is as far as it goes. 854 00:55:09,600 --> 00:55:13,840 Speaker 1: Cynical Jonathan, I'm more familiar with this one. Cynical Jonathan 855 00:55:14,320 --> 00:55:20,440 Speaker 1: thinks this is going to continue largely unchecked at various 856 00:55:20,520 --> 00:55:28,960 Speaker 1: levels of law enforcement agencies, including the federal level, largely 857 00:55:29,080 --> 00:55:32,480 Speaker 1: used in in agencies that have no transparency because of 858 00:55:32,480 --> 00:55:34,080 Speaker 1: the nature of what they do, things like the n 859 00:55:34,200 --> 00:55:39,760 Speaker 1: s A, and that there will be a growing number 860 00:55:39,920 --> 00:55:44,239 Speaker 1: of people very much interested in using it for purposes 861 00:55:44,280 --> 00:55:46,840 Speaker 1: that it was not intended, in other words, not to 862 00:55:46,960 --> 00:55:50,600 Speaker 1: investigate crimes alone. I mean they would still want to 863 00:55:50,719 --> 00:55:54,840 Speaker 1: use it for that obviously, but also to perhaps push 864 00:55:54,840 --> 00:56:01,319 Speaker 1: specific political agendas by suppressing opposing political agendas. And this 865 00:56:01,400 --> 00:56:05,239 Speaker 1: sounds very Orwellian and very like I'm not again, I'm 866 00:56:05,280 --> 00:56:08,839 Speaker 1: not a Norwellian type of dystopia person, but it makes 867 00:56:08,880 --> 00:56:12,320 Speaker 1: it so easy to do that even if you weren't 868 00:56:13,080 --> 00:56:16,520 Speaker 1: consciously setting out on that path, you could do it 869 00:56:16,640 --> 00:56:20,320 Speaker 1: just by happenstance of your other methods of going about 870 00:56:20,320 --> 00:56:24,640 Speaker 1: your business. And because we have no rules or regulations 871 00:56:24,680 --> 00:56:28,040 Speaker 1: for it, there's no protection against it. So I find 872 00:56:28,080 --> 00:56:32,920 Speaker 1: it very concerning. I think it needs to be addressed 873 00:56:33,160 --> 00:56:39,320 Speaker 1: and and it needs to be consistently looked at as 874 00:56:39,360 --> 00:56:41,880 Speaker 1: something that we have to fix, because if we don't, 875 00:56:42,600 --> 00:56:45,239 Speaker 1: the abuse of it is going to be rampant. It's 876 00:56:45,280 --> 00:56:48,399 Speaker 1: going to be disruptive. And I mean, if you are 877 00:56:48,480 --> 00:56:53,480 Speaker 1: not a member of whichever privileged group is not consistently 878 00:56:53,560 --> 00:56:56,160 Speaker 1: targeted by this, you might not see a problem with it. 879 00:56:56,200 --> 00:57:00,319 Speaker 1: But for everybody else, it's it's a huge it's un fair, 880 00:57:00,440 --> 00:57:06,799 Speaker 1: it's a threat, it's not cool. Uh So, yeah, I'm 881 00:57:06,840 --> 00:57:09,200 Speaker 1: I'm afraid we're going to a real dark place. And 882 00:57:09,239 --> 00:57:11,879 Speaker 1: it's literally stuff they don't want you to know. When 883 00:57:11,880 --> 00:57:15,120 Speaker 1: we talk about where this data goes, where it's aggregated, 884 00:57:15,440 --> 00:57:19,960 Speaker 1: who's talking to whom? Right? This is? This is not 885 00:57:20,080 --> 00:57:23,560 Speaker 1: only I don't know. I think you make an important 886 00:57:23,600 --> 00:57:30,040 Speaker 1: point when you say it's not illegal, because there's no 887 00:57:30,160 --> 00:57:38,040 Speaker 1: measure of legality yet. But legal and ethical are not 888 00:57:38,080 --> 00:57:40,600 Speaker 1: always the same thing. No, I mean, there are plenty 889 00:57:40,600 --> 00:57:44,960 Speaker 1: of things that I think you could argue are unethical 890 00:57:45,120 --> 00:57:47,600 Speaker 1: that are perfectly legal, and vice versa. There are some 891 00:57:47,640 --> 00:57:51,360 Speaker 1: things that you might argue are illegal but