ShotSpotter is a private company leveraging audio technology and proprietary algorithmns to locate gunfire with an incredibly high degree of accuracy. It's currently deployed across hundreds of square miles of the US -- many proponents argue it saves lives, but critics allege it has some serious problems... including corruption and conspiracy. In the second part of this two-part series, the guys explore the claims, concerns and controversies surrounding ShotSpotter -- including the stories other media made about the company, before those claims were retracted for inaccuracy. They don’t want you to read our book.
From UFOs to psychic powers and government conspiracies. History is riddled with unexplained events. You can turn back now or learn the stuff they don't want you to know. A production of I Heart Grady. Hello, welcome back to the show. My name is Matt Our colleague Nol is not here today, but we'll return tomorrow very soon. They called me Ben. We're joined as always with our super producer Paul Mission controlled deconds. Most importantly, you are you. You are here. That makes this the stuff they don't want you to know. Before we get started, shot Spotter did reach out to us regarding some of the statistics provided in our episodes and their position with respect to the Williams case. Shot spot a disclaims all responsibility and Mr Williams incarceration and disagrees the associated press is factual assertions regarding changing locations. If you're interested in reviewing additional information from shot Spotter on these topics, please visit www dot shot spotter dot com. Heads up, fellow conspiracy realist. This is part two of a two part series. We are continuing our discussion of the famous or infamous, certainly controversial technology known as shot spot Er. Previously, we discussed the terrifying reality of gun deaths in these United States. We also looked at the ways in which law enforcement, local and federal level political bodies, and average average citizens right just the folks you would see on next door trying to figure out a solution. Um. And it's funny, Matt, because in the interim between you Noel and I recording Part one of shot Spotter and Part two of shot Spotter, I I went out, I hit the streets and I looked for some shot Spotter arrays, and I'm pretty sure I found some. Maybe they're not brand name, but I'm pretty sure. I like knowing now what we know about sort of the combo meal of how these things are set up. Just for a quick refresher for everybody. Shot Spotter is using some pretty impressive algorithmic software and some pretty cool microphones. In part one, we were not able to discern the specific brand name of these microphones, but we do know the following there's something like fifteen to twenty per square mile, and these microphone arrangements also include the following a GPS and onboard computer transmitter and some never we have some of the algorithmic stuff on site, so it can quickly do a calculation. Is that a dump truck? Is that someone with a legendary fart. I think we nailed down the science, right. No one can fart such that shot Spotter will mistake it for a gunshot. Right to our knowledge, that's not possible. It hasn't happened yet. That isn't to say, you know, some fast food company will invent something to allow that to to occur. Just on the microphones, I'm going to say, their quote audio detection devices unquote are likely proprietary. So they probably are labeled shot Spotter right as like a brand name or something like that, And we just don't know what they are be because it is it's technology that they've invented or at least patented, right. So the big thing to be aware of its expensive for a police department to run this thing in a city, and it costs taxpayers a ton of money. And what we're gonna talk about today is that the system itself. While the company you know says, hey, this thing works great, while there are a lot of people on the internet you can find that say hey, this thing is amazing and it works great. A lot of listeners to stuff they don't want you to know are saying the same thing, yes, or at least they're aware of how function it seems to function well. But there are a lot of detractors, people who say, there's a whole other side to this coin, and we're going to explore that today. Yeah, and when we do a two parter. We don't do these super often, but when we do a two parter, it's usually because we have, um, we've got a little bit deeper in the rabbit hole, and we didn't have time to make a brief. You know, That's one of the infamous jokes about writing Matt, which I'm sure you've heard, is I can't remember who it was. Fellow conspiracy realists help us out here. But there are multiple instances of these writers corresponding with each other and sending their version of an essay length text and ending it with sorry in a hurry, didn't have time to make this brief, so we didn't We didn't have the time to make this brief. And you know, when Matt, Noel and I are hanging out on air or off, we often run into these sorts of these sorts of moments where we say, how can we accurately concentrate the facts, the controversy, the conspiracy into something that doesn't become a four are our long weird hangout session. If you want to learn what our endless hangout sessions are like without the you know, without the constraints of podcasting, do check out our book. Stuff. They don't want you to know. We we even had this is how weird we get with it, folks. We even had um our publishers come to us several times and say, hey, this chapter is a little long, but I think we're able to get a lot of stuff in there. What do you think. Yeah, to my knowledge, we were able to fit everything we wanted to in there. We did have to pare it down to government only stuff from all the research we gathered right right the the cryptic chapter cryptozoology chapter is one of my favorites, and maybe oh yeah, yeah, oh gosh, we went deep on that too. Basically, we have another book, is what we're saying. We kind of have another book that will have