1 00:00:00,640 --> 00:00:03,480 Speaker 1: So we've been talking about this. This is an interesting one. 2 00:00:03,680 --> 00:00:08,320 Speaker 1: As facial recognition software, it gets better and better. It's 3 00:00:08,600 --> 00:00:11,480 Speaker 1: it's got some cool ways to use it. Obviously, unlock 4 00:00:11,520 --> 00:00:13,640 Speaker 1: your phone or if if you ever end up working 5 00:00:13,720 --> 00:00:16,040 Speaker 1: in a building where it uses your face as your 6 00:00:16,120 --> 00:00:18,000 Speaker 1: key card to get into building or that sort of stuff, 7 00:00:18,040 --> 00:00:20,360 Speaker 1: that could be cool. But on the other hand, China 8 00:00:20,440 --> 00:00:22,599 Speaker 1: is using it to uh, to keep an eye on 9 00:00:22,640 --> 00:00:26,680 Speaker 1: everybody all the time. It is absolutely a cherished new 10 00:00:26,720 --> 00:00:30,440 Speaker 1: tool for totalitarian regimes. So we're a little uncomfortable with 11 00:00:30,440 --> 00:00:33,479 Speaker 1: it and exploring it. Daniel Castro is the vice president 12 00:00:33,520 --> 00:00:37,080 Speaker 1: for the Information Technology and Innovation Foundation, Director of i 13 00:00:37,240 --> 00:00:40,640 Speaker 1: t i F Center for Data Innovation. Joins us. Now, Daniel, 14 00:00:40,680 --> 00:00:43,040 Speaker 1: how are you, sir? I'm good, Thanks for having me on. 15 00:00:43,360 --> 00:00:45,839 Speaker 1: You are pro this sort of technology, is that correct? 16 00:00:46,600 --> 00:00:48,920 Speaker 1: That's right. I see a lot of potential benefits and 17 00:00:48,960 --> 00:00:51,800 Speaker 1: how can be used. So are you a communist? Have 18 00:00:51,880 --> 00:00:53,760 Speaker 1: you ever been a member of the Communist Party? Well, 19 00:00:53,760 --> 00:00:55,640 Speaker 1: that's an interesting thing that you would ask that because 20 00:00:55,640 --> 00:00:58,320 Speaker 1: San Francisco is working on becoming the first city in 21 00:00:58,320 --> 00:01:02,040 Speaker 1: America to ban facial ignition software. There are a lefty 22 00:01:02,040 --> 00:01:04,600 Speaker 1: city and they think it's a scary things. I don't 23 00:01:04,600 --> 00:01:07,959 Speaker 1: know what the politics of this are. Well, San Francisco 24 00:01:07,959 --> 00:01:10,160 Speaker 1: seems to like to ban a lot of new technologies 25 00:01:10,200 --> 00:01:14,560 Speaker 1: to come after sidewalk robots, they've gone after the EA scooters, 26 00:01:14,600 --> 00:01:17,400 Speaker 1: and now they're doing facial recognition and to the history 27 00:01:17,400 --> 00:01:21,080 Speaker 1: of moving quickly in banning technology and the technology of straws, 28 00:01:21,600 --> 00:01:25,959 Speaker 1: right right, So listen, help us understand first of all 29 00:01:25,959 --> 00:01:29,360 Speaker 1: the practical uses of the technology. Let's get excited about 30 00:01:29,360 --> 00:01:31,600 Speaker 1: the technology. Then we'll get into the civil rights thing 31 00:01:31,600 --> 00:01:35,120 Speaker 1: a little bit. Sure. So there's both commercial and government 32 00:01:35,160 --> 00:01:38,399 Speaker 1: uses right now of facial recognition. So people are increasingly 33 00:01:38,440 --> 00:01:40,440 Speaker 1: familiar with how you're using it on your phone to 34 00:01:40,520 --> 00:01:43,800 Speaker 1: unlock your device. Uh, you see it on social networks 35 00:01:43,880 --> 00:01:46,560 Speaker 1: to automatically sort photos based on who your friends are 36 00:01:46,560 --> 00:01:49,000 Speaker 1: in search through those. And then you're starting to see 37 00:01:49,040 --> 00:01:53,720 Speaker 1: applications in the hospitality industry, for example. So the idea 38 00:01:53,760 --> 00:01:56,480 Speaker 1: is that you might be able to register for a 39 00:01:56,560 --> 00:01:59,000 Speaker 1: hotel and then when you walk in, instead of having 40 00:01:59,080 --> 00:02:01,960 Speaker 1: to give your name and you know, your credit card 41 00:02:02,000 --> 00:02:04,720 Speaker 1: and everything taken to have a camera that just recognizes 42 00:02:04,760 --> 00:02:07,120 Speaker 1: you and they can creach you automatically. That's pretty cool. 43 00:02:07,160 --> 00:02:09,760 Speaker 1: You just walk in the front lobby of the hotel. 