Latest / Cyber Crime Podcast / AI, Automation, and the Future of Cybercrime: An FBI Profiler's View
Transcript
- 0:05Speaker 1: Welcome to the Cybercrime Podcast, where we discuss the issues, trends,
- 0:12Speaker 1: and cases of past and present cybercrimes. Let's get started.
- 0:25Speaker 2: We actually got our first sponsor that we've actually accepted
- 0:28Speaker 2: a deal, and I'm very excited to announce the new partnership.
- 0:31Speaker 2: And it's with Titan Nutrition. So I know I had
- 0:33Speaker 2: a lot of you working out. A lot of you
- 0:36Speaker 2: would lift weights, resistance training, aerobic training, whatever it may be.
- 0:40Speaker 2: So I highly recommend that you use some supplements for
- 0:42Speaker 2: sure to be able to help you with your workouts
- 0:45Speaker 2: to get you stronger, to be able to last longer.
- 0:46Speaker 2: Creatine is a given. Protein powders are great because it's
- 0:49Speaker 2: a great way to try to get your protein intake up.
- 0:51Speaker 2: A lot of times we don't have enough protein in
- 0:53Speaker 2: our diet, and that could be a big problem. So
- 0:57Speaker 2: I encourage you to go check out TitanNutrition.net. Titan, T-I-T-A-N. Nutrition.net.
- 1:03Speaker 2: They're evidence-based, which I love a lot. They look at
- 1:06Speaker 2: the research to see what products work and which ones
- 1:09Speaker 2: don't work. And, by the way, you can also get
- 1:11Speaker 2: 10% off if you use the discount code DRCARLOS at
- 1:15Speaker 2: the checkout. And free shipping, too. So if you use creatine,
- 1:19Speaker 2: if you use protein powders, branched-chain aminos, whatever it may be,
- 1:24Speaker 2: all those can be helpful. And we'll talk about those
- 1:27Speaker 2: as well in another podcast. You can catch my other podcast,
- 1:29Speaker 2: the Muscle Gain and Fat Loss Podcast, You can check
- 1:32Speaker 2: that out. But now, to our podcast story about crime, AI,
- 1:38Speaker 2: and technology. Welcome back, everybody. Welcome back, Andy.
- 1:45Speaker 3: Hey, Carlos. It's been a little while since we've had
- 1:48Speaker 3: our last podcast. Nice to hear your voice.
- 1:50Speaker 2: Thank you. Likewise, my friend. It seems like always there's
- 1:53Speaker 2: a little bit of a pause between those things. But today,
- 1:57Speaker 2: you know what's kind of interesting because we're kind of
- 1:59Speaker 2: going on almost like on a series. And I know, folks,
- 2:02Speaker 2: if you followed us, you're probably going, where are they
- 2:04Speaker 2: going now? It looks like we're sticking with artificial intelligence
- 2:07Speaker 2: because Andy did some great research and bumped into a
- 2:10Speaker 2: lot of artificial intelligence and its relationship with crime in
- 2:15Speaker 2: more ways than one. And Andy's going to highlight a
- 2:17Speaker 2: lot of those factors for us in a second. Before
- 2:19Speaker 2: we get started, you know what to do. Share, subscribe, hit.
- 2:22Speaker 3: That like button.
- 2:23Speaker 2: You know we like it. So let's not waste any
- 2:25Speaker 2: more time and let Andy run with it. So Andy,
- 2:27Speaker 2: what did you discover? Because you were rattling off stuff.
- 2:30Speaker 2: And I was really, I mean, I'm always focused on
- 2:33Speaker 2: what you're saying, but I was hyper-focused this time. There
- 2:37Speaker 2: was a lot of interesting stuff that you discovered.
- 2:39Speaker 3: Well, I think what's interesting about artificial intelligence and machine
- 2:44Speaker 3: language is that they represent two very useful tools that
- 2:49Speaker 3: have improved our lives and will continue to reshape our
- 2:53Speaker 3: lives in the future. helping us to predict and overcome
- 2:59Speaker 3: some very challenging issues. At the same time, they're tools.
- 3:03Speaker 3: And we've talked about this as it relates to guns
- 3:06Speaker 3: and gun laws or other tools like computers. And it's
- 3:10Speaker 3: the operator's intent behind the tool that often determines whether
- 3:15Speaker 3: it's being used for good or evil. And we're going
- 3:19Speaker 3: to talk about that concept today because And the research
- 3:23Speaker 3: that we've done, we found that AI and machine language is,
- 3:26Speaker 3: in fact, used for both, for helping police detect and
- 3:32Speaker 3: predict where crime may occur, as well as by scammers.
- 3:37Speaker 3: And we're going to talk about both of those topics today.
- 3:40Speaker 3: So why don't we talk about, first, the good side,
- 3:42Speaker 3: the good before the evil, right? So let me start
- 3:47Speaker 3: by talking about a couple of products that I've discovered
- 3:50Speaker 3: that use AI machine language to better and more efficiently
- 3:55Speaker 3: predict and detect what crime is. One is a CCTV
- 4:01Speaker 3: operated process, but it's not just a camera and it's
- 4:05Speaker 3: not just a sensor. It's the software, the artificial intelligence
- 4:09Speaker 3: software and geospatial software that they combine with this technology.
- 4:15Speaker 3: It's called a ShotSpot and it was developed during the
- 4:21Speaker 3: Iraqi war. It's being used according to the company in
- 4:24Speaker 3: 90 different cities. And the ShotSpotter is a machine that
- 4:32Speaker 3: has a camera sensor on it. And it has an
- 4:37Speaker 3: auditory capability that is, um, programmed to listen to different gunshots,
- 4:43Speaker 3: the reports of a gun. And then it triangulates. When
- 4:47Speaker 3: it hears the gunshot go off, the camera spins 360
- 4:50Speaker 3: to the location of where the gunshot went off to.