to will have to be on the way too soon, and the best way to help us make that happen is to check out this first book. We hope that you enjoyed it's available wherever you find your favorite books, whether that's your local bookstore, whether that's Amazon. Well, yeah, and and let let daddy bezos know how you feel about the book by going on Amazon and reviewing it. Seriously, if you've got any kind of Amazon account and you do like the book, please review it. That that helps a ton, it does it? Weirdly helps a ton? Uh? And yeah, we want to be we want to be your five star crew, you know what I mean. And because we're asking, we're asking you, our fellow listeners uh for for a little help at the top in terms of ratings, especially, we want to give you something in return. Right, So we spend a lot of time and shot Spotter, here are the facts. Back to the backstory. Okay, Matt, you and I haven't spoken for a weird amount of time for us, because we've both been doing some crazy things. Matt is executive producer on one of my favorite seasonal shows, Thirteen Days of Halloween. You and are pal Alex Williams, and I guess our pal Clancy Brown. We can say it now right, Uh? Yes that guy. Uh. I've been working on this wonderful horror show and it's coming out now as you hear this episode, you should be able to hear all thirteen episodes of Thirteen Days of Halloween. I'm super biased. I always get a chance to to write a couple of episodes and it's um it makes my year, honestly. But in that interim, I was also working on a secret project, which may hear Matt and Noel allude to cryptically in our listener mail and strange news segments. I hate to miss stuff, but it's it's been that kind of thing, you know, been in the absence of transparency. M hm, speculation. That sounds good. Who wrote that? So the the thing I found about shot Spotter, It's fascinating. It was I I found, but had not have not confirmed, some sort of audio tracking relays here in the neighborhood of our old office. Additionally, you know, as we get, as we get closer and closer to Halloween, I heard more and more gun shots, and I heard two very close gun shots, and I was thinking back to um an earlier conversation we had. I can't remember whether it was on air or off where we were. I think it's when we were reading letters from fellow listeners in law enforcement and we started asking each other which of us had ever called the police to report a gunshot. I don't remember what you said, did you know? For me, it was pointing out that where I lived previously, I would hear gunshots pretty consistently. Um. I would also hear fireworks, and it was difficult at times to tell what was what. Depending on the distance right you could you could kind of make it out. If I was, like, if you're standing outside right at the time when one of these shots or fireworks occurred, you could usually discern and you could kind of tell even the caliber, often depending on several factors. But what it sounds like, you just kind of get used to the sound of certain weapons. Um. But if you're inside your house, especially like if it's late at night, many people, including myself, wouldn't be able to discern and then likely wouldn't call the police unless you know, you're a particularly I'm not crabby person, but somebody who's really concerned and is often in contact with the police, right right, right, right, Um, what may be called a busy body sometimes. But also we have to we have to remember that, you know, there's a reason good Samaritan laws and so on exists. Many people are rightly concerned about engaging involving themselves in this situation, and so often people won't call contact first responders or emergency services unless they feel those gunshots are aimed at them. Even then, you know, if you think about it rationally, if you call your local police department, depending on how close you are to that department, by the time a shot has been fired, you've heard it, you've called the police, they've dispatched an officer or two, and then the officer makes it out there, it's very unlikely the devil police officer will be able to do anything to prevent further crime or stop whatever is happening. Uh. It is often if there was an actual shooting, that officer is going to be a first responder to a gunshot victim rather than you know, stopping a crime. The response time is huge. Also side note, if you are ever in a situation where you feel someone is outside your house or threatening you with entering your house or something, or you know, your apartment, wherever you live, your domicile, always remember a lot of people mess this part up. Don't say I'm going to call the police or I'm I'm going to call nine one one say I have called, they are on their way. You know what I mean, Because what you're doing if you threaten to call instead of just making it is you're essentially telling people there is a time interval to to your point, Matt, there's a time interval before these folks arrive, right, And Look, this can be nasty stuff, but the idea of shot Spotter and similar technologies is to fight against this, to mitigate that interval of time between a gunshot being fired and someone getting to the scene. When it works, as we explain in Part one, it can definitely absolutely save lives, save lives of law enforcement individuals, first responders e m t s can know what they're getting into. Uh, it can save the lives of civilians. It can also help cut short spree shooters, you know, mass shooters who might have might be proceeding on foot through an brewood right or through some certain region or community. So this can make a difference when it works. Yeah, and according to shot Spotter, it also increases the likelihood that ballistics evidence will be recovered, which would help close a case if there was an actual crime that occurred that led to a trial or something. Yes, perfect, that's perfect. Foreshadowing here because you see part two of this of this series could probably be called shot Spotter goes to Court. So what happens what happens when this private