44 00:02:09,840 --> 00:02:12,960 Speaker 1: Everything recognize you, you walk up, your door opens with 45 00:02:12,960 --> 00:02:16,240 Speaker 1: your credit card or whatever, and you go exactly and 46 00:02:16,240 --> 00:02:19,000 Speaker 1: your things much faster. I mean, we're seeing this in 47 00:02:19,040 --> 00:02:21,040 Speaker 1: airports right now, so you have some of these airlines 48 00:02:21,040 --> 00:02:23,440 Speaker 1: that you just you don't have to have a you know, 49 00:02:23,480 --> 00:02:25,720 Speaker 1: your ticket anymore. You just walk straight past the gate, 50 00:02:25,960 --> 00:02:28,000 Speaker 1: look into a camera, and move on and you're done. 51 00:02:28,320 --> 00:02:31,440 Speaker 1: So lots of benefits there. What people are mostly concerned 52 00:02:31,440 --> 00:02:34,040 Speaker 1: about right now is on the government side, particularly around 53 00:02:34,040 --> 00:02:37,400 Speaker 1: police using it for different reasons. They can use it 54 00:02:37,440 --> 00:02:40,480 Speaker 1: to quickly search for a suspect from you know, some 55 00:02:40,520 --> 00:02:44,000 Speaker 1: surveillance footage. They can use it to create virtual lineups 56 00:02:44,040 --> 00:02:47,160 Speaker 1: of potential suspects. They can use it to help find 57 00:02:47,400 --> 00:02:49,880 Speaker 1: a lost child or an elderly adult for one of 58 00:02:49,919 --> 00:02:52,280 Speaker 1: those Amber Silver alerts. They can also use it to 59 00:02:52,280 --> 00:02:55,520 Speaker 1: help identify where their offices are in the city. Lots 60 00:02:55,520 --> 00:02:58,360 Speaker 1: of different uses there. And what's interesting is that, you know, 61 00:02:58,400 --> 00:03:00,800 Speaker 1: these are things that a lot of times police officers 62 00:03:00,800 --> 00:03:02,760 Speaker 1: are doing right now manually, so right now when they're 63 00:03:02,760 --> 00:03:06,320 Speaker 1: looking for a suspect. Sometimes these departments have just binders 64 00:03:06,400 --> 00:03:08,920 Speaker 1: full of photos of potential suspects and they just have 65 00:03:09,000 --> 00:03:11,639 Speaker 1: to leave through these looking for potential matches. And that's 66 00:03:11,639 --> 00:03:15,720 Speaker 1: a really slow and time intensive process. And what they're 67 00:03:15,720 --> 00:03:18,440 Speaker 1: doing now with the technology is they're able to search 68 00:03:18,480 --> 00:03:21,480 Speaker 1: through millions of records very quickly and identify with more 69 00:03:21,520 --> 00:03:25,080 Speaker 1: accuracy who potential suspects might be. Then they give that 70 00:03:25,120 --> 00:03:28,359 Speaker 1: information to officers whould go out and actually the those 71 00:03:28,400 --> 00:03:30,880 Speaker 1: are some some pretty good upsides to the whole thing. 72 00:03:30,880 --> 00:03:33,360 Speaker 1: And did we make it clear that you're you're for 73 00:03:33,440 --> 00:03:36,080 Speaker 1: this facial recognition stuff here. You don't like the idea 74 00:03:36,080 --> 00:03:39,520 Speaker 1: of San Francisco banning it? Um? Yeah, But now there 75 00:03:39,560 --> 00:03:41,680 Speaker 1: are all kinds of things that would make it easier 76 00:03:41,720 --> 00:03:45,640 Speaker 1: to you know, find criminals or whatever that we've avoided 77 00:03:45,680 --> 00:03:47,720 Speaker 1: throughout the history of our country. We have never had 78 00:03:47,840 --> 00:03:51,200 Speaker 1: a national fingerprint and database for instance, or that sort 79 00:03:51,240 --> 00:03:53,720 Speaker 1: of thing where it make it easier to find people. Well, 80 00:03:53,760 --> 