- 4:54Speaker 3: And then based on geospatial software, can triangulate to the
- 4:59Speaker 3: very location, about 10 feet. It has an error span
- 5:02Speaker 3: of about 10 feet. And then automatically contacts the police
- 5:06Speaker 3: department And then a responding officer or officers can go
- 5:11Speaker 3: to identify where that shot was fired. The company claims
- 5:16Speaker 3: that only 20% of all gunshots are reported by 9-11
- 5:19Speaker 3: through a telephone call to 9-11. Only 20%, that means
- 5:23Speaker 3: 80% of gunshots go unreported. So you can imagine in
- 5:27Speaker 3: cities like Chicago, where this is being utilized, the importance
- 5:32Speaker 3: of being able to identify very quickly where the gunshot
- 5:36Speaker 3: occurred so that the police can report. Now, this is
- 5:40Speaker 3: a detection tool. This is not a prevention or prediction tool.
- 5:44Speaker 3: Gunshot's already gone off. And so as we've talked in
- 5:47Speaker 3: the past, this represents a tool for law enforcement, right?
- 5:51Speaker 3: Because the crime has already been committed. Gunshot goes off
- 5:54Speaker 3: in Chicago. It might be a gang shooting. Police are
- 5:57Speaker 3: now off to, but more accurately than they have been
- 6:00Speaker 3: in the past. And so the response time has been quicker.
- 6:05Speaker 3: That's one example.
- 6:06Speaker 2: You know, it's interesting. If I can hold on, jump
- 6:08Speaker 2: on that for a second. It's fascinating because you mentioned
- 6:11Speaker 2: Chicago and any cities that have a lot of gang issues.
- 6:14Speaker 2: One of the issues they have in criminology, they have
- 6:18Speaker 2: that concept of collective efficacy. And you can't have that
- 6:20Speaker 2: in some of these cities because of the retribution by
- 6:22Speaker 2: that gang member. They figure out, hey, Andy snitched on me.
- 6:26Speaker 2: I'm coming back for that guy. And if they now
- 6:28Speaker 2: now you don't. One less thing to have to worry
- 6:30Speaker 2: about because the shot spotter can at least help them. Hey, look,
- 6:33Speaker 2: shots went off. This isn't normal. I'm assuming some people
- 6:36Speaker 2: are even desensitized where they're not even going to, you
- 6:38Speaker 2: know what, it's another gunshot. Just hit the deck and
- 6:40Speaker 2: once it's over, we'll get back up. I'm assuming that's
- 6:43Speaker 2: part of it too. But a lot of it could be,
- 6:45Speaker 2: I don't want to say anything if I look out
- 6:46Speaker 2: my window and I see who's shooting who. I want
- 6:48Speaker 2: to keep myself quiet. Psychologically, it's interesting. I wonder how
- 6:52Speaker 2: it's changed the communities a little bit in the sense
- 6:55Speaker 2: of providing at least some kind of response. And security, yeah.
- 7:00Speaker 3: Yeah, I mean, I could see where if you're in
- 7:03Speaker 3: a high crime area of any city, having a tool
- 7:10Speaker 3: like the ShotSpotter would give me maybe, in some cases,
- 7:15Speaker 3: a false sense of security. At the same time, if
- 7:20Speaker 3: I'm a criminal and I know where the location of
- 7:23Speaker 3: these tools are, right, and I don't know how they
- 7:25Speaker 3: hide the ShotSpotter, but it could become a target of
- 7:30Speaker 3: the criminal to shoot them so that they can't respond.
- 7:35Speaker 3: And any CCTV camera, and we're seeing more and more
- 7:38Speaker 3: CCTV camera coverage in cities, but the unique aspect of
- 7:43Speaker 3: AI and ML, machine language, is that they're now combining
- 7:49Speaker 3: the technologies. So it's not just a camera or sensor,
- 7:53Speaker 3: it's a camera or sensor with an algorithm attached to
- 7:56Speaker 3: it that can identify, in some cases predict, human behaviors.
- 8:03Speaker 2: That's the minority report thing right.
- 8:05Speaker 3: That's yeah it is kind of close to minority report
- 8:09Speaker 3: and now let's switch to prediction you're right there's a
- 8:13Speaker 3: tool called predpol and this company is out of china
- 8:17Speaker 3: they developed a an ai camera but with advanced visual
- 8:22Speaker 3: analytics and And so this camera can identify individuals that
- 8:28Speaker 3: can triangulate on your face and do facial recognition, but
- 8:31Speaker 3: they also look for anomalies, including gait patterns. And so
- 8:38Speaker 3: the gait analysis, and that's your stride, of your normal
- 8:41Speaker 3: stride when you walk, but then increases if you're agitated
- 8:44Speaker 3: or if you're perhaps engaged in criminal behavior and trying
- 8:49Speaker 3: to run away. So this particular AI camera looks at
- 8:54Speaker 3: facial recognition and gait analysis. It's called CloudWalk in this
- 9:00Speaker 3: particular program. And then it looks for anomalies in the
- 9:04Speaker 3: behavior of the individual and then reports those anomalies to police. Now,
- 9:09Speaker 3: according to the company, they used it in Seapoint, South Africa,
- 9:17Speaker 3: and they had a 65% drop in crime. Wow, that's
- 9:23Speaker 3: a lot. Yeah, pretty big. Now, a similar tool was
- 9:27Speaker 3: used in Tacoma, Washington, where they saw a 22% drop
- 9:32Speaker 3: in crime. That tool was called PredPol. And again, it
- 9:36Speaker 3: looks for anomalies of behavior. It looks for unattended baggage.
- 9:42Speaker 3: These types of tools could be useful at, you say,
- 9:44Speaker 3: the Boston Marathon. Remember the bombing there?
- 9:47Speaker 2: Oh, yeah.