company it starts helping save or put people away. We're gonna pause for word from our sponsor and we'll be back. Here's where it gets crazy, Matt. You set this up so very well. That's one that's we're back in the land of controversies about this technology. And again, to be clear, we are exploring this stuff that critics and proponents have said about shot Spotter. Additionally, shot Spotter is not the only audio tracing software or system out there. It's like how Coca Cola is not the only soda you can buy at the grocery store. It's just the one you might be most aware of. So shot Spotter probably gets a lot more public attention because it is more pervasive than a lot of these audio tracking other audio tracking private entities. What you need to know is there is more than one. However, shot Spotter is the and that comes up the most because they right now have the most data out of again the private companies. And we're going to get into some sticky stuff about the future man. I know you saw the stat Matt shot spotter itself. How accurate is it? Right? Like, how many times has someone responded on the ground to a shot spotter report and found it was just a dump truck, a superhuman farter road construction, you know, stuff like that. How accurate is it? I saw the statistic and the statistic feels bold. Yeah, it is a bold statistic. Shots Water claims that their system has a nineties seven percent accuracy rate when it comes to identifying gunshots versus other loud sounds in a city where it's placed seven percent. It's very important to remember that is accuracy into scarning whether or not the sound emanated that it's tracking was a shot right, right? And and with that statistic, we also have to acknowledge that statistics can be so slippery. What they're saying is cent of the time they are right when they say something is or is not a gun shot. That's not the same thing as saying cent of the time they're right about the specifics of the firearm, not at all. It's just it's like up there with um. You know, before Google went full alphabet, they had that motto don't be evil, and we used to We used to joke back and forth about how there's a world of difference between don't be evil and be good. You know. So, so they're not saying of the time they can tell you, you know, this was an a k versus something else. Right. What they are saying is that they know the difference between the stuff that is just you know, the burps of farts of your average city versus actual discharge of a firearm. The thing is a lot of folks dispute this. One of the biggest customers for shot Spotter, and specific in the United States is the beautiful city of Chicago. Chicago, we miss you. We're gonna hit the road soon and say hello in person. But the thing about Chicago for our purposes is that as it is such a huge customer of shot Spotter, it also generates a lot of data to pour into to look at this claim of accuracy. The conspiracy here from the critics is the idea that shot Spotter is uh being purposely misleading about that percent statistic, which again is bold. Another other thing we have to realize about this statistic is you will see different claims. They're They're always going to be you know, a pretty high percentile, but you will see different specific claims about the accuracy. We went to shot Spotter's own website, we looked at their literature and the nine seven percent claim is from twenty nineteen, and they said the system overall had an aggregate accuracy rate across the entirety of their customer base, which you know, you and I work with marketing departments. We love ours. But it does sound a little bit like marketing, doesn't it matter? Yeah, it does sound quite a bit like marketing. It's strange. They're in that same statement, they're saying there's a very small false positive rate of less than half of a percentage of all reported gunfire incidents. That's that's inten okay. Yeah. They also point out in their own literature that the system and the process are evolving constantly. Right, work in progress. We are building the playing that we are flying, which can be taken one of two ways. One it can be seen by the more optimistic as an inspiring thing. You know, one day, right, there will be a hundred percent accuracy. The other, more cynical response would be, hey, why don't you finish building it before you put it out? You know what I mean? And that just I call it the Larry David scale, Like, how Larry David, are you going to be about some stuff? But this is unfortunately, this is not you know, this is not complaining about the big salad or whatever. People's lives are at stake, and the question becomes, is this a tool deployed in good faith to save human lives or is there's something else at play? And that's that's the scary part, and that's why we have to talk about these statistics, which get a little crazy. So Chicago huge customer for shots Spotter. Shot spotter itself says percent accuracy overall as of twenty nineteen. Other cities, other local journalists in the community will will have better stats about how shot spotter functions in their neck of the global woods. Uh now, I think I'm thinking of that San Francisco Examiner article right when they point into it a different um was that different overall accuracy rate for shot Spotter? Uh yeah, I can read you a quote from it. It comes from seventeen and again it's a statement from shot Spotter. They, according to this article, guarantee accuracy of the time and that is speaking again. Well, I don't know, I don't know the context here. I guess that's in detecting an actual gunshot. Yeah, and that comes from the CEO, Ralph Clark. Ralph Clark is named in multiple articles on shot Spotter. Often he's talking about things he feels journalists got wrong, you know what I mean. And and it's another get we made a joke about in part one. But shot Spotter is litigious, so they're I mean, they're not scientology level litigious. But that's why if you look at outfits like Associated Press or Vice News, you will see some specific retractions or