are perfectly ethical, right, 892 00:57:51,400 --> 00:57:55,320 Speaker 1: Like it's legal to uh seeing that Journey song and 893 00:57:55,320 --> 00:57:58,200 Speaker 1: get karaoke, But is it ethical? I would think not that, 894 00:57:58,240 --> 00:58:00,240 Speaker 1: I would think at this point, don't stop, believe even 895 00:58:00,400 --> 00:58:06,720 Speaker 1: has been overdone. Go with a different Journey song. So 896 00:58:06,880 --> 00:58:10,480 Speaker 1: at this point we've talked about the past, the present, 897 00:58:11,760 --> 00:58:16,120 Speaker 1: and the possible future of facial recognition. And Jonathan, I 898 00:58:16,200 --> 00:58:18,480 Speaker 1: want to thank you so much for coming on the 899 00:58:18,520 --> 00:58:24,000 Speaker 1: show and illuminating us choosing that word very carefully, uh 900 00:58:24,160 --> 00:58:28,440 Speaker 1: with with the facts behind something that people hear about. 901 00:58:28,480 --> 00:58:31,640 Speaker 1: And this is a morphous boogeyman sort of way. But 902 00:58:31,800 --> 00:58:34,040 Speaker 1: it turns out that even when you shed this light 903 00:58:34,080 --> 00:58:37,360 Speaker 1: on it for us, it's still a boogeyman. Oh yeah, now, 904 00:58:37,440 --> 00:58:41,880 Speaker 1: this is this is something that is happening. It's happening 905 00:58:41,880 --> 00:58:49,400 Speaker 1: without any real oversight it. It is certainly incorporating uh 906 00:58:49,600 --> 00:58:54,400 Speaker 1: data from people who have no criminal background whatsoever, and 907 00:58:55,080 --> 00:58:59,120 Speaker 1: it is it's very disturbing, Like it's It's one thing 908 00:58:59,200 --> 00:59:01,680 Speaker 1: if everyone were to go with the rules that the 909 00:59:01,760 --> 00:59:05,040 Speaker 1: FBI originally set, which is that we're only using the 910 00:59:05,120 --> 00:59:08,640 Speaker 1: data that we've collected from criminal cases. Even then you 911 00:59:08,640 --> 00:59:12,360 Speaker 1: could argue, well, we don't necessarily know that everybody who's 912 00:59:12,360 --> 00:59:15,880 Speaker 1: in that criminal database actually did something criminal. There might 913 00:59:15,880 --> 00:59:18,160 Speaker 1: be some innocent people in there, but there probably are 914 00:59:18,200 --> 00:59:20,400 Speaker 1: some innocent people in there, and you're also missing out 915 00:59:20,440 --> 00:59:22,920 Speaker 1: on everybody who has yet to commit a crime. But 916 00:59:23,040 --> 00:59:28,640 Speaker 1: that that part would be less concerning from the average citizen, right, Matt. 917 00:59:28,680 --> 00:59:31,440 Speaker 1: I mean, if you're FBI, you hate it because it 918 00:59:31,480 --> 00:59:33,360 Speaker 1: makes it harder for you to find a lead. But 919 00:59:33,400 --> 00:59:37,080 Speaker 1: if you're an average citizen, you're relieved because you know 920 00:59:37,160 --> 00:59:41,120 Speaker 1: you're not gonna get targeted. But the problem is that no, 921 00:59:41,640 --> 00:59:44,880 Speaker 1: those rules, that's not what everyone's following. People are in fact, 922 00:59:45,280 --> 00:59:48,240 Speaker 1: we don't know what everyone is doing. And it's terrifying 923 00:59:48,400 --> 00:59:53,240 Speaker 1: because you are there's no way of knowing how frequently 924 00:59:53,600 --> 00:59:56,520 Speaker 1: your face has already been in virtual lineups. It probably 925 00:59:56,560 --> 00:59:58,760 Speaker 1: has been at some point. It may have been that 926 00:59:58,840 --> 01:00:02,000 Speaker 1: it reached a very early stage before it was dismissed. 