- 9:47Speaker 3: Yeah. Then you had two subjects that were pacing back
- 9:50Speaker 3: and forth, casing the joint. They had backpacks on. Eventually
- 9:54Speaker 3: the backpacks were left unattended. These tools are more predictive
- 9:59Speaker 3: in the sense that nothing has happened yet, but if
- 10:02Speaker 3: they see the bag that's been left unattended for whatever
- 10:05Speaker 3: the time set is, maybe 15 minutes, 10 minutes or less,
- 10:11Speaker 3: if they see an individual pacing back and forth, if
- 10:14Speaker 3: they identify a face and they triangulate that face and
- 10:17Speaker 3: then run it against DMV and and search for outstanding warrants.
- 10:21Speaker 3: They might have a person who is a known terrorist
- 10:25Speaker 3: who had a backpack, dropped the backpack, all of this
- 10:27Speaker 3: being done without any human interaction, all through the AI software.
- 10:34Speaker 3: And then a phone call goes 911 and the police respond.
- 10:37Speaker 3: And then perhaps, you talked about minority report, perhaps in
- 10:42Speaker 3: the near future, we have these tools helping us prevent
- 10:46Speaker 3: crimes from occurring. like the Boston Marathon bombing.
- 10:51Speaker 2: Yeah, it's really fascinating. Go ahead.
- 10:55Speaker 3: Incredible. It's just incredible utilization of advanced technologies by police
- 11:02Speaker 3: to help prevent a crime from occurring. Having said that,
- 11:09Speaker 3: the bad guys have also looked at using these tools.
- 11:15Speaker 3: And particularly tools where there are situations where there's human engagement.
- 11:21Speaker 3: So when I was doing this research, there's a lot
- 11:24Speaker 3: of different frauds out there, financial frauds, one of which
- 11:28Speaker 3: are known as romance scams, where AI is being utilized.
- 11:35Speaker 3: And what I found in the research is that according
- 11:37Speaker 3: to the FTC, the Federal Trade Commission, $ 1. 3 billion in frauds.
- 11:41Speaker 3: 2022 was lost through romance scams.
- 11:47Speaker 2: It's a lot.
- 11:48Speaker 3: It's a lot of money. I'll take $ 1. 3 billion. The
- 11:53Speaker 3: average cost to a victim is $ 4, 400. Now, when I
- 11:59Speaker 3: was at the FBI's behavioral science unit, we actually did
- 12:03Speaker 3: some work and study looking at how terrorist organizations can
- 12:09Speaker 3: utilize a computer to, uh, to develop, garner trust with
- 12:16Speaker 3: a person in the United States, have them convert to Islam,
- 12:21Speaker 3: go overseas and blow themselves up yelling Ali Akbar. We
- 12:25Speaker 3: were fascinated in the process of developing rapport and trust
- 12:31Speaker 3: through what is known as computer mediated communications. So you
- 12:36Speaker 3: and I right now, as we engage, we're doing it
- 12:39Speaker 3: through this wonderful internet, and Zoom, of course, but this
- 12:44Speaker 3: is a form of computer-mediated communications. This is different than
- 12:48Speaker 3: being engaged live with someone in the presence of that
- 12:52Speaker 3: other person and developing rapport and trust through that process.
- 12:56Speaker 3: It was thought years ago that it was very difficult
- 12:59Speaker 3: to develop trust or even develop rapport or use rapport
- 13:04Speaker 3: to build trust using computer-mediated communications, and there was a
- 13:08Speaker 3: lot of research done on this area. It turns out
- 13:12Speaker 3: that the same processes that lead an individual to develop
- 13:17Speaker 3: trust in person are found in computer media communications and liking.
- 13:24Speaker 3: And there's a, um, great book called influence by help
- 13:30Speaker 3: me out here.
- 13:31Speaker 2: Um, Yeah. You mean, uh, that's it.
- 13:36Speaker 3: That's it. And, um, The book talks about seven ways
- 13:41Speaker 3: of influencing behavior, one of which is liking. And what
- 13:46Speaker 3: artificial intelligence is allowing criminals to do is use technology,
- 13:51Speaker 3: one of which is called deepfake. Are you familiar with deepfake? Oh, yeah.
- 13:58Speaker 3: So they use tools such as MRF. It's called MRF.AI
- 14:03Speaker 3: and another tool called FOTOR, F-O-T-O-R. to con victims on
- 14:09Speaker 3: online dating scams, right? What they'll do is they'll use
- 14:13Speaker 3: deep fake, which is a video that is not real.
- 14:18Speaker 3: So if, you know, we're on Zoom, so you can
- 14:21Speaker 3: see my face, our audience can't, but unfortunately you can.
- 14:25Speaker 3: And so, but you're not, if I was using deep fake,
- 14:29Speaker 3: I would be talking with this visual, but this wouldn't
- 14:33Speaker 3: be me. So my lips would be moving. You would
- 14:37Speaker 3: hear my voice, but or maybe another voice. And but
- 14:42Speaker 3: the visual that you see would be fake. And so
- 14:46Speaker 3: what happens is there's these Nigerian scammers are using fake video.
- 14:51Speaker 3: If you go to YouTube, our listeners can go to
- 14:53Speaker 3: YouTube and search romance scams or catfishing scams. And there's
- 15:00Speaker 3: some videos where these Nigerian scammers have been outed by
- 15:05Speaker 3: intrepid amateur sleuths. And the thing is, they're using artificial
- 15:11Speaker 3: intelligence to generate these artificial, these fake, deep fakes to
- 15:21Speaker 3: allow them to develop trust and rapport with potential victims.
- 15:28Speaker 3: So that's just one example of this process.
- 15:33Speaker 2: That was really crazy because they don't even So I
- 15:35Speaker 2: know that I saw Russia doing it with Ukraine. They
- 15:38Speaker 2: actually had one of their they deep faked a former
- 15:41Speaker 2: Ukrainian ambassador before they started the war. And they joined
- 15:45Speaker 2: in on a private Zoom meeting between Ukraine and some
- 15:49Speaker 2: other countries. And they thought it was him till he
- 15:51Speaker 2: called and said, I'm running behind. I'll be there in
- 15:53Speaker 2: about 20 minutes. Yeah, they didn't count on that part.