clarifications published. Shot Spotter wants people two avoid criticizing it, which is, you know, that's what every business wants, right. However, people are disputing the statistics, and the people who are disputing these claims are not um, you know, your uncle on Facebook or whatever. There are places like the MacArthur Justice Center in Chicago. They studied more than forty thousand shot Spotter dispatches and they found that eighty nine percent of those dispatches resulted in no gun related crime. That's a quote from MacArthur. Uh, and they said eighty six percent resulted in no crime at all. That's interesting. We see another three percent discrepancy, right, So so it's like the did the cops just show up and someone was like throwing rocks and orphans and they were like, that's not a gun shot, that's a crime. Well, again, that's what we're talking about. It doesn't matter how fast and accurate shot spotter is in detecting a gun shot. By the time an officer is able to arrive on scene, there may be no evidence of a gun being fired. Theoretically, there would be shell casings on the ground if someone just was firing a weapon and then you know, got out of there quickly, as is often the case, and they would probably not take the time to pick up any shell casings. But it depends on where the thing where the shots occur. It depends on what the you know, ground looks like. There. Is it gravel, is it dirt, is it asphalt? Is it, uh, you know, tall grass, And there's so many factors. It's tough for me thinking about that resulted in no gun really crime, no crime at all when it comes to because that really doesn't mean that shot spotter is failing. That means that when officers arrive on scene, there's no evidence of a crime. There's nobody to apprehend that, there's nobody to even probably talk to unless there is somebody there, right, and then they become highly suspicious, no matter what they had to do or didn't have to do with firing a weapon. You know, we're talking about someone who is maybe just hanging out on the street, maybe someone who's unhoused, right, maybe someone who lives there and is on their porch, you know what I mean. And that's not that's not bad police work either. That's hey, let me get an eye witness, you were here, tell me what's going on. Or it could be bad police work too. It could be because that person, that person could automatically be seen as a suspect as or somebody potentially had something to do with the gunfire. So it's just it's especially if that person has priors, right, if they have a jacket already, you know what I mean. This is amusing jack as the slang term for your record in a folder in an incarceration system. But if they already have some priors, then of course suspicion automatically falls on them. You could be a good person in the wrong place at the wrong time, and all of a sudden you're jammed up, But consider the other side. We're still in Chicago with this um. The Chicago Inspector General kind of backed up these findings from the MacArthur Justice Center and they found that this was expensive for police departments. Just like in part one, we talked about this really fascinating deep dive into St. Louis, Missouri, uh in Police Chief magazine, which I still didn't still surprised that's a magazine, but but there's this deep dive into the hidden cost of systems like this. Four police departments and therefore four taxpayers in those cities, in those towns, in those communities. Anyway, Chicago Inspector General backs up MacArthur Justice which also backs up some of the findings in St. Louis, and they say, these police officers are changing their practice. They're stopping and searching people, right, stopping frisk basically for no other reason than they were in that place where shot spotter frequently gives alerts in the vicinity. Yeah, so not having to do with with responding to a shot spotter hit. Yeah, not a specific pop just hey right, that sounds a little bit like an attempt at predicting crime. Just a little bit. Sounds like trying to get in front of something, but it's also missing the big step one. Why are there many shots spotter alerts? Because you have shot spot ter there, you wouldn't have those alerts because there would still be the old dispatch who calls the police kind of system. But also some l e O s are not super into this law enforcement officers because they're thinking of the time it takes. Right, if I normalize responding to this, I get the call, I get the buzz, it sounds like it's legit. I'm on the way, it's over. There's no victim, you know what I mean, there's no case sing, there's nobody on the street to even you know, detained for a while. Well, then how many other calls have I missed while doing that right in a resource strapped environment? Well, and if you also, I think it would be hard for a police department to justify not responding. If you're shot spot er software goes off right or the hardware detect a gun shot, how could you justify not investigating that or sending out a car to at least check it out? Uh? Yeah, which is puts a real drain. Ye shot spotter told us there was automatic gunfire at a place that has a lot of reports of gun crime, and we just chose to ignore it. That doesn't look good. And there are all of these factors that we're considering in this episode, and it is obviously extremely complicated. This is not a clear cut issue, right. Uh. One of the ways that we could better evaluate this is if we had more hard information, information that wasn't let's say, put together by a sales team to go out and sell more shot Spotters in more cities. If we if we could actually see how does this system actually function in the real world. Um, but we're not really getting that. We are getting specifics coming out of cases that have arisen because of a Shot Spotter call. Though m HM and I love what you just did. Their men. You subtly slid in a reference to a quote by a Shot spot Ter employee, Paul Green, who said, quote, our guarantee was put together by our sales and marketing department, not our