ShotSpotter requests to be held in contempt rather than disclose its methodology
chicagoreader.com
chicagoreader.com
When I mentioned it to a family member who is also a police officer, he related that a shotspotter had detected an explosion "right in a guy's backyard." It was clear to me that no device on the market, nor algorithm could figure out that close of a pinpoint. Of course I kept my mouth shut so as to not ruin a holiday. But that was just more undoing of my trust and more indoctrination in one is supposed to remain in a state of denial or cognitive dissonance.
Anyway, just another police technology like polygraphs, DNA, and fingerprints with two stories. One, what it actually does, and two, what police pretend it does.
If sharable, any links to these documents?
I really wanted to include references to the research I did, but I was laid off at short notice (ahem, fired) for unspecified reasons, but it might have had to do with not introducing myself with pronouns. Not that I'm against it, I just felt that my freedom included the freedom to not use them. I was let go on a dime and didn't get to collect my research before I left.
But to briefly summarize how these systems specify their accuracy, it's usually with a range of angles. e.g. +/- 10 degrees horizontal, and +- 15 degrees vertical. I just did a quick google search to try to see if anyone has rolled up the various devices like I did, but nothing. However, here's one decent, non-paywalled paper that shows results under very ideal conditions:
http://nms.lcs.mit.edu/papers/ipsn07-girod.pdf
As far as my words: "no device on the market, nor algorithm could figure out that close of a pinpoint", not your words: "why it doesn't work," consider the accuracy in an empty field in the above paper, but then add in several factors, including: 1. reflections, particularly in a dense urban environment, can causing reverberation 2. humidity and temperature fluctuations, particularly as they vary vertically causing acoustic lensing effects. also temperature fluctuations affecting sensor differently from air. e.g. direct sun heating sensor. 3. extremely high dynamic range of the signal. 4. synchronization of detectors, 5. vandalism of detectors 6. accuracy of layout of detectors, 7. characterization of detector beam. 8. detector distance
This is not to say most of these can't be mitigated and improved, but they cost more money and time to work on all of them at the same time. The highly idealized results (open blank field) of a single shot in the above paper gave results +/- 8m in one direction and +/- 5m in the other direction with an average sensor distance of ~60m. So you're seeing about ~15% error in one direction and 6% in the other under ideal conditions and within a very short range. As sound propagates, the PV field evolves exponentially and it blurs and gets quieter and background noises interfere more.
Note that I mentioend the systems generally specify angles (polar coords), but i used rectilinear here bc I have a diff job and don't have the time to convert just for the sake of this convo, but that would be a step I wish I had time for.
If you want even more inside information from that side of the family, I will let you know that parallel construction with pseudo-private information is very real. This member, again trying to relate to me, brags about using social media to figure out many cases, hinting at having advanced access to e.g. facebook. I would go so far as to say that if you want to get away with crime, police will likely defer their attention to the low-hanging fruit of what they can solve with social media.
for those who asked for papers, here some that I vaguely recognize from my research. Most of these are paywalled, and I used my privileged access from a fancy school to get to them:
https://ieeexplore.ieee.org/abstract/document/5680318
https://tj.uettaxila.edu.pk/index.php/technical-journal/arti...
https://ieeexplore.ieee.org/abstract/document/8635092
https://www.spiedigitallibrary.org/conference-proceedings-of...
https://link.springer.com/chapter/10.1007/978-81-322-1823-4_...
https://ieeexplore.ieee.org/abstract/document/6637700
What's the point of amassing knowledge if you're not gonna use it to change things?
"I looked into that and it's actually this... I still love you dad. You're just wrong about a thing that's all. It's even ok if you still want to think I'm the one who's wrong because hey maybe I am."
Helps people learn that domination and submission, or utter seperation, are not the only ways to have peace and cooperation.
The first system was in Redwood City, CA. I once went to a talk by the designer. They had leased analog audio lines to each microphone. At the microphone end, they had something like a logarithmic amplifier, which traded dynamic range for frequency response. The idea is that you don't want to saturate when it picks up a bang. The audio feeds went to an IBM PC at the police station with some multichannel analog input cards. A program written in LabView running under DOS did the analysis. It was looking for a sharp upward spike from multiple microphones. Then it took the time deltas and computed a position. No attempt to compensate for bounces off buildings, etc. Fairly coarse resolution, down to the side of a block, not a building. It was pretty simple.
Today, there are steerable microphone arrays.[1] That technology was first demonstrated around 2010. Now it's commercially available. Simple ones with a few channels are used for video conferencing. Surveillance microphone arrays that record a few hundred microphone channels individually exist. These allow steering the beam during playback, so the surveillance operator can choose where to listen long after the recording. Needs a lot of bandwidth and storage, but no worse than HD video. Still, the range of that sort of thing for voice is maybe 10 meters. They're installed on ceilings, not outdoors.
When I worked in telecom we had similar problems with reflections in phone lines, which is a highly controlled environment compared to anything that could potentially reflect sound in a city. And even then we weren’t trying to locate something in 3-dimensional space with reflections, just decode information in spite of them. The more I think about it the more this technology sounds like vaporware except in really ideal conditions, like an open field in the middle of farmland or something.