- 15:56Speaker 2: But yeah, yeah, it was really that scared me.
- 16:00Speaker 3: Well, that's a sophisticated form for intelligence gathering, and certainly
- 16:05Speaker 3: it can be used for those applications as well. The
- 16:08Speaker 3: point being that these criminals are now using deepfake. They're
- 16:14Speaker 3: using chatbots, and these chatbots are trained to act like
- 16:19Speaker 3: humans in the conversations online. Now, there are some red
- 16:24Speaker 3: flags because chatbots are not yet sophisticated enough to change
- 16:28Speaker 3: their verbiage. So- If asked similar questions, they'll repeat the
- 16:33Speaker 3: answers verbatim. And a lot of these catfishing scams do
- 16:39Speaker 3: the same thing. You'll know you're on a bot when
- 16:41Speaker 3: they follow a pattern. And usually the pattern is, hi,
- 16:45Speaker 3: how are you? Whatever, handsome, pretty, whatever. And then where
- 16:49Speaker 3: are you from is another one that you usually start with.
- 16:52Speaker 3: Where are you from? What do you do? And these
- 16:55Speaker 3: are very short questions. sentences because that's where the machine
- 16:59Speaker 3: language technology is right now. But it's a force multiplier
- 17:05Speaker 3: because they can steal somebody's profile and pictures. And by
- 17:10Speaker 3: the way, they will ask for your picture. What they're
- 17:13Speaker 3: hoping for are pictures that cannot be reverse looked up.
- 17:17Speaker 3: You're familiar with reverse lookup, right?
- 17:19Speaker 2: Yep.
- 17:20Speaker 3: There's plenty of sites online where you can take a picture.
- 17:22Speaker 3: You put the picture in the browser. it reverse looks
- 17:28Speaker 3: throughout the internet to see where the original source is.
- 17:32Speaker 3: So it'll come to Twitter, it'll come to Instagram, it'll
- 17:34Speaker 3: come to Facebook, whatever the picture was at. They want
- 17:37Speaker 3: virgin pictures. They want clean pictures. So that when a
- 17:41Speaker 3: potential victim uses reverse lookup, it doesn't come back. So
- 17:46Speaker 3: that gives them more legitimacy. The whole idea in this
- 17:49Speaker 3: process is using these technologies to develop rapport and therefore
- 17:55Speaker 3: trust Once you have the trust, then you can try
- 17:59Speaker 3: to seek the monetary gain you're looking for. It may be, hey, sweetie,
- 18:05Speaker 3: I need $ 300 so that I can fix my iPhone.
- 18:08Speaker 3: Or my grandmother's living with me and she needs $ 30
- 18:12Speaker 3: for pills. They don't care about the $ 30 or the $ 300.
- 18:16Speaker 3: What they're looking for is your financial information. And that's
- 18:21Speaker 3: the second piece. Once you're willing to give them the money,
- 18:23Speaker 3: then they're going to ask for your bank routing information.
- 18:27Speaker 3: They're going to try to get your identifying information. Then
- 18:29Speaker 3: they're going to steal everything they can from you. There's
- 18:33Speaker 3: all kinds of different variations of these phishing schemes. But
- 18:37Speaker 3: one critical common element in cat phishing schemes is that
- 18:44Speaker 3: they're going to use a romance platform, a dating platform.
- 18:49Speaker 3: And eventually, within a week, certainly, they're going to ask
- 18:53Speaker 3: you for some money in some form. It could be
- 18:57Speaker 3: an Apple gift card for their son. It could be
- 19:00Speaker 3: the pills for grandma. It can be to fix the
- 19:02Speaker 3: cell phone. It could be, can you put some money
- 19:05Speaker 3: on my food plan here at college? But whatever it's
- 19:08Speaker 3: going to be, the idea is for you to share
- 19:11Speaker 3: your financial information with them. And that's why the Federal
- 19:15Speaker 3: Trade Commission says over 36,000 people reported being conned by
- 19:21Speaker 3: people pretending to be their friends or family using artificial
- 19:24Speaker 3: intelligence and AI chat bots.
- 19:27Speaker 2: In 2020.
- 19:27Speaker 3: That's a lot, man. That's a lot. And what happens
- 19:34Speaker 3: to me when I look at these things, when I
- 19:35Speaker 3: was doing this research, I'm like, how can people be
- 19:38Speaker 3: that dumb? But was it Barnum and Bailey said a,
- 19:44Speaker 3: what do you say, a fool is born every day?
- 19:52Speaker 3: And a fool and his money are quickly parted. So
- 19:55Speaker 3: there are a lot of suckers. There's a sucker born
- 19:58Speaker 3: every day and there are a lot of suckers out
- 19:59Speaker 3: there that become much less wealthy because they are desperate
- 20:05Speaker 3: for affection or, intimacy or relationship. And that's what these
- 20:10Speaker 3: people prey on.
- 20:12Speaker 2: And that's the sad thing. And it hasn't gone away.
- 20:14Speaker 2: I think one of the most common actors that people
- 20:17Speaker 2: tend to use who they say they're representing that actor
- 20:21Speaker 2: or that they are that actor is Keanu Reeves. It's
- 20:23Speaker 2: kind of surprising. He has the most, he has the
- 20:26Speaker 2: most imposters of anybody else. And it's kind of funny
- 20:29Speaker 2: because I actually know someone who fell for that. She went,
- 20:35Speaker 2: she went out with the guy, the guy, He kind
- 20:37Speaker 2: of had a different MO than what I've heard before.