engineers. That's referring to the accuracy quote right right, and then, as we as we know, was apparently seven percent according to their marketing folks. So you point out a great thing we can look at specific cases. We can look at something like the case of Corey Ali Mohammad, who was a mass shooter apprehended in April with help from shot Spotter. So being able to triangulate where these shots were occurring played a massive role in helping law enforcement in Fresno catch this guy before he killed more people. Um, this was you know, this was a hate crime, racially motivated kind of spree killing, and more people probably would have died if the law wasn't able to get to this criminal when they did. And that's that's the shot Spotter story. But the thing is, that's like one, one specific case, right, that's a success story, as weird as it sounds, because this guy did get convicted of four murders, but it's not five, it's not eight, and you know it's not twenty. And that's what shot Spotter saying. They helped create the alped to mitigate the situation. And this is the thing. Shot Spotter is supposed to be a diagnostic outfit. It's supposed to be an investigative tool for the police. An investigative tool is not the same thing as primary evidence in a court of law. In the United States, evidence is a different thing, and so some of the critics are saying this is a bridge too far, This is the wrong rubicon to cross for shot spot Er data to be shared out as evidence in a court of law. And this is not just us on on stuff they don't want you to know saying that stuff, is saying this can be a sketchy operation. This is a particular concern to places like the a c l U. They say they had a quote that really stood out to me because their accusation is that shot Spotter is kind of cherry picking the data, the information that it does or does not share, and it's not showing its work either, but it's doing it in court, meaning that people can go to prison without transparency on this stuff. Yeah, at best, I think shot Spotter could and probably should be used as supporting evidence. You can if you're going through a trial and as a prosecutor you are trying to prove that a gun was fired in this vicinity at this specific time, you would have time stamps associated with that weapon being fired with shot Spotter. But again that's not that's not the primary evidence to prove that whoever your suspect is, is the person who fired that weapon that is just supporting the fact that a weapon was fired at that time. Does that make yes? Yeah, I was just I don't know if we're on video, if this will make it into into our YouTube channel, check that out. But I was, I was doing a lot of non verbal hyping and agreements. Yes, you have described it better than I could have. Supporting evidence. There's a line, you know, so a c LU says, in particular, they say quote shot spot as methodology is used to provide evidence against defendants in criminal cases, but isn't transparent and hasn't been peer reviewed or otherwise independently evaluated. That simply isn't acceptable for data that is used in court. A c l U is still on that mid mid Atlantic or transatlantic voice, and then he goes further than this. There are allegations of some knoughty things going on at shot Spotter and between shot Spotter and police departments and other law enforcement agencies, talking about some corruption, some cover ups, the things that this show likes to talk about. And we're going to address that right when we return from a word from our sponsor, and we have returned cover ups, corruption, conspiracy. There's a lot here, and just to walk you to like mid rabbit hole. Let's call this mid rabbit hole because we can't we can't go too far until the next book, right unless you know someone's legal department or our own legal department comes for us. So being objective here, here's what we can say. Mentioned Vice News Associated Press earlier. They stated with what appeared to be pretty compelling quantitative evidence that shot spotters human analysts were changing their judgments on results. So that's the human peace right We know that um shot spotter engineers and management themselves that said, yes, we have algorithms, we have some very fancy monitors and microphones, but we always have a human they're at the past right two to help steer. There's the same reason, or the same way in which autonomous vehicles in testing phases still have someone sitting behind the driver's wheel. Right, there's still some there's still a body there, is what they're saying, Well, there is a body there, and that's one of the big things that shot spot are always tells there's a human being making the determination. Highly trained human being will be able to discern the difference between a gunshot and whatever backfire. And this is super tricky because shot spot are also teuts that that analysts, that highly trained person is going to make it to termination within I think a minute or ninety seconds or something some time frame that is super short. So if that person does make that determination initially this was a gun shot. The coordinates that were given to us by the shot Spotter system are here, and it's a gun shot for sure. Then that person goes back and changes it to something else, after officers have already been dispatched, Like is that for shot Spotters records? Is that for, you know, so that they can change their accuracy rating and put it on their website, Like what is that for? Why would they do that? Or hypothetically, what if it gets changed after conversations with with the local law enforcement right before something goes to court. Spooky stuff, you know. This is a shot Spotter, by the way, disputes this idea of the company's analysts changing those judgments, and they dispute I mean very diplomatic when I say dispute. By the way, they dispute the idea of law enforcement being able to put their thumb on the scale right and say no no no no, no, no no no call this a gunshot or or no no no no, no, no, change the time stamp here. Uh, they did have Vice