It isn't used as evidence; it's used to manufacture pretext to justify collecting "real" evidence. SpotShotter is not an ignorant party: they understand that the service they sell to police departments is a pretextual laundromat for busting up whoever the police feel did the crime.
That's the crux of this case: SpotShotter is being asked to produce evidence that they are compelled to produce, and they'd rather take the loss in this instance (and continue to use it for parallel construction) elsewhere.
(If I'm being pedantic, "fruit of the poisonous tree" usually refers to illegally obtained evidence, which ShotSpotter is technically not. It's purely a source of investigatory pretext.)
In this case, wouldn't they just keep getting punished for contempt (racking up fines presumably) forever until they fulfil the court's request?
Sounds strategic
I don’t see it as any different than if a police officer is out patrolling, hears a car backfire, and confuses it for a gunshot. If he then detains someone as a part of his investigation and discovers some other crime, he still has probable cause because it wasn’t unreasonable for him to confuse a loud bang for a gunshot and for him to investigate.
The same holds true for shotspotter, regardless of if it makes mistakes. Even if shotspotter has some bug in their acoustic analysis that identifies loud bangs where there isn’t one, ie: total false positives, so long as it sometimes does detect a real gunshot, the police have probable cause to investigate.
The proper analogy is to say that this company is not allowing for the experience and training testimony to be shown to the court. That would be how I would argue against probable cause in this case. However, as another comment in the thread discusses, this is probably not an issue since there is a heavy dose of parallel construction argument going on.
Probably not. Cop said it, court saw it written, that settles it. Now prove your innocence without the ability to face your accuser. They're a cop, of course they're 100% correct in everything they wrote down, so... yeah, pay your ticket.
I've seen this a few more times than I'm the slightest fan of, for me and other people accused of "impossible things" in cars (like "being clocked far faster than the car is capable of having accelerated to based on where it turned onto the road").
The prosecution has no problems with, "But the cop wrote that they saw you doing 70 in a 35, wrote the ticket for 45, and they're not here, so you're guilty."
Now imagine all it did was randomly pick a number from 1 to 100 each day and if it were a 1 or a 2, it "detected" a crime, giving probable cause to police to search people near the box.
Now imagine I almost exclusively put the boxes in areas where black people without a lot of money lived.
You could see this would be something of a problem, right?
Would you put those magic boxes near where previous gun crimes have been committed or would you put those boxes in places based on the race and sexual orientation and disability status and gender of the people living in the city?
First consideration would be applicable legal standard, which practically always errs on the side of the State, second is the reasonableness of statistical methodology presenting the locations based on crime statistics.
I would not personally argue for its constitutionality, but given the precedents stemming from pretextual patrolling cases, it isn’t as clear cut as one might hope.
Logically I would say spending the money on gun control would be better, but politically that seems like a no no. Make carrying in the street illegal with jail time, and use the money in a gun amnesty.
My gut feeling is that shotspotter mostly works and is mostly accurate, regardless of what Vice reported. I can imagine how I’d build such a system and I don’t think it would be very complicated. You don’t need any powerful signal processing or AI, this entire idea was feasible 20 years ago.
And secondly, I think that if any technology is rolled out by a city in a way that’s discriminatory, that falls squarely on the city and their police department, not on the technology company that sold it to them.
The problem is that this specific company seems to be strong on marketing claims and light on engineering legwork. There have also been some implications that point to (in my opinion) post-indictment editing of records to gin up post-hoc probable cause in some cases. No smoking gun (ugh) to prove it but concerning.
What is known is that ShotSpotter reports seem to rarely create useful leads for police. It would hardly be the first (or second, third, fourth, et al) time someone promised to revolutionize policing only to hand over snake oil. Then when the money faucet begins to dry up some companies just can't help fabricating proof of their technology or even falsifying evidence to make sure police "see the value" in continuing to pay for it. Most police departments are being taken for a ride. Some fully understand the wink-wink-nod-nod going on and appreciate the company's "flexibility".
I don't want to base our legal system on your gut feelings. If ShotSpotter is mostly accurate, then they should have no problem releasing the requested material and proving that's the case.
Otherwise I will have to assume it's just the magic crime detection box as described above.
Now part of that might be because of the defacto police strike in response to "defund the police", but the noise went on long enough and was heard in nearby wealthy cities enough that actually motivated police forces were pushing to fix the problem and yet shotspotter proved useless.
This leads to the allegation that shotspotter doesn't actually work at all and is just there so when a hothead young cop does something wrong without thinking the police department can, after the fact, get shotspotter to gin up some probable cause.
If you think it's wrong for police to investigate a shotspotter alert because it might be a false positive then the logical conclusion of that line of thinking is that it should be illegal for the cops to show up to investigate when someone calls in to report gunshots, since those are mostly false positives too.
He is saying they could be selling pure snake oil, and worse using it for discrimination (but not sure what the basis was for that, if any).