927 01:00:02,320 --> 01:00:06,959 Speaker 1: It never went further than that. But right now, as 928 01:00:07,040 --> 01:00:10,880 Speaker 1: we're talking, as your listeners are listening to this, there 929 01:00:10,880 --> 01:00:15,120 Speaker 1: are cold algorithms looking for that numeric match and your 930 01:00:15,160 --> 01:00:18,640 Speaker 1: face might be it. And that reminds me, Uh, do 931 01:00:18,680 --> 01:00:21,120 Speaker 1: you guys want to take a picture for Instagram after 932 01:00:21,160 --> 01:00:24,800 Speaker 1: we after we closed the show doesn't have to be 933 01:00:24,880 --> 01:00:29,440 Speaker 1: of us? All right, we'll break out I patches or something. 934 01:00:29,520 --> 01:00:34,360 Speaker 1: Maybe this is all true. This is not a conspiracy theory. 935 01:00:34,960 --> 01:00:39,480 Speaker 1: This is there is an active collection of groups that 936 01:00:39,560 --> 01:00:44,560 Speaker 1: are cooperating using your data in what could be and 937 01:00:44,600 --> 01:00:49,320 Speaker 1: what have been very dangerous ways. And Jonathan, I think 938 01:00:49,360 --> 01:00:53,280 Speaker 1: you've ended it for us on the perfect note because 939 01:00:53,360 --> 01:00:59,320 Speaker 1: this is only weirdly enough the beginning. So if people 940 01:00:59,320 --> 01:01:02,080 Speaker 1: want to learn more about not just this, but all 941 01:01:02,480 --> 01:01:04,600 Speaker 1: things tech related, could you could you tell us where 942 01:01:04,640 --> 01:01:08,840 Speaker 1: to check you out? Absolutely? I host the podcast tech 943 01:01:08,960 --> 01:01:12,840 Speaker 1: Stuff uh we uh we published twice a week Wednesdays 944 01:01:12,840 --> 01:01:16,840 Speaker 1: and Fridays. We cover all sorts of technology topics, from 945 01:01:17,040 --> 01:01:19,720 Speaker 1: things that are scary, like I did a recent episode 946 01:01:19,760 --> 01:01:23,800 Speaker 1: about near misses with nuclear war. The Times when we 947 01:01:23,800 --> 01:01:26,560 Speaker 1: were the world was on the brink of nuclear devastation 948 01:01:26,960 --> 01:01:30,720 Speaker 1: and cooler heads prevailed because clearly that did not happen. 949 01:01:31,200 --> 01:01:33,360 Speaker 1: But I also do fun, silly stuff. I did one 950 01:01:33,400 --> 01:01:37,040 Speaker 1: that was all about the top memes of the last 951 01:01:37,080 --> 01:01:40,120 Speaker 1: few years, so talking about where those came from and 952 01:01:40,120 --> 01:01:43,160 Speaker 1: and uh and and how they evolved over time. So 953 01:01:43,480 --> 01:01:47,040 Speaker 1: it's the whole gamut of technology stories. So if you 954 01:01:47,120 --> 01:01:50,240 Speaker 1: want to check that out, go to your favorite podcatching 955 01:01:50,280 --> 01:01:54,200 Speaker 1: service such as iTunes podcasts and listen to tech Stuff. 956 01:01:54,200 --> 01:01:56,640 Speaker 1: If you like it, subscribe if you really like it. 957 01:01:56,920 --> 01:01:59,680 Speaker 1: You can watch me record it live because I'm doing 958 01:01:59,720 --> 01:02:04,240 Speaker 1: something crazy on Wednesdays and Fridays over at twitch dot 959 01:02:04,320 --> 01:02:09,400 Speaker 1: tv slash tech Stuff, I live stream the recording sessions 960 01:02:09,920 --> 01:02:12,120 Speaker 1: of my show, so you can go and watch me 961 01:02:12,320 --> 01:02:15,800 Speaker 1: record live. And here's the cool thing or lame thing, 962 01:02:15,920 --> 01:02:18,760 Speaker 1: depending upon your point of view. You can witness all 963 01:02:18,760 --> 01:02:20,920 Speaker 1: the times I mess up or have to take a break, 964 01:02:21,160 --> 01:02:23,000 Speaker 1: or have to cough, or have to drink some water 965 01:02:23,360 --> 01:02:26,440 Speaker 1: or whatever it may be. Because it's live, it's all 966 01:02:26,480 --> 01:02:28,840 Speaker 1: the stuff that we cut out before we publish the show. 967 01:02:29,000 --> 01:02:31,680 Speaker 1: You can witness it and you can say I heard 