- 20:39Speaker 2: Maybe you've heard before as well. He kind of actually,
- 20:41Speaker 2: he bought her dinner. He took her out. And then
- 20:46Speaker 2: about two or three dates later is when he started
- 20:50Speaker 2: saying things like, have you ever seen such, such movies?
- 20:53Speaker 2: And she said, no. And she goes, well, this is me.
- 20:56Speaker 2: I wanted to tell you now. And she's like, oh,
- 21:00Speaker 2: he knew she wasn't up to speed in the movie world.
- 21:03Speaker 2: She just wasn't that kind of person. But he looked
- 21:07Speaker 2: like him. So he showed her a picture of himself
- 21:10Speaker 2: in the movie, and she said he really looked like him,
- 21:13Speaker 2: and he did look like him. And eventually, we kept
- 21:20Speaker 2: warning her that, you know what, it's probably not going
- 21:23Speaker 2: to be him. The odds are really high that it's
- 21:25Speaker 2: not going to be him.
- 21:25Speaker 3: Not going to be him, right, yeah.
- 21:27Speaker 2: And eventually, we found out she wasn't because he supposedly
- 21:31Speaker 2: had a heart condition, and he had to fly to Washington, D.C.
- 21:34Speaker 2: to do the operation, and then but he has to
- 21:37Speaker 2: keep it quiet because he doesn't want the paparazzi all
- 21:39Speaker 2: over it. And then now he had a flight to
- 21:40Speaker 2: India because they're the only ones that could do this
- 21:42Speaker 2: special operation. Oddly enough, he never went for the money.
- 21:46Speaker 2: So we don't know if he just, maybe he thought
- 21:47Speaker 2: there wasn't enough money there to bother with. So he
- 21:49Speaker 2: kind of ended it and moved along.
- 21:51Speaker 3: Right. Well, I, you know, again, it's, I, I come
- 21:56Speaker 3: back to the PT Barnum quote about a sucker born
- 22:00Speaker 3: every day or a fool is born every day. I,
- 22:04Speaker 3: Some of these scams that I've read about, it's so
- 22:08Speaker 3: obvious to me, right, that these are scams. But when
- 22:11Speaker 3: you're that desperate and that close to it, it's certainly
- 22:14Speaker 3: less obvious. There are some red flags. And one of
- 22:20Speaker 3: those red flags is if some, you know, first and foremost,
- 22:23Speaker 3: if somebody is asking for money that you just met
- 22:25Speaker 3: on the Internet less than a week ago, probably a scam.
- 22:30Speaker 3: If somebody is unwilling to video and this has gotten
- 22:34Speaker 3: a little more sophisticated, a little harder because of the
- 22:37Speaker 3: deep fakes. But in most cases, they won't video chat
- 22:40Speaker 3: with you because it's a scam. And, you know, they're
- 22:43Speaker 3: they're a Nigerian male pretending to be, you know, a
- 22:47Speaker 3: 35 year old female. So then that's that's a good
- 22:51Speaker 3: way to determine it. You don't ever give your social
- 22:56Speaker 3: security number or your routing number or your account number
- 23:00Speaker 3: to anyone that you don't know. I mean, those are
- 23:02Speaker 3: clearly commonsensical things, but $ 1. 3 billion tells me that there
- 23:08Speaker 3: are a lot of fools out there that don't really
- 23:13Speaker 3: care about their personal security or information security. This company
- 23:18Speaker 3: that was talking about how they use the software to
- 23:21Speaker 3: prevent crime, they talk about the development of deep neural networks.
- 23:30Speaker 3: And that's really the basis of these sophisticated AI software
- 23:35Speaker 3: programs is that they create new knowledge. They're not human, certainly,
- 23:43Speaker 3: but these deep neural networks learn from the information of
- 23:47Speaker 3: their past. And so, these computers are becoming more and
- 23:51Speaker 3: more sophisticated to understand the patterns of behavior. And you've
- 23:56Speaker 3: heard of routine activities theory, of course, where you have
- 23:59Speaker 3: a motivated offender and you have a potential victim without
- 24:05Speaker 3: a capable guardian. And so if you can see where
- 24:10Speaker 3: patterns of crime have occurred in a particular town over
- 24:16Speaker 3: the course of the last five years, crime can become
- 24:19Speaker 3: predictable in the future. And using these security cameras and
- 24:23Speaker 3: deep neural networks, combining the optics with the algorithm, not
- 24:31Speaker 3: only takes away the bias that you might have as
- 24:33Speaker 3: a police officer, but gives a predictive process to the
- 24:39Speaker 3: department so that they can create and define a hotspot,
- 24:43Speaker 3: put police in that area, and basically prevent crime by
- 24:48Speaker 3: by creating a presence of threat to the perpetrator. And
- 24:53Speaker 3: that's the future, I think, in a lot of policing
- 24:55Speaker 3: in prevention, right? Now, here's the thing that I find interesting.
- 25:00Speaker 3: How do you prevent or how do you measure the negative?