Either hearing a gunshot is not probable cause to pull someone over (in which case Shotspotter's accuracy and methodology is irrelevant), or hearing a noise that sounds sufficiently like a gunshot is indeed probable cause, in which case you can just listen to the audio yourself and make a determination about whether or not there was a true gunshot-like noise or the company invented a fake alert out of thin air. I don't see any scenario in which the methodology or accuracy of Shotspotter's devices is at all relevant to this case. The relevant fact here is whether this particular bit of audio can be reasonably classified as a gunshot, not the broader accuracy of the system over the whole data set.
>>Attorneys from the Cook County Public Defender Office, contending the police pulled the man over solely based on a ShotSpotter alert, subpoenaed the company as a third party to find out whether that was the case.
but that seems unlikely -- it would be really expensive and low-yield to pull over everyone in the vicinity of a gunshot. It would be about the same as with regular human ears hearing a bang: go to the source, then pull over cars with evidence further differentiating them as a possible locus of crime.
So probably shotspotter got the cop to that location, where he saw other, more crime-specific evidence. But without more info about this case I couldn't say for sure.
Expensive and low-yield for whom? For the officer who's been directed their by the robot, they're incentivized to do so; it makes them look like they're doing their jobs instead of doing nothing. What's built into the system to penalize them for inconveniencing / endangering a bunch of people with nothing to show for it?
And if you’re economizing on time, I can assure that a better use of it would be to answer the challenge you kept dodging (or “misinterpreting” as a opportunity to show off irrelevant knowledge):
https://news.ycombinator.com/item?id=32259971
At this point I’m just going to assume you have no idea if 7% is a good threshold or not, but still see so problem with shouting it because it sounds good.
[1] In case you forgot, the only reason you were outraged was because the search did catch a criminal that it shouldn’t have. So obviously someone cares about catching a criminal!
It's expensive for the community, to be sure, but the cop's not incentivized immediately to care about that.
It wasn't in that comment, it was in the premise of all your previous posts that it is outrag^W^W^W^W^W suboptimal for someone to be pulled over and eventually arrested for DUI when the original instigating cause was a false positive for gunfire.
IOW, the entire reason you're talking about this case, is because they did not pull over everyone in the vicinity but prioritized those that were more likely to be committing crime, based on other signals. That's how we got to this case to begin with! And yet, I'm having to remind you that you're arguing with yourself.
Do you consider it too much to ask that you remember the context of what you're talking about? And for others not have to remind you?
Also, thanks for the confirmation that you don't have a substantive answer to the substantive question, "what probability threshold is appropriate for police to enter in preparation for an active shooter" -- I had always assumed your comment was bad-faith to begin with. Especially when you deny that comments such as these are expressing something similar to outrage:
> In one case, a 65-year-old grandfather was held in Cook County Jail for nearly a year at the height of the COVID-19 pandemic before his case was dismissed due to insufficient evidence.
I mean, WTF.
https://www.chicagotribune.com/news/criminal-justice/ct-shot...
Prosecutors accuse him of shooting his passenger, but he claims that another car pulled up next to him and shot his passenger instead. In either case shotspotter both correctly identified a gunshot and the location of the gunshot, so what exactly did they do wrong here? By the defendant's own admission the gunshot that triggered the alert was real! Unless crucial details are being omitted somewhere it seems like a hatchet job to me.
For background: https://iaals.du.edu/sites/default/files/documents/publicati...
Honestly, given their history, they probably just don't want to reveal that their entire business is snake oil, basically existing solely to make money by enabling law enforcement to create crimes out of thin air.
If their science was sound they wouldn't have to resort to such tactics...
Look how accurate raspberry pi powered ADSB plane tracking networks are — and that’s a much harder problem: they have no funding, they’re doing triangulation at the speed of light, and they’re doing it without GPS clock sources. You could also point to the lightning tracker networks that work in roughly the same way.
Given a few million dollars to outfit their nodes with good GPS units and good microphone arrays, I’d expect their technology to be very good.
https://www.flightradar24.com/blog/how-we-track-flights-with...
Sometimes my HN replies sound more like compiler errors.
- https://igchicago.org/2021/08/24/oig-finds-that-shotspotter-...
"Police watchdogs in Chicago have issued a report about the city’s use of ShotSpotter’s acoustic gunshot detection system, finding that only 9 percent of ShotSpotter alerts led to evidence of gun crime."
- https://www.vice.com/en/article/5db8vk/police-are-using-guns...
"Court documents recently reviewed by VICE have revealed that ShotSpotter, a company that makes and sells audio gunshot detection to cities and police departments, may not be as accurate or reliable as the company claims. In fact, the documents reveal that employees at ShotSpotter may be altering alerts generated by the technology in order to justify arrests and buttress prosecutors’ cases."
"And according to the testimony of a ShotSpotter employee and expert witness in court documents reviewed by VICE, claims about the accuracy of the classification come from the marketing department of the company—not from engineers."
- https://www.eff.org/deeplinks/2021/07/its-time-police-stop-u...