and associated Press change part of their original reporting after threatening lawsuits. So if you read these Vice articles, if you read the AP articles, you read a couple other things, you'll see at the very end there's this, Hey, we got this court case and this specific thing is not exactly the way it may have sounded in the original article. But that's not the same thing as saying, oh wait, there's no problem at all. There's huge problems. There are huge problems. I don't think that's a hot take. I don't think that's too far for us to say that there is opportunity for corruption and conspiracy in this thing, right because also, I mean this is this could run into some of the same problems as um automated facial recognition technology bias against people of color. It's a little more complicated because it goes into intergenerational things, right, kind of like how um god, I love that food deserts episode. Man, Big food deserts are an accidental conspiracy. They're just people not knowing, people in power, not understanding nor caring about the full consequences of their actions for short term gain. And now you know uh, with shot Spotter. We see you know, really went to bat with him. I know you read this. Two man who really went to bat with shot Spotter was the A C l You And they're like clarifications in corrections. Part of their report is much longer, and it's all about how they had a fight on the phone and how the the CEO called the folks at the a c l U and said no no, no, no, no no no. But you know the A c l U is lawyers, so threatening them with a defamation lawsuit that's not really going to swing the needle. So back to our guy Clark, the CEO of shot Spotter. He was saying, look, these algorithms are doing the math. They're not spooky robo minds. They're doing basic math that could be done by hand, pen and paper. They're just doing it faster. Again, it's all about the response time to your earlier pointment. But he then he talks about the more complex algorithm. Both of these are proprietary, by the way. That's the black box stuff. Also, the CEO, Mr Clark does not like us using the phrase black box, but it kind of is a black box. No one's been under the hood, No one can peer review this stuff. That's a huge problem if you're going into a court of law. Right you're saying, Hey, I have proof that this is the kind of gun that was fired here. So what's your proof? Do you have by witnesses? No? Do you have a recording? Yes? How does it work? Magic? It just doesn't. It doesn't look good. And I again, I'm not trying to dunk too hard. I know these are good faith actors, but if you can't show your work, then the law gets really tricky, right, Yeah, it does. It gets extremely tricky. Why don't we jump to what the other critics are saying, Like there's a there's a lot of criticism of this thing that we're talking about Chicago. Earlier in this episode we talked about nineteen and as of I believe it's this year and maybe the summer of two. There's a case right now, a lawsuit that is attempting to remove shot Spotter completely from the city of Chicago because there's a thirteen year old boy named Adam Toledo who was killed, and there's a ton of information coming out about how shot Spotter technology is actually being used in the city right now, And I want to stay in Chicago there because we're it's had problems for a long time with this system. In March of this year, the Associated Press reported how there's a person named Michael Williams living in Chicago, but he went to jail in August one for the murder of of somebody in his neighborhood. Okay, he went to jail for murdering somebody. And a key part of that evidence, guess what it will us It was a quote clip of noiseless security video showing a car driving through an intersection and allowed bang picked up by shot spotter. Uh yeah, okay, okay, So again, is this wrong place, wrong time, or is it working perfectly? Is that exactly what happened? Or is that exactly what happened? How would you put how can you put those things together? I mean, if they if they're time stamped, then logically it would make sense if you can track that sound down and you know, triangulated the way shot Spotter allegedly does and it's supposed to do, then maybe that's great evidence, but it does seems it seems tough to connect an audio and video feed from separate sources like that and call it enough evidence to send somebody to jail. Yeah, that's that's the thing. Like, um, he does go to jail. Williams ends almost a year in jail. And this is this guy has lived in this neighborhood for a while. He's got grandkids, you know what I mean. He's not some hot shot, unhinged, lunatic nineteen year old pulling pranks for TikTok cloud. He's a dude. He's a dude who has children and grandchildren. He's not the person you would look at and automatically think criminal. But Shah Sponner said he was because it pinpointed where that gunshot came from. Supposedly. And again, like we said in part one, the that system, when it works, it has a high degree of fidelity or sophistication. Right, I don't want to say accuracy, sophistication, I'll say but but good news for Williams here. Eventually the case is dismissed. He us is almost a year of his life at an age where one year matters quite a bit. And the prosecutors dismissed this because they ultimately say they have insufficient evidence. They say, this video footage, this, uh, the statements of people in the neighborhood, and particularly this data from shot Spotter, which again overall the process cannot be pure reviewed. They say it's just not enough. So Associated Press goes on to say, look, our investigation found some really troubling stuff here, like some Orwellian style surveillance economy kind of stuff. Uh, it's the idea that we'll just give you the quote already did a quote? Do you want to do this quote? Yes. According to the AP quote, shot Spotter employees can and often do change the source of sounds picked up by its sensors after listening to audio recordings, introducing the possibility of human bias into the