00:03:56,760 Speaker 1: the really scary part about facial recognition though, is that, 81 00:03:57,120 --> 00:03:59,840 Speaker 1: unlike other systems like fingerprints, for instance, it can be 82 00:04:00,040 --> 00:04:03,600 Speaker 1: is for general surveillance in a passive way. It doesn't 83 00:04:03,640 --> 00:04:05,840 Speaker 1: require me to hold out my thumb or give a 84 00:04:05,880 --> 00:04:08,480 Speaker 1: fingerprint or let you check my fingerprints. I mean, it 85 00:04:08,600 --> 00:04:13,640 Speaker 1: is an incredibly valuable tool for totalitarians or those who 86 00:04:13,720 --> 00:04:16,960 Speaker 1: would oppress anybody. I mean, you show up at a rally, 87 00:04:17,160 --> 00:04:20,400 Speaker 1: the authorities are going to know it. They'll just generate 88 00:04:20,440 --> 00:04:22,640 Speaker 1: a list, and so that sort of thing scary. Does 89 00:04:22,760 --> 00:04:24,880 Speaker 1: the industry have a position on that sort of thing, 90 00:04:24,960 --> 00:04:28,560 Speaker 1: Daniel Daniel Castros with us. Yeah. Absolutely. I think you know, 91 00:04:28,760 --> 00:04:31,440 Speaker 1: nobody wants the United States to become China in that 92 00:04:31,520 --> 00:04:36,839 Speaker 1: regards what the type of totalitarian surveillance isn't something that anyone, 93 00:04:37,200 --> 00:04:40,000 Speaker 1: an industry or outside industry is calling for. But the 94 00:04:40,000 --> 00:04:42,159 Speaker 1: whole point is there's a there's a big gray area 95 00:04:42,760 --> 00:04:45,600 Speaker 1: in between, and there's also kind of just replicating things 96 00:04:45,600 --> 00:04:48,919 Speaker 1: that were already doing today with the technology, but more efficiently. 97 00:04:49,160 --> 00:04:51,240 Speaker 1: And that's what the problem with the DAN is. It 98 00:04:51,240 --> 00:04:54,520 Speaker 1: would be completely fine that San Francisco was saying, let's 99 00:04:54,520 --> 00:04:57,760 Speaker 1: make sure this technology is never used to, for example, 100 00:04:57,920 --> 00:05:00,080 Speaker 1: you know, track people throughout the entire city. Let to 101 00:05:00,120 --> 00:05:02,880 Speaker 1: make sure this is never used to track people at 102 00:05:02,920 --> 00:05:05,159 Speaker 1: political rallies. But that's not what they're doing. They're saying, 103 00:05:05,240 --> 00:05:08,160 Speaker 1: let's make sure we never use the technology anywhere across 104 00:05:08,160 --> 00:05:12,080 Speaker 1: the entire government, and that's a problem. Interesting. Daniel Castro 105 00:05:12,240 --> 00:05:16,280 Speaker 1: is the VP for the Information Technology and Innovation Foundation, 106 00:05:16,320 --> 00:05:19,360 Speaker 1: Director of I T i F Center for Data Innovation. Daniel, 107 00:05:19,440 --> 00:05:22,520 Speaker 1: we appreciate the chat. Thanks, happy to be here, all right. 108 00:05:22,800 --> 00:05:25,279 Speaker 1: You know some of that stuff about finding uh, you know, 109 00:05:25,320 --> 00:05:28,719 Speaker 1: a lost kid or an old ster with Alzheimer's wanting around. 110 00:05:28,920 --> 00:05:31,160 Speaker 1: You know, obviously those are those are good arguments for that, 111 00:05:31,240 --> 00:05:34,599 Speaker 1: but you got you gotta trust that once you allow 112 00:05:34,640 --> 00:05:37,200 Speaker 1: these tools to be in place, that then then they 113 00:05:37,240 --> 00:05:39,159 Speaker 1: aren't gonna be misused. And I just don't have that 114 00:05:39,200 --> 00:05:42,840 Speaker 1: trust you. Well, you would have to have a stringent 115 00:05:43,080 --> 00:05:46,960 Speaker 1: and utterly clear set of restrictions on them. But I 116 00:05:46,960 --> 00:05:49,279 Speaker 1: will I will tell you the history of mankind is 117 00:05:49,320 --> 00:05:53,279 Speaker 1: that those who would misuse whatever technologies are opportunities that 118 00:05:53,360 --> 00:05:56,919 Speaker 1: exist to oppress people, find ways around