I'm skeptical of this company in particular. See this other comment: https://news.ycombinator.com/item?id=32259253
Edit: I think a novel implementation here is that they often use one mic array (several mics mounted around one tower or vehicle) as opposed to a dispersed network at them. If they made that work effectively, I think IMO it's different enough as to be novel and patent worthy. But it also strikes me as the kind of technology that should be defensible on its own merits, and legitimate copycats easy enough to identify, rather than blanket squelching tactics and going after journalists. That's what snake oil salesman do, not companies with actually superior technology. Read the Vice reports. Their effectiveness record isn't great except in their first party sponsored research.
Bad part of oakland? Plenty of random gun shots happening, but no idea where exactly they are coming from - really it’s random spots.
But ShotSpotter can give you a high precision (really random) fix on someones house! Wait until it’s someone you want to search, and voila you can.
If the cops think they’re committing a crime, they should actually find enough probably cause to get actual paperwork, not just create junk to justify going after the ‘usual suspects’.
*Not Legal Advise*
I frequently misunderstand how features work in common applications, especially nuances of models and processes; so I can only imagine the difficulty gaining a precise understanding and context for a non technical person; e.g. a lawyer or jury.
Unrelated: as someone not from Chicago, it's surprising that a company exists solely to detect gunshots. Depressing.
(1) Their methodology is about to be debunked, or;
(2) It's about to be revealed that they were working with police to allow them to manufacture reasons to arrest people.
I am not versed in Illinois contempt law, but judges have very broad constitutional authority to sanction people or companies for intentional failure to comply with a lawful order. In this case, you could expect the judge to do something like order the company to pay five thousand dollars per day until they disclose the records.
There could be an argument that the judge could legally order the sheriff to seize their servers or throw the CEO in jail until the company complies, but for some reason stuff like that doesn't seem to ever happen in the corporate context.
It's a thing
https://cdn.ymaws.com/www.idc.law/resource/resmgr/quarterly_...
> This procedure is often called “friendly contempt,” a recognition that the contemnor has not acted “with contumacious disregard for the court’s authority,” but requires a contempt finding and sanction for the purpose of obtaining appellate review. See Klaine v. S. Ill. Hosp. Servs., 2016 IL 118217, ¶ 6; Zagorski v. Allstate Ins. Co., 2016 IL App (5th) 140056, ¶¶ 16-17. In many cases, the contemnor has even requested the finding and sanction for that purpose. See Ctr. Partners, Ltd. v. Growth Head GP, LLC, 2012 IL 113107, ¶ 20.
So this is the lawyer asking to appeal the original judge's order of disclosure, and this is the obscure process that makes that happen.
Maybe a night in jail isn't all that common though. Sure seems reasonable and effective.
I can't think of any off the top of my head, but I'm sure there are a few where the evil journalist or priest is held in contempt hindering finding the bad guy.
3) They don't want their "secret sauce" made publicly available.
That would both open them up to commercial competition, and make their product less reliable in court, as anti-Shotspotter experts would start figuring out how to effectively argue against the product's methodology.
In general I think technology like this should have to be fully transparent if it is used as the basis for public entities' decisions, but so far the arguments against that practice seem to have won out. For example, good luck getting source code or design documents to contest an automated red light or speeding ticket.
I sincerely doubt there's any secret sauce to protect when they have "analysts" sitting around regularly "correcting" evidence.
In general though, I think there are legitimate reasons for manual analysis. Any machine learning approach to a problem like this is going to have to balance false positives and false negatives, and there is necessarily going to be a somewhat arbitrary cutoff. Detecting bang-like noises in a large city is likely going to have a pretty conservative cutoff to avoid DDOSing the police with calls. For example say the company sets their cutoff for automatic reporting at 90% confidence, but when the police ask them to review a specific time period, it turns out there was a shot detected with 89% confidence, and manual analysis indicates a false negative. That is probably still useful information, and it would definitely be useful to include this data point with the correct classification in future iterations of the model.
https://en.m.wikipedia.org/wiki/Independent_component_analys...
Plus basic triangulation logic synced across a bunch of microphones?
Heck, a truly calibrated surround sound system requires using a microphone to measure the volume in place.
Now that's a lot of work.
And how often would it have to be repeated?
I doubt they've solved (or even have non trivial mitigations for) any of those subproblems, based on stories of how often they apply human "corrections". The portion they've implemented may be very easy to replicate.
You can’t make rulers while also trying to keep the definition of an inch secret.
I don't really care about a red light camera (as the image shows the license plate and the red light at the same time), but there are tons of other technical solutions that need to be examined.
The actual technology is not all that complex. Transient detection and triangulation by decorrelation from a microphone (edit: not microwave!) array. You could probably prototype it in a week if you leaned on an established FFT library. Obviously they would not want to reveal their source code and so on, but if you know digital audio you know this isn't that hard.