gunshot detection algorithm. That's a great point. In the end, it is a human using their training and their understanding to make the determination and possibly leveraging their relationships. Right because still, you know, if you're listening to this and you're human, you know the score, you know the game. You know humans are often so AP continues and they claim quote employees can and do modify the location or number of shots fired at the request of police, according to court records. And in the past we're still in this quote. And in the past, city dispatchers or police themselves could make some of these changes end quote. Not a good look. Well yeah, and of course shot spotter says, no, that is not true. No, no, And they say that too many claims that are similar to this, um And ultimately it's not it's not necessarily like you're completely wrong. It's more like this is misleading. This isn't quite accurate, this is not accurate. Nice. Yeah, I mean, of course, of course there are going it's a business. This is not a nonprofit, this is not a foundation. They're out here to hustle and sell and hopefully, um, hopefully save human lives. Again, that's that's the idea. But because it's a for profit business, what are they going to say? What are they can be like, oh, geez a, p you got me, good game. No, they're going to use some of these massive profits to sue the Christ out of people who are making them look ad. That's just rational actors. Right. So shot Spotter spoiler Alert vigorously contest these and similar claims. And like you said, Matt, there saying these are missing nuance, misleading, they don't paint the full picture. Look, we require this human participation. Our algorithms are constantly being evaluated tweaked you know what I mean, we're building the plane we're flying, and we think that as a good thing, just like Vice. Associated Press later issues a statement correcting their original claims because both articles are talking about uh, these court cases, right, specific instances of court cases. But this is a moment where we need to exercise not um analysis of anecdotes, not analysis of specific cases. We need to exercise structural thoughts. Right. The problems are bigger than one sick tree in the forest. We need, we need what we need. What the suits at our job called the what was it? There's the forty foot view and then high level views. Yeah, I mean you go, you can see most things in your general area. Yeah, the full picture, sure, the whole shebey. So shot Spotter is still a business. It is actively is actively expanding or attempting to of course, as all businesses do. It is actively you know, um monitoring about nine hundred and eleven square miles of the United States twenty four hours a day, seven days a week, no holidays. But a number of cities are spooked by this, by this court of public opinion and stuff going on, right, and they've stopped using shot Spotter technology because they going back to what you pointed out earlier, Man, they have decided that shot Spotter creates too many false positives and false negatives. Could you break down for us what what is a false positive and false negative? In this context? False positive would be shot Spotter says there was a gun shot when there was not. A false negative would be shot Spotter missing a real gun shot that occurred in the city, missing a gun shot, calling it something else. Right, Oh wow, Yeah, that's just a fart, That's just a dump truck. I gotta get off this fartest gunshot thing. It's it's ruining the show. But the but but so a false false negative. It's kind of like going to a McDonald's and they don't have fries. Right, that's exactly right. Isn't this your job? Is it? Is it that like your main thing to do? Yeah? What profits center? Right? One of McDonald's biggest profits as a corporation is actually real estate. Just like Target, the department store here in the US makes a ton of money off surveillance systems. We're fun at parties here. We know we're going long. But one last piece of this episode is incredibly important. We hope We have done an all right job of giving you the lay of the land, the forty foot view, and showing you some of the support, some of the reasons why people believe in systems like this, and some of the reasons why people are concerned. But we have to end on the future. Shot Spotter is not going away. Technology like this is not going away. It is evolving. We found this stat the Department of Justice says that overall, UM, quite recently Uncle Sam has spent six point nine million dollars on gunshot detection systems. This includes shot Spotter. But this number, it feels kind of small. Six point nine million, right, but that doesn't count many more millions spent by state and local governments. It's just the fetes. This is a good business, and we we kind of teased cynically the idea of predicting crime. This is Yeah, this is the spookiest part, at least for me, Matt. This is the spookiest part. Yeah, because shot spot already has all this technology. It's already in all of these cities across the United States. And then in they went and did a little thing called an acquisition. They acquired another company. Sorry, in the words, UM they acquired a company called hunt Lab, and then hunch Lab takes its AI model and integrates it with shot spotters shot detection models and guess what they can predict where gunshots may happen or are going to have a an increased likely likelihood of happening. Wait wait, wait, wait wait wait, the algorithms and the machine learning and the data can predict or the marketing department says they can predict because this is part of the problem. Right, Let's put it this way. It's again it's building the plane as you go kind of thing. So it will likely increase. Their accuracy will likely increase if they put it in enough places, if they have it active for long enough, then it probably will increase and it will get more predictive. But right now, who knows, well, I mean they've had a couple of years. We'll see what happens. How is that not violating