them. It's like 119 00:05:57,000 --> 00:06:00,760 Speaker 1: campaign finance reform. You give your your bundlers and your 120 00:06:00,800 --> 00:06:03,480 Speaker 1: experts ten minutes, they'll read the law, they'll say, oh, 121 00:06:03,480 --> 00:06:05,279 Speaker 1: no problem, we'll just do this and route it through that, 122 00:06:05,320 --> 00:06:09,320 Speaker 1: and they just beat it. And that's what concerns me. Yeah, 123 00:06:09,520 --> 00:06:12,440 Speaker 1: it seems that the biggest thing that Georgia Orriel missed 124 00:06:12,440 --> 00:06:16,040 Speaker 1: about the surveillance state from was that we're just gonna 125 00:06:16,080 --> 00:06:19,480 Speaker 1: opt into it. It's gonna be so convenient, and it's 126 00:06:19,480 --> 00:06:22,479 Speaker 1: gonna just seem like such a a nonstarter. Yeah, of 127 00:06:22,480 --> 00:06:25,200 Speaker 1: course I'm gonna have my phone linked to my location 128 00:06:25,240 --> 00:06:28,320 Speaker 1: because that makes it easier for when I'm mapping directions 129 00:06:28,400 --> 00:06:31,760 Speaker 1: right neat a ride sharing thing. This whole getting on 130 00:06:31,800 --> 00:06:34,360 Speaker 1: a plane, getting into a hotel thing is so convenient, 131 00:06:34,400 --> 00:06:37,279 Speaker 1: I'll overlook the other part of it. It's seductive, literally, 132 00:06:37,320 --> 00:06:40,520 Speaker 1: and that's what Huxley got right. We'd be seduced into oppression, 133 00:06:40,600 --> 00:06:45,760 Speaker 1: not beaten into it. Um. But you know the other hand, 134 00:06:45,760 --> 00:06:48,400 Speaker 1: I do hate standing there with my bag on my 135 00:06:48,400 --> 00:06:51,880 Speaker 1: shoulder waiting for the damn hotel clerk. I mean, who 136 00:06:52,000 --> 00:06:56,960 Speaker 1: enjoys the freaks? Right? We just stroll into the hotel, 137 00:06:57,120 --> 00:07:00,160 Speaker 1: smile at the camera there it's Joe, I'm here, and 138 00:07:00,279 --> 00:07:03,480 Speaker 1: walk up to my palatial suite and then and then 139 00:07:03,520 --> 00:07:07,839 Speaker 1: that will be like phone records are currently that the 140 00:07:07,880 --> 00:07:10,360 Speaker 1: government can access pretty much any time they want to, 141 00:07:11,120 --> 00:07:14,280 Speaker 1: because it's known that any time they ask for that information, 142 00:07:14,360 --> 00:07:17,480 Speaker 1: it's always given to them right there. It will require 143 00:07:17,520 --> 00:07:20,680 Speaker 1: a warrant unless you know it's voluntarily handed over, and 144 00:07:20,720 --> 00:07:24,800 Speaker 1: the government says to uh, large hotel chain, listen to 145 00:07:25,080 --> 00:07:28,320 Speaker 1: some of your environmental practices aren't so great, and I'd 146 00:07:28,320 --> 00:07:29,920 Speaker 1: hate for the e p A to get all up 147 00:07:29,920 --> 00:07:33,320 Speaker 1: in your business. How about does facial recognition records years? 148 00:07:33,320 --> 00:07:35,600 Speaker 1: Do you mind hand them over? Oh? Yeah, yeah, we 149 00:07:35,640 --> 00:07:39,120 Speaker 1: could do that. Sure, Okay, that's how it works. Study 150 00:07:39,200 --> 00:07:43,440 Speaker 1: history fools. On the other hand, this is really cool technology. 151 00:07:43,680 --> 00:07:45,600 Speaker 1: And again, the politics are kind of interesting. You've got 152 00:07:45,640 --> 00:07:48,040 Speaker 1: a lefty city like San Francisco looking at being the 153 00:07:48,040 --> 00:07:51,280 Speaker 1: first city to ban the facial recognition stuff. There will 154 00:07:51,320 --> 00:07:55,880 Speaker 1: probably be some you know, right wing city that's full 155 00:07:55,920 --> 00:07:59,160 Speaker 1: of conservatives using finger quotes because these these labels are 156 00:07:59,160 --> 00:08:01,600 Speaker 1: not very good in these instance. Indeed, who don't want 157 00:08:01,600 --> 00:08:05,560 Speaker 1: it in their city either. Yeah. Yeah, certainly your