I think what they really want to conceal is not the technology but its use in operation. I don't know what Chicago pays for the service, but Sacramento CA spends about $1m a year on it and Sac is a fairly small city. https://www.sacbee.com/news/local/article231997777.html
Wouldn't "red teaming" ShotSpotter solve for this to see how accurate they are? A team would perform controlled firings in known locations under various auditory scenes with the city's approval, and then request the data from ShotSpotter to see how accurate they are (without revealing you are performing an exercise). ShotSpotter then doesn't have to reveal their methodology, as their data provided from a blind test will reveal the accuracy regardless of how they arrive at conclusions. The system is either accurate or it isn't (with a measured level of confidence), and this can be tested for without ShotSpotter's consent (as long as they continue to provide data to the customer per their contract).
A bigger problem is that this doesnt approach any of the process problems with the human moderated parts of the system. For that you would need cooperation from the law enforcement customer. The customer would need to make a request like "are you sure you didnt find any gun shots around this time and place?" and see how the company responds.
I think the most reasonable inference is that they are outright trying to hide. And probably because the truth is totally damning.
If their methodology has weaknesses which allow those arguments, then its kind of a travesty of justice not to allow them.
Waving a scientific magic wand from a crony-connected company is a great asset to would-be abusers in the justice system.
I was on a jury where we were forced to acquit a guy of vehicular manslaughter because a vendor representative couldn’t properly explain how NTP works on a red light camera under cross examination. It was gross.
It's pretty different from "Let's just search everyone's houses. Probably someone is committing a crime."
Promotional material issued by ATSC claimed that the ADE 651 could detect such item as guns, ammunition, drugs, truffles, human bodies, contraband ivory and bank notes at distances of up to 1 kilometre (0.62 mi), underground, through walls, underwater or even from aircraft at an altitude of up to 5 kilometres (3.1 mi).[12] In a promotional video, McCormick claimed that the device could detect elephants from 48 kilometres (30 mi) away.[13]In so far as I can tell, it seems to be doing its job, though I understand there are several institutions here which obviously are not trustworthy.
https://www.vice.com/en/topic/shotspotter
https://www.techdirt.com/2022/07/08/judge-tosses-defamation-...
Also, the devices don't "enforce crimes", they only alert approximate areas where a gunshot/firework may have gone off at.
That's the sort of thing that would be useful in court if it's true.
Are local police required to publish where they are deploying ShotSpotters? Shootings and racial demographic data is certainly already public.
I've read the entire Vice court report and don't remember seeing that data. Haven't read the Chicago one.
Edit: reading the original Motherboard source all of the examples were described as small 'pilot programs' so they were usually only a single part of each city, in Cleveland for example :
> the Cleveland Police Foundation—has been deployed in a small section of Southeast Cleveland that city officials have described as a high-crime area.
> Those neighborhoods also happen to be among the least white in the city and those hit hardest by economic recession and the pandemic, according to local community organizers.
> The area the division chose for Shotspotter accounted for 37 percent of all shots fired calls, 37 percent of felonious assaults, and 45% of all homicides within district four, Sergeant Jennifer Ciaccia wrote to Motherboard in an email.
https://www.vice.com/en/article/88nd3z/gunshot-detecting-tec...
(Motherboard didn't provide crime stats in the other 3 example cities cited, only the demographic profile which was consistently non-white)
For example, perhaps literacy tests in order to vote in the Jim Crow south were purely based out of concern that only informed voters should be able to vote. Assuming they were applied evenly to both whites and minorities (they weren't), one could draw the conclusion that even though most of the people prohibited from voting were minorities, the intent wasn't racist.
But that's not how most racism works.
Which, in the sum of everything, ultimately out-ranks the value of attempting to address the homicides and gun violence that plagues their communities through law enforcement. Therefore it's a net negative for society?
(Apologies if I'm being overly literal here, my brain tends to seek the most rational explanations for everything).
Edit: Also to be clear I agree that the vast majority of people aren't these actors. The problem is it is very easy for someone with ill intent to get things moving in the direction they want. Very few systems stand up to real world abuse by a sufficiently motivated bad actor.
The reason violent crime enforcement disproportionately affects black people in the United States is because black people disproportionately commit violent crime. Black people, for example, commit most of the murders in the United States despite not being most of the population.
It is not racist to notice and react to crime. On the contrary, it is racist to insist that law abiding black people, the vast majority, endure high rates of crime. Reducing and hobbling law enforcement necessitates that citizens who might be protected are not.
That's without getting into the socio-economic factors that lead people to pursue sports as a career rather than playing in school and then going on to be an architect or whatever it is.
The rest of your comments have been dissected and examined to such an extreme over the last like... 80 years... that I don't have anything new to add to the discussion.
The issue is that black people are overrepresented in the NBA. This is a "racist outcome" per the language of an ancestor comment. Is it because the league is prejudiced against white people and keeping them out? Prejudiced against black people and expressing the prejudice, for some reason, by admitting more black players to the league? Or, could it possibly be that black people are just more likely to be good at basketball? Black people are ~15% of the US population and ~75% of the NBA. Must be an awful lot of racism there...