fundamental aspects of US governance? Right, Innocent until proven guilty, you know, like you are, you are prejudged. That's the idea. Ultimately, that's that's where this goes, right, the we and we don't have visibility on how granular this might get. And I know that's how that also sounds like corporate gobbledegook. But since we don't know whether they're saying hey the month this month in December, we we statistically will see higher levels of carth after something versus three forty two pm April seventeen on the corner of ninth and two below. We need to have two cop cars because they're going to be there's going to be a heist, you know what I mean. We don't know how how close they can zoom in on the future, but they're definitely trying to zoom in on the future a little bit. And that should scare you because there are no real laws yet. There's no real precedent to protect civilians right and to protect law enforcement. To be about it, there are no real laws in effect to protect human beings in a situation that becomes closer and closer to prophetic algorithmic time travel. And I know I'm like styling a little bit on this Welcome to our creepy accidental spoken word performance. But but you do need to worry about this. Whether you are a supporter of things like shot spot or whether you are an opponent. The reason you need to worry is that I would posit this evolution in technology is already happening. It's on the way, It's inevitable. And Matt, I thought of you when I was when I was thinking this part. It feels like a classic Matt thing man. People are so worried about these proprietary audio surveillance systems. Oh, microphones are everywhere. What kind of world is that gonna be the big hurdle? Oh, I'm sure I'm doing the class. I'm doing your move man, where you like you hold up the phone whenever people talk about you know, I can't be tracked, and then they're like tweeting us and stuff. I had that conversation yesterday with a couple of really nice A D T reps that came out to the house to try and sell me a security system, and we ended up talking for about forty minutes outside on my front stoop. And that was there. One of the big concerns was like government spying and that kind of thing, and I did exactly that, just held up my phone. Was like, you guys were really worried about the government. I said, meta in alphabet already got you, bro? You know what help us mess up the search terms I'll I'm pretty sure I know how the I'm pretty sure I got under the hood of the targeted advertise on social media and not I'm nine seven percent sure that I figured it out, Matt, And it might be an off conversation. Yeah, we gotta figure You and I have to have to, you know, huddle up and decide whether or not it's worth going to air with. But it doesn't involve the dismantled microwave you do, you know one, look, don't try it at home. But it's like, it's kind of cool, right, you would have done it if you were if you were hanging out. This is before we knew each other, you would have been so down. We should have continued anyway, the like. So the surveillance state, dismantled mike waves aside, I think the legal department requires us to tell you not to do it, dismantled mic waves aside the to The only real big hurdles to a constant surveillance state now would be privacy concerns, if you want to put it really simply, privacy concerns, right, the software of the mind and the community and society. That's what we mean with privacy concerns, how could it be normalized? And the second thing would be the physical aspect infrastructure. But as we said, we're holding up our phones. The infrastructure is already there. It is technically easier. I know this. Some people may have problem with this. It is technically easier to deploy a nationwide, pervasive, around the clock surveillance system than it is to have everybody driving electric cars. Because you need you need physical infrastructure. It's already there. It's the easiest part of the equation to solve millions of Americans or US residents, I should say, well, no, North American continent, South American continent millions, millions, millions already carry a phone constantly, and ethics aside, How difficult is it to do the Christopher Nolan Dark Knight Batman trick and just link them all up and have the best echolocation ever, the kind of stuff that would make a bat jealous or a batman. That was awkward, but let's keep it in You like, uh, Jez thinks, well, you know any landing you can walk away from, right shot spotter, build your play. No, that was too far. Alright, Well that's our show. Thank you so much for tuning in our pal Noel will be back soon. I personally can't wait to listen to our strange news listener mail segments. Uh. And we also can't wait for you to be a part of the show, folks, So let us know your thoughts again. One of the reasons this was a two part episode is entirely because we received so much in depth, well thought out correspondence with people who clearly are on different have different perspectives about whether or not shot Spotter is good, whether or not this technology is inevitable, whether or not it's even worth the money, right, and everybody came with fire, you know what I mean? And this this is one of the reasons we do this show. So we can't when we say we can't wait to hear what you think, we mean it. We try to be easy to find online Facebook, Twitter, YouTube. Do we have the MySpace Yeah, no, still know MySpace one day man. One thing. We do have an Instagram conspiracy stuff. We do have a phone number one eight three three st d w y t K. Just call in. It's a voicemail, super easy, three minutes. Give yourself a nickname and let us know if we can use your message and voice on the air. If you don't want to do that, why not instead send us a good old fashioned email. We are conspiracy at i heeart radio dot com. Stuff they don't want you to know. Is a production of I heart Radio. For more podcasts from my heart Radio, visit the i heart Radio app, Apple Podcasts, or wherever you listen to your favorite shows.