libertarian 158 00:08:05,600 --> 00:08:07,720 Speaker 1: type conservatives are going to be uncomfortable with it. The 159 00:08:07,720 --> 00:08:11,440 Speaker 1: interesting thing about the left, and again, those those words 160 00:08:11,440 --> 00:08:15,840 Speaker 1: are are decreasingly useful. But the interesting thing about the 161 00:08:15,960 --> 00:08:21,520 Speaker 1: left is that they really cherish freedom from the authorities 162 00:08:21,600 --> 00:08:24,920 Speaker 1: until they become the authority. Then they are so utterly 163 00:08:24,960 --> 00:08:29,400 Speaker 1: convinced of their own righteousness they oppress everybody, um and 164 00:08:29,440 --> 00:08:33,040 Speaker 1: then you know, they become indistinguishable from the right. Really well, 165 00:08:33,320 --> 00:08:36,640 Speaker 1: get down. Here's a question for you. Do you think 166 00:08:36,640 --> 00:08:39,200 Speaker 1: in I don't know how many years, twenty years, fifty years, 167 00:08:39,200 --> 00:08:42,720 Speaker 1: how long it will take, we'll have a similar system, 168 00:08:42,760 --> 00:08:45,400 Speaker 1: if not the federal government, at least like local government 169 00:08:45,480 --> 00:08:47,760 Speaker 1: or state government. Where they got the cameras on the corners, 170 00:08:47,760 --> 00:08:49,960 Speaker 1: will have We've already got those everywhere. There will be 171 00:08:49,960 --> 00:08:51,679 Speaker 1: even more of them, and they'll do the same thing 172 00:08:51,679 --> 00:08:53,360 Speaker 1: to do in China where it picks up your faces 173 00:08:53,360 --> 00:08:56,120 Speaker 1: you cross the street on computers, picking up on the okay, 174 00:08:56,120 --> 00:08:59,040 Speaker 1: here's somebody. They have some outstanding parking tickets, they're behind 175 00:08:59,040 --> 00:09:01,920 Speaker 1: on their taxes. All that information will be constantly they'll 176 00:09:02,000 --> 00:09:04,040 Speaker 1: keep a track of where you go. I absolutely think 177 00:09:04,080 --> 00:09:07,280 Speaker 1: that will happen. Yeah, I'm afraid it will. And and 178 00:09:07,520 --> 00:09:10,240 Speaker 1: sorry to be checking around all sorts of political theory 179 00:09:10,280 --> 00:09:12,320 Speaker 1: at you, but there are quite a few on on 180 00:09:12,400 --> 00:09:15,600 Speaker 1: the libertarian side of things, including iron Rand, whatever you 181 00:09:15,600 --> 00:09:17,600 Speaker 1: think of her. Her idea was that there would be 182 00:09:17,720 --> 00:09:22,120 Speaker 1: so many crimes large and small that everybody would be 183 00:09:22,160 --> 00:09:25,280 Speaker 1: guilty of something. And that's about where we are right now. 184 00:09:25,520 --> 00:09:30,000 Speaker 1: So once you are a criminal, the government is justified 185 00:09:30,000 --> 00:09:32,920 Speaker 1: in doing about anything they want, including you know, the 186 00:09:32,960 --> 00:09:35,760 Speaker 1: stuff you were describing. So yes, I definitely think we'll 187 00:09:35,800 --> 00:09:38,000 Speaker 1: have that system, at least in big cities in the 188 00:09:38,040 --> 00:09:41,640 Speaker 1: future where they're doing that facial recognition thing and everything 189 00:09:41,679 --> 00:09:44,120 Speaker 1: about you is you know. Uh brought up on a 190 00:09:44,160 --> 00:09:46,320 Speaker 1: computer and kept track of At the same time, Old 191 00:09:46,360 --> 00:09:48,680 Speaker 1: iron Land Iron Rand was a little over the top 192 00:09:48,720 --> 00:09:51,880 Speaker 1: and wrote unnecessarily long and dull passages in her books, 193 00:09:52,040 --> 00:09:55,440 Speaker 1: but she got a lot right, And uh, I think 194 00:09:55,520 --> 00:09:58,679 Speaker 1: we're seeing some of it happening. You know. You you 195 00:09:58,840 --> 00:10:02,920 Speaker 1: give a government capability, they will misuse it eventually. It's 196 00:10:02,960 --> 00:10:04,480 Speaker 1: just the way humans work.