I provided an example of basically the blackest city in America deploying it within their own city. Why did they deploy a tool of oppression on themselves?
It seems you were very distracted that I pointed out that whites are a minority within Jackson MS, but that was really just an offhand comment and not the point of the post.
The OP said nothing like this. You're projecting a lot of personal biases onto their comment. But you know that, and are trying to hone in a single example in order to strawman/red-herring valid concerns about the technology and its effects.
> It seems you were very distracted that I pointed out that whites are a minority within Jackson MS, but that was really just an offhand comment and not the point of the post.
That was the whole point of your post. You literally said nothing else.
Yes, I will admit I am projecting, or rather assuming, bad-faith on their part, or at least sneaky rhetorical tactics. I used to assume good faith, but the past 5 years or so have made me realize that very few people operate this way. If I see someone making a point, and an underspecified part of their post or a part where they use vague language may give a certain impression, and it just so happens that that certain impression furthers the point they were trying to make(rather than diminish it, or be orthogonal to the point), then I assume there was intent to do such a thing.
OP said:
> It doesn't work all that well and is selectively deployed in minority areas and is used by police to selectively enforce crimes, sometimes also altering the evidence.
Yes, I suppose it sometimes may be selectively deployed in minority areas in order to oppress them, but without any kind of mention of the rate that that happens, this appears to be OP trying to convince you that it may be worse than it is. If you didn't know anything about cars, and I said that cars are used to intentionally murder people, you would probably have a worse perception of cars than if you had a more thorough understanding of their usage.
Therefore, my point was to show that shot spotter is not just used to oppress minorities, I provided an example of a predominantly black city choosing to use the technology, which almost certainly means there are uses of the tech that aren't about oppressing minorities.
> That was the whole point of your post. You had nothing else substantive to say.
I'm sorry, but I understand my point better than you do. I apologize for not being more explicit earlier. Hope my post helped.
https://www.vice.com/en/article/88nd3z/gunshot-detecting-tec...
https://www.vice.com/en/topic/shotspotter
https://www.vice.com/en/article/qj8xbq/police-are-telling-sh...
https://igchicago.org/2021/08/24/oig-finds-that-shotspotter-...
https://www.eff.org/deeplinks/2021/07/its-time-police-stop-u...
https://www.techdirt.com/2022/07/08/judge-tosses-defamation-...
No, it would be silly to claim it's being deployed only in those areas.
Surely there is some merit there that is not merely keeping minorities down, or maybe all the black people running the government just secretly hate blacks.
Even if the response that used it differently was meant to make a valid point (I believe it was, as they expanded on it later), the way they used minority did nothing to help that point and instead caused unnecessary contention.
It would have been far more effective for their point to say "Jackson MS has a population and police department that are both majority black, and they still choose to deploy shotspotter in black neighborhoods.". Then we could jump right to exploring reasons why that might be, including valid uses of shotspotter because of effectivenessas well as other incentives regardless of it's effectiveness.
1) Jackson MS is 80+% black.
2) The (mayor + city council) of Jackson MS is 75% black
3) The command staff of Jackson MS police department is 100% black.
4) The city and police department decided to deploy shot spotter in Jackson
5) (don't get distracted by this please) The city has not deployed shot spotter in areas that are predominantly white
Given those 5 points, where is Jackson deploying their shot spotter tech? Is it in predominantly black/minority neighborhoods? Yes.
Why are they doing that? Does the predominantly black government intend to keep black people down?
This may be non obvious but for example black cops tend to use disproportionate force against black suspects in areas where such behavior is culturally and systemically tolerated. Cops and suspects are in an adversarial situation and being part of team cop or team suspect is more important than sharing a skin color.
An infallible argument; the oppressee can just as well be opressor, because otherwise we'd have to reconsider our preferred conjecture.
The alternative is like a mirror of the argument that a fellow isn't racist because he has a black friend. It's parody masquerading as analysis.
Also it's my experience that if an all black community decides to enact public policy that goes against the narrative, it's almost exclusively young white progressives that come up with these sorts of agency-removing arguments to protect their preferred worldview
The topic of discussion is a company that produces a product that due to echos and analysis doesn't actually know where or if guns are fired being deployed in shitty crime ridden neighborhoods where disproportionately minorities being poorer are obliged to live. In the context of the gun shot imaginer system they have been literally caught doctoring data to match what the police wanted to be true in order to obtain a conviction.
Multiple things can be true. The shot spotter can be deployed in areas where gun fire actually and truly is more common, it can be misused to shake down or frame people in that area, such behavior can primarily be misused against people in that area, because more poor black people are obliged to live their they can in turn fall victim to such bad behavior more frequently even if not specifically targeted by law enforcement. Further faced with multi decade scare tactics and shitty public defenders poor people and by extension disproportionately minorities flinch and cop pleas.
This is how you fill your prisons up with black people.
This is literally what systemic racism means. Its a racist outcome of a fucked up system even when said system isn't consciously malicious and lets be a little honest with each other. We both know that there is plenty of honest bigotry. There are plenty of folks in government and plenty of folks in uniform that wouldn't mind donning a sheet and burning a cross in their neighbors lawn.