- 25:06Speaker 3: How do you prove that negative? How do you say
- 25:08Speaker 3: that the deep neural network and the cameras prevented crime
- 25:13Speaker 3: because there's been a drop in the crime rate? Because
- 25:17Speaker 3: there's a lot of factors that play into crime rates
- 25:21Speaker 3: or crime in general, right? So anecdotally, you can say
- 25:26Speaker 3: that the presence of the police officer, now you have
- 25:30Speaker 3: a capable guardian over the potential victim. And because most
- 25:34Speaker 3: criminals are reasonable or reason, now the routine activity theory,
- 25:40Speaker 3: the perpetrator decides not to act in a criminal manner,
- 25:44Speaker 3: or at least not in that location, right? And that's
- 25:47Speaker 3: the theory. So that's and then, you know, in closing
- 25:51Speaker 3: on this topic, because I find this whole topic when
- 25:53Speaker 3: we can talk about this, you know, for three or
- 25:54Speaker 3: four podcasts. But by putting these cameras in place and
- 26:00Speaker 3: you were talking about L.A. and some of the cessation
- 26:03Speaker 3: of these activities. By putting these cameras in place, particularly
- 26:08Speaker 3: in areas that are high crime areas, which coincidentally, many
- 26:13Speaker 3: of these high crime areas are in areas of people
- 26:16Speaker 3: of color. people of color, you have serious civil liberty
- 26:23Speaker 3: issues that are advanced because they're like, wait a second,
- 26:27Speaker 3: you're putting these surveillance cameras and the sophisticated software in
- 26:33Speaker 3: my neighborhood, but not in, you know, Beverly Hills. And so,
- 26:40Speaker 3: you know, you're, you're, you know, playing big brother because
- 26:43Speaker 3: I can't afford to live in a fancy house. And
- 26:46Speaker 3: I don't know what your thoughts are on that, but
- 26:49Speaker 3: it is an issue. And so many communities have backed
- 26:53Speaker 3: off on the use of these technologies that could be
- 26:56Speaker 3: used to prevent crime because of the civil liberty issues.
- 27:00Speaker 2: I think it's a mixed bag. I can see both
- 27:02Speaker 2: sides of the story. There's the side of, hey, you're
- 27:04Speaker 2: picking on us because we don't have the ability to
- 27:06Speaker 2: move into another area or because we're a particular ethnicity
- 27:09Speaker 2: or dominant in this ethnicity. But the other side of
- 27:11Speaker 2: the coin is, no, we're trying to help you because
- 27:13Speaker 2: you have a high crime problem or we're trying to
- 27:15Speaker 2: reduce the crime. So I don't know if there's even
- 27:18Speaker 2: a halfway point there. It depends on how those arguments
- 27:22Speaker 2: are reconciled. I get both sides of the argument. How
- 27:26Speaker 2: do you quantify that? I don't even know if you can.
- 27:29Speaker 2: How do you be able to define it clearly enough
- 27:31Speaker 2: to show that all sides have good intentions? I guess
- 27:37Speaker 2: it'd be another one. It's difficult.
- 27:39Speaker 3: It's funny how a lot of our conversations, you and I,
- 27:42Speaker 3: in these podcasts come back to the original point and
- 27:45Speaker 3: You know, we talked about in our opening comments that
- 27:50Speaker 3: we're going to talk about a tool. You know, the
- 27:53Speaker 3: combination of these computer, you know, these cameras and these
- 27:58Speaker 3: sensors coupled with the algorithms to help predict or detect
- 28:02Speaker 3: where crime is occurring. But ultimately, these tools have humans
- 28:06Speaker 3: behind them, and it's the intent that the human introduces
- 28:10Speaker 3: that determines whether it's being used for good or evil.
- 28:14Speaker 3: And even when it's being used for good by police
- 28:16Speaker 3: departments to help prevent crime, it can be interpreted as
- 28:19Speaker 3: evil by those that live in that community and feel
- 28:23Speaker 3: oppressed by these technologies that are intended to help them.
- 28:28Speaker 3: And that's a much deeper philosophical conversation that may warrant
- 28:34Speaker 3: a future podcast, but I don't have any answers for.
- 28:39Speaker 3: I've often said that crime is a relational issue. of
- 28:44Speaker 3: unpredictable behavior. You've heard me talk about all relationships are
- 28:47Speaker 3: predicated on trust and control. When trust goes down, controls
- 28:50Speaker 3: go up. So if your relationship in your community is
- 28:54Speaker 3: between the people and the police that serve them, but
- 28:58Speaker 3: there's a high level of unpredictable behavior in that community,
- 29:02Speaker 3: namely there's a high level of crime, public trust will
- 29:05Speaker 3: go down. And the only way to stabilize that environment
- 29:09Speaker 3: is to increase social control, more police presence. has very
- 29:14Speaker 3: little to do with the race of the people or
- 29:17Speaker 3: the ethnicity of the people or the religion of the people.
- 29:20Speaker 3: And you see this demonstrated throughout the world. And so
- 29:25Speaker 3: where there's high crime, you're going to have higher presence
- 29:28Speaker 3: of police because you need social controls, because there's a
- 29:33Speaker 3: lower level of public trust, because there is a lack
- 29:37Speaker 3: of predictability of behavior as it relates to the base
- 29:41Speaker 3: or laws. Now, the question becomes, what's the intent behind
- 29:46Speaker 3: the police? Is it to maintain, quote unquote, security or
- 29:50Speaker 3: law and order? And if that's the case, then you
- 29:54Speaker 3: have a buy-in in the community. And where that's the case,
- 29:57Speaker 3: you don't really have a control. hoot and holler from
- 30:01Speaker 3: the people because they're cooperating and partnering with the police themselves.
- 30:07Speaker 3: So ultimately, in my opinion, whether it's these sophisticated tools
- 30:12Speaker 3: or just the cop on the beat, it comes down
- 30:15Speaker 3: to managing public trust against social control. And that's the
- 30:19Speaker 3: heart and base of relational policing. So it's not the
- 30:24Speaker 3: tool or the use of the tool. It's how they
- 30:26Speaker 3: sell it to the community, in my opinion.
- 30:28Speaker 2: Yeah, I think that's exactly what I was going to
- 30:30Speaker 2: get to was the part of, it's really the communication,
- 30:33Speaker 2: like you were mentioning, between the community and the police department,
- 30:36Speaker 2: the mayor, whoever is going to agree with it, they
- 30:39Speaker 2: should probably present it before they implement it and then
- 30:41Speaker 2: maybe even have a vote on it, I guess. You
- 30:43Speaker 2: can have some kind of weird vote to be able
- 30:46Speaker 2: to see if the community is willing to try it.
- 30:49Speaker 2: And if they sell it and they give it a try,
- 30:51Speaker 2: then it's a whole different ballgame.
- 30:53Speaker 3: Right.