968 01:02:31,680 --> 01:02:34,240 Speaker 1: that show a month before it published, because that's how 969 01:02:34,280 --> 01:02:38,400 Speaker 1: far out I am. That's fantastic, And I think Matt 970 01:02:38,480 --> 01:02:40,840 Speaker 1: sometimes helps you produce that is that he does. Matt 971 01:02:40,880 --> 01:02:43,200 Speaker 1: has I have pulled Matt in on multiple occasions to 972 01:02:43,200 --> 01:02:46,360 Speaker 1: help set up the both the recording and the Twitch 973 01:02:46,480 --> 01:02:50,600 Speaker 1: streaming side of that. Matt has appeared on camera briefly 974 01:02:51,520 --> 01:02:53,440 Speaker 1: as part of that. I don't make him stay for 975 01:02:53,440 --> 01:02:55,680 Speaker 1: the whole thing because he's cut. It turns out he's 976 01:02:55,680 --> 01:02:59,880 Speaker 1: got stuff to do sometimes. Yeah, so, uh, do check 977 01:03:00,000 --> 01:03:02,960 Speaker 1: out Jonathan and that if you are interested in the 978 01:03:03,000 --> 01:03:05,640 Speaker 1: future of technology and want to feel a little bit 979 01:03:05,680 --> 01:03:09,600 Speaker 1: better about the state of humanity than check out Jonathan's 980 01:03:09,680 --> 01:03:14,120 Speaker 1: other show, Forward Thinking, which is available on YouTube, and 981 01:03:14,200 --> 01:03:17,280 Speaker 1: the Facebook's Yeah Yeah if you wanna. If you wanna 982 01:03:17,400 --> 01:03:21,160 Speaker 1: see my face getting identified all over the place, check 983 01:03:21,200 --> 01:03:23,160 Speaker 1: out Forward Thinking. That one's that was kind of like 984 01:03:23,240 --> 01:03:26,760 Speaker 1: the the Shiny Happy Cousin two Stuff they don't want 985 01:03:26,800 --> 01:03:31,720 Speaker 1: you to Know's dark, dark Emo Cousin. I think we 986 01:03:31,920 --> 01:03:35,280 Speaker 1: prefer to think of it as conspiracy realism. I look, 987 01:03:35,440 --> 01:03:38,160 Speaker 1: I'm not making any kind of judgments here. I'm merely 988 01:03:38,160 --> 01:03:41,240 Speaker 1: making an observation. I love all the children of how 989 01:03:41,280 --> 01:03:45,320 Speaker 1: stuff works. Equally all right, Well, now we're definitely putting 990 01:03:45,320 --> 01:03:47,840 Speaker 1: a picture of you on Instagram, and you may be 991 01:03:47,960 --> 01:03:52,280 Speaker 1: asking yourselves, ladies and gentlemen, wait, the guys have an Instagram. 992 01:03:52,360 --> 01:03:54,960 Speaker 1: Oh my gosh o MG, where can I find it? Well, 993 01:03:55,080 --> 01:03:58,640 Speaker 1: the answer is Conspiracy Stuff Show. You can also find 994 01:03:58,720 --> 01:04:01,560 Speaker 1: us on Facebook and Twitter, where we are Conspiracy Stuff. 995 01:04:02,280 --> 01:04:05,320 Speaker 1: And that's the end of this classic episode. If you 996 01:04:05,360 --> 01:04:09,440 Speaker 1: have any thoughts or questions about this episode, you can 997 01:04:09,480 --> 01:04:12,080 Speaker 1: get into contact with us in a number of different ways. 998 01:04:12,280 --> 01:04:13,840 Speaker 1: One of the best is to give us a call. 999 01:04:13,880 --> 01:04:17,640 Speaker 1: Our number is one eight three three std w y 1000 01:04:17,720 --> 01:04:19,880 Speaker 1: t K. If you don't want to do that, you 1001 01:04:19,880 --> 01:04:22,360 Speaker 1: can send us a good old fashioned email. We are 1002 01:04:22,640 --> 01:04:27,040 Speaker 1: conspiracy at i heart radio dot com. Stuff they Don't 1003 01:04:27,040 --> 01:04:29,600 Speaker 1: Want you to Know is a production of I Heart Radio. 1004 01:04:29,880 --> 01:04:32,160 Speaker 1: For more podcasts from my heart Radio, visit the i 1005 01:04:32,240 --> 01:04:35,120 Speaker 1: heart Radio app, Apple Podcasts, or wherever you listen to 1006 01:04:35,160 --> 01:04:35,960 Speaker 1: your favorite shows.