America was born in racism and as much as we hope to generation by generation make it more equal some of us remain more equal than others and a lot of us like it that way.
>In Chicago, a government watchdog says Shotspotter alerts have only led to evidence of gun crime in 9 percent of cases.
>fewer than 2 in 10 investigatory stops following ShotSpotter alerts resulted in the recovery of a gun
9% actually seems like a lot? Considering that most criminals don't stick around after illegally discharging a firearm. And 1 in 5 resulting in the recovery in a gun? Vice is really bending over backwards to pretend that these numbers are bad.
If you're in a particularly crime ridden area, my concern is how many of those 9% are the right evidence against the right person. It's pretty terrible if all it took to nab the wrong person was shoddy evidence from a shady and secretive company that is accused of having manipulated the evidence in the past. It is still quite terrible even if it's often correct, because that makes it much harder for someone who is innocent to do anything about it.
Still, it's possible that the numbers Vice is showing does not demonstrate the problem.
I would not want to live in a community so policed.
Maybe we should focus on why there's so many gunshots in these specific communities.
The gun violence is just the camouflage that provides plausible deniability. If ShotSpotter had been founded in London they'd have called it "StabSpotter."
I also hear kids shooting fireworks, and those modifications people do to their motorcycles and cars to make them sound like they're backfiring (even if they aren't; I don't know how these things work).
And while I like the concept behind ShotSpotter, the company, it's methods and its data governance are... Shoddy. That's probably the most charitable I can put it from what I've read about them.
Another thing I didn't point out in my comment is that ShotSpotter also changes its data if police tell them to. Which seems like either an admission that their sensor is not very good, or that they're willing to compromise good data when told to.
This is no different than other technologies. DNA testing wouldn't be admissible in legal actions either if the tests were essentially black boxes that couldn't be scrutinized.
You have an interesting concept of the rule of law.
The average police response time in Chicago is 5-6 minutes[0]. I'm going to bet that active shootings are on the fast side of the average. Having lived in Chicago for 7 years, I can tell you there are areas where you regularly see police patrolling.
0. https://mediaprogram.maps.arcgis.com/apps/Minimalist/index.h...
In any case, this is all supposition. We aren’t being allowed to review the actual evidence, so for all I know SpotShotter is just telling the police what they’re asking it to.
Eg https://chicago.suntimes.com/crime/2022/7/27/23280415/chicag...
Yes, I do think that.
https://www.npr.org/2022/05/26/1101496374/chicago-police-13-...
It seems to me that to think otherwise is a "the omelet requires breaking eggs" argument.
That said, I don't think anyone is opposed to technology that can locate gunshots and deliver law enforcement to the right places. The gripe is that this company in particular has a history of being opaque and trying to cover up their failings, leading to questions about the effectiveness of their underlying technology and data integrity. The public can't tell whether this is a legitimate social good based on an effective tool or if the company is just ripping off tax payers and sometimes outright working with crooked cops to fabricate probable cause.
Not every cop is crooked, and not every acoustic triangulation technology is fake. But this company isn't being transparent or well intentioned enough for the public, or even a limited group of outside experts, to make that determination. And they have a history of being willing to corrupt their data, altering evidence to facilitate false records from crooked cops.
The right tools in the wrong hands can still be a problem. We don't know if this is the right tool, but the company has shown itself to be the wrong hands.
https://news.ycombinator.com/item?id=32259925
But yes, "tighten up" those Fermi estimates for a question no one asked and which has no bearing on the discussion.
No, I wouldn't, because your facts are wrong. Yes, London police are generally unarmed, but they still get called to respond to situations where there is a likely armed suspect, and in those cases, they send in special units that are armed.
The question here -- that you avoided -- is, what is the appropriate probability before sending in such an officer, and is it higher than the 7% in the situation you cited? You were asking others to be outraged that there was "only" a 7% chance of an actual armed shooter. In an attempt to move the discussion forward, I asked what the appropriate value is, so as to appropriately judge whether such intervention was unwarranted.
To which your entire response was to shoehorn in an irrelevant factoid about London and falsely imply I'm not aware of it. (And, of course, work in a Fermi reference, because, why not?)
If I may make a suggestion: next time someone asks a substantive question, and you don't have a substantive answer, just don't reply. That's a much better approach for the forum than what you just did.
In that context, ShotSpotter's mechanism bordering on ascientific prognostication should be considered dangerous to the community because it is increasing the chances of fatal encounters for no benefit. It's basically SWATting with a fancy corporate branding.
What part of my comment is that (or anything else you've said) responding to? I asked a simple question: what is the appropriate "probability of a shooter" that justifies going in with a protocol that expects one? Now you're just guessing things that have little to do with what I said.
Are you under the impression that the only evidence of a firearm being discharged is carried away by the person firing the gun? I was under the impression that evidence of a gunshot is kind of famously objective because of the presence of bullets, bullet holes, casings, victims, property destruction etc.