- 30:54Speaker 2: Because I think if you just do it, then it
- 30:56Speaker 2: scares a little bit the community. they had no say
- 30:58Speaker 2: in it and something's being implemented where they're not really
- 31:02Speaker 2: sure how it works and all this other stuff.
- 31:06Speaker 3: I think that's a very valid point. I think that
- 31:10Speaker 3: the way I've often been told, and this is just me, Andy,
- 31:17Speaker 3: it's not what you say, it's how you say it.
- 31:20Speaker 3: And that's very true, I think, with a lot of
- 31:23Speaker 3: police departments, right? You said maybe if they introduced it
- 31:26Speaker 3: through a community you know, response, community outreach, before they
- 31:35Speaker 3: implemented the tools, then the community would see that it's
- 31:38Speaker 3: being used to help them and prevent crime as opposed
- 31:42Speaker 3: to oppress them and unfairly watch them.
- 31:49Speaker 2: And I don't know if it's a situation where law
- 31:52Speaker 2: enforcement says, you know what, but if we say too much,
- 31:54Speaker 2: then the bad guys are going to figure this out.
- 31:57Speaker 2: I don't know if this is one of those situations
- 31:59Speaker 2: or not, because I know some things, yeah, you want
- 32:00Speaker 2: to kind of keep quiet. I get it. Like the
- 32:02Speaker 2: shot spotters, you're not going to start showing, oh, look,
- 32:04Speaker 2: it's over there in that corner. So I get that.
- 32:09Speaker 2: I'm just kind of, I don't know, I haven't really
- 32:10Speaker 2: thought about it, whether or not this system would actually
- 32:13Speaker 2: be something that they would want to keep on the
- 32:15Speaker 2: down low, as they say.
- 32:16Speaker 3: Yeah, I think that it depends on the culture of
- 32:20Speaker 3: the community. And in some communities where you have a
- 32:24Speaker 3: very good relationship with the police department, I think that
- 32:28Speaker 3: these types of technologies are welcomed and embraced. And I
- 32:32Speaker 3: think where there's a lack of public trust in other communities,
- 32:36Speaker 3: then they're much more hesitant to have these and they're
- 32:38Speaker 3: more likely to cry out a civil liberty issue.
- 32:43Speaker 2: One final note from my side, at least. Not really
- 32:47Speaker 2: related to it, to the AI world, but related to
- 32:50Speaker 2: a tactic that was implemented to get a lot of controversy.
- 32:54Speaker 2: But I was stunned. The droppage of homicide per capita rate.
- 33:01Speaker 2: For the last, I don't know how many years now,
- 33:04Speaker 2: El Salvador has had a huge homicide per capita rate
- 33:07Speaker 2: with one of the top five in the world. Usually
- 33:09Speaker 2: it's like a hundred for every a hundred thousand people.
- 33:12Speaker 3: Yeah.
- 33:13Speaker 2: And just to give you a comparison folks, I think
- 33:15Speaker 2: like some of our most violent cities in the U
- 33:17Speaker 2: S is about 25 to 35 per a hundred thousand.
- 33:19Speaker 2: So this was like triple the most violent that we
- 33:22Speaker 2: have in the U S.
- 33:23Speaker 3: Yeah.
- 33:23Speaker 2: And it was almost, I mean, you look at those
- 33:26Speaker 2: hotspot maps and everything that was red was a big
- 33:29Speaker 2: crime area. Almost the whole country was red. It was
- 33:31Speaker 2: insane to look at some of these things. And this
- 33:33Speaker 2: was about 2016. Yeah, 2016 or 17. And then I
- 33:37Speaker 2: made the mistake, and I admit my mistakes. I made
- 33:40Speaker 2: a mistake a few months ago in a lecture saying,
- 33:43Speaker 2: you know, the homicide rate in El Salvador. But, you know,
- 33:46Speaker 2: they're implementing this new thing in the last year. They
- 33:48Speaker 2: put in like over 40,000 people in jail or prison.
- 33:53Speaker 2: So then I said, you know what? I wonder if
- 33:54Speaker 2: this thing is doing anything. So in the middle of
- 33:57Speaker 2: the class, I had a breakout for a second. I
- 33:59Speaker 2: decided to look at it myself. And I'll be damned,
- 34:01Speaker 2: that was wrong. It dropped about a 12 or 15
- 34:04Speaker 2: for every 100,000. It went from 100 down to about
- 34:08Speaker 2: 12 to 15. That's like an 80% drop. Yeah, right.
- 34:12Speaker 2: I get the civil liberty issue. I understand that. But then,
- 34:16Speaker 2: of course, they do the anecdotes where they go around
- 34:18Speaker 2: asking people. And they say, it's the first time I
- 34:19Speaker 2: can walk at night. It's the first time I can
- 34:21Speaker 2: do this. The first time I can open a business
- 34:23Speaker 2: and not worry about somebody strong arming me for money.
- 34:26Speaker 3: And are they using these technologies?
- 34:29Speaker 2: No, I have nothing related to that, but it was
- 34:30Speaker 2: related to an intervention that was controversial. I guess that
- 34:33Speaker 2: was my point about it. It was an intervention that
- 34:35Speaker 2: a lot of people said, oh, you can't do that.
- 34:37Speaker 2: And I get it. I understand it.
- 34:39Speaker 3: Yeah.
- 34:40Speaker 2: But boy, did it drop a lot.
- 34:41Speaker 3: I mean, what was the intervention? What was it?
- 34:45Speaker 2: Targeting these individuals and putting them away. So they were
- 34:47Speaker 2: just grabbing everybody who was part of these groups and
- 34:49Speaker 2: just boom, boom, boom. And I don't know if it
- 34:51Speaker 2: was a lack of corruption at this point compared to
- 34:55Speaker 2: other past administrations. I have no idea about that. If
- 34:59Speaker 2: it was a coordinated effort with everybody, did they get
- 35:01Speaker 2: money from another country? I have, you know, sometimes that helps. Look,
- 35:05Speaker 2: I can't do this because they're funding these corrupt politicians.
- 35:08Speaker 2: If you help us out, I'll do this. I promise you.
- 35:10Speaker 2: And maybe that was the case because I've never seen
- 35:12Speaker 2: a change like that. I mean, you can look at
- 35:14Speaker 2: those numbers. Maybe we'll talk about it next time. That
- 35:15Speaker 2: was really a surprise how fast that drop was.
- 35:18Speaker 3: Well, you know, it's funny, too, you know, not funny,
- 35:20Speaker 3: but ironic that in places like San Salvador, and I
- 35:24Speaker 3: actually lectured in El Salvador, San Salvador, back in 2015,
- 35:30Speaker 3: I think, or 14 or 15.
- 35:31Speaker 2: That was high problems right then.
- 35:33Speaker 3: Yeah, and I flew in. When I flew in for
- 35:35Speaker 3: this lecture, I came in, and then when I got
- 35:40Speaker 3: off the airplane, just outside the airport, you're driving and
- 35:45Speaker 3: you see these people armed shotguns or M16s, but they
- 35:50Speaker 3: weren't in front of like government buildings. They were in
- 35:52Speaker 3: front of businesses, like everyday businesses, like be in front
- 35:56Speaker 3: of Walmart. There's a guy with an M16. It was
- 36:00Speaker 3: very shocking to me, the number of guards in front
- 36:04Speaker 3: of these houses and businesses, right? Part of the problem in,
- 36:10Speaker 3: you know, these, these, violent cultures, particularly in South America,
- 36:13Speaker 3: I think a lot of it has to do with
- 36:16Speaker 3: the culture itself, the honor culture. So if, you know,
- 36:19Speaker 3: and they, in the number of gangs and organized criminal groups, right.
- 36:24Speaker 3: And if, so if you disrespect somebody in one of
- 36:27Speaker 3: these cultures, the response is usually violent. The other thing too,
- 36:32Speaker 3: is you talked about the, the crackdown and the response,
- 36:36Speaker 3: the laws in many of these South American, Central American countries,
- 36:41Speaker 3: They don't have a death penalty. And the reason is
- 36:45Speaker 3: because historically, most of them are Catholic. And so the
- 36:49Speaker 3: Catholic Church has been against the death penalty forever. And
- 36:54Speaker 3: so you could kill a cop in some of these
- 36:57Speaker 3: South American countries and you're not getting the death penalty.
- 37:00Speaker 3: You could have premeditated murder. You're not getting the death penalty.
- 37:04Speaker 3: So these individuals go to jail and oftentimes they're not
- 37:09Speaker 3: in jail for the rest of their life either. And
- 37:12Speaker 3: so it creates and breeds, I think, a violent culture
- 37:16Speaker 3: in many of those countries. I was interested to see if,
- 37:21Speaker 3: in fact, the drop in crime was due to these technologies,
- 37:25Speaker 3: because the response from police in these countries is also,
- 37:29Speaker 3: we talk about the violence of our police departments here
- 37:32Speaker 3: in the United States. They are almost paramilitary in those
- 37:38Speaker 3: countries south of us, so.
- 37:41Speaker 2: I think it would be a great discussion for our
- 37:44Speaker 2: next podcast is maybe have an international or somewhat of
- 37:47Speaker 2: an international look at crime and policing and seeing the differences.
- 37:51Speaker 2: Because I'm always bugged when people try to compare it
- 37:53Speaker 2: because they're not apples to apples.
- 37:55Speaker 3: Not even.
- 37:56Speaker 2: No, not even close.
- 37:57Speaker 3: And it's funny because I taught police executives for 15
- 38:02Speaker 3: years at the National Academy. I had a student, I
- 38:07Speaker 3: believe it was Newfoundland. No, it was Greenland. Greenland. And
- 38:13Speaker 3: they don't have guns. They don't carry weapons. If there
- 38:16Speaker 3: is an issue on the island, they have to go
- 38:18Speaker 3: to the headquarters and the gun vault and they get
- 38:21Speaker 3: a gun. So, you know, there's a broad spectrum. And
- 38:27Speaker 3: I think it certainly would be worth a podcast to
- 38:30Speaker 3: look at policing around the world and how those processes work.
- 38:37Speaker 2: I think so.
- 38:38Speaker 3: Yeah, I don't think the United States is nearly the
- 38:40Speaker 3: most violent policing culture or the most passive. I think
- 38:45Speaker 3: it's somewhere in the middle.
- 38:49Speaker 2: Again, that's hard to compare.
- 38:51Speaker 3: It's very hard.
- 38:51Speaker 2: We can have an easy another podcast right now. So
- 38:54Speaker 2: let's be careful. I can tell already we're digging that
- 38:57Speaker 2: rabbit hole pretty fast.
- 38:58Speaker 3: Yeah, no, I think that'd be a good a good
- 38:59Speaker 3: topic in the future.
- 39:01Speaker 2: Sounds good to me. Again, folks, hey, check out BehavioralScienceUnitLLC.com, BehavioralScienceUnitLLC.com
- 39:08Speaker 2: to find out what Andy's up to and to learn
- 39:10Speaker 2: about his services over there. And by the way, folks,
- 39:13Speaker 2: you can also make sure to share, subscribe, and hit
- 39:16Speaker 2: that like button if you like our show. Of course,
- 39:19Speaker 2: hopefully if you're listening, you do. Thank you, everyone, for listening.
- 39:22Speaker 2: Stay safe. Watch out for those deep fakes and whatever
- 39:24Speaker 2: else AI is up to.
- 39:25Speaker 3: And the deep snow in California.
- 39:28Speaker 2: Yeah, the deep snow.
- 39:29Speaker 3: All right, buddy. Talk to you later.