By your reasoning, it’s a feasible proposition that criminals regularly patch up bullet holes and what, fireman carry their victims away from the scene?
Find the bullet hole, and you've found the property destruction, and the bullet is either embedded in the property destruction or fragmented all of over the place, as bullets often do. So you could have just said "property damage."
But tell you what, I'll go downtown and shoot a black paintball in a random direction, and tell you roughly where I was when I shot it, within 100m. It will be your job to find where that paintball hit. Good luck finding your evidence.
>casings
Casings are easier to find, but still very difficult if you don't have the precise location of where the shooting took place.
>victims
Firearm discharges don't always result in victims, or even victims willing to come forward.
What are you honestly expecting cops to do when there's a gunshot report? Spend hours combing the city for some rogue bullet holes? It's not realistic. The best evidence you're going to get is rushing to the area where shots were reported and finding someone fleeing.
That’s a fascinating position! Based on how difficult it is to find any evidence whatsoever, would you conclude that gunshots are essentially inconsequential 91% of the time? If I believed that, I would advocate for the legalization of recreational firearm discharge in populated areas, since obviously they rarely actually cause any noticeable issue.
Edit:
A tech company going “Beep boop loud noise! Give us more public funds for our technology that we refuse to explain!” is not the same thing as an actual report of gunfire to me. As for what I expect the police to do? Nothing. Literally nothing at all, anywhere at any time. They’re not obligated to, and in my experience, that’s what they usually do. Nothing.
I forgot to add “witnesses and complainants” as something that could be considered evidence of gunfire, but I’m assuming they can also be carried away
It's the same problem that Stop and Frisk[1] in NYC had: you can make the contraband recovery numbers look fantastic, if you're willing to sweep peoples' civil rights under the rug.
[1]: https://en.wikipedia.org/wiki/Stop-and-frisk_in_New_York_Cit...
But some minority demographics commit disproportionate (edit: violent) crime per capita so you expect some proportional scaling (more crime = more need for crime detection) if they happen to live in cities/areas where that's the case. (Usually big cities)
I'm surprised to see this is somewhat true in the US. In Canada rural crime rates are higher than urban ones, but that trend doesn't seem to hold for the US (per sources below)
Of course extrapolating from "average crime rate" to "location of hotspots" might be suspect in both cases...
Stats Canada: https://www150.statcan.gc.ca/n1/pub/85-002-x/2019001/article...
Bloomberg article with some nice graphs, and really good links to sources (at the end of it) - the text focuses on NYC but the second line chart in particular (based off of linked CDC data) does a good job of answering this question: https://www.bloomberg.com/opinion/articles/2022-06-07/is-new...
For example, it's well documented wage theft is a much bigger crime in terms of monetary value than all forms of robberies combined. But those criminals don't get nearly the same level of enforcement as robbers do.
Other Spotters for things such as wage theft, fraud, etc would be great.
You could (and I would) argue some whole industries are heavily criminal (fraud) and bait people into moral hazard.
I don’t think police is deploying this based on skin color
Police said ShotSpotter, a surveillance system that uses hidden microphone sensors to detect the sound and location of gunshots, generated an alert for that time and place.
Except that’s not entirely true, according to recent court filings.
That night, 19 ShotSpotter sensors detected a percussive sound at 11:46 p.m. and determined the location to be 5700 South Lake Shore Drive—a mile away from the site where prosecutors say Williams committed the murder, according to a motion filed by Williams’ public defender. The company’s algorithms initially classified the sound as a firework. ...
But after the 11:46 p.m. alert came in, a ShotSpotter analyst manually overrode the algorithms and “reclassified” the sound as a gunshot. Then, months later and after “post-processing,” another ShotSpotter analyst changed the alert’s coordinates to a location on South Stony Island Drive near where Williams’ car was seen on camera.
In 2016, Rochester, New York, police looking for a suspicious vehicle stopped the wrong car and shot the passenger, Silvon Simmons, in the back three times. They charged him with firing first at officers.
The only evidence against Simmons came from ShotSpotter. Initially, the company’s sensors didn’t detect any gunshots, and the algorithms ruled that the sounds came from helicopter rotors. After Rochester police contacted ShotSpotter, an analyst ruled that there had been four gunshots—the number of times police fired at Simmons, missing once.
Paul Greene, ShotSpotter’s expert witness and an employee of the company, testified at Simmons’ trial that “subsequently he was asked by the Rochester Police Department to essentially search and see if there were more shots fired than ShotSpotter picked up,” according to a civil lawsuit Simmons has filed against the city and the company. Greene found a fifth shot, despite there being no physical evidence at the scene that Simmons had fired. Rochester police had also refused his multiple requests for them to test his hands and clothing for gunshot residue.
Curiously, the ShotSpotter audio files that were the only evidence of the phantom fifth shot have disappeared.
Both the company and the Rochester Police Department “lost, deleted and/or destroyed the spool and/or other information containing sounds pertaining to the officer-involved shooting,” according to Simmons’ civil suit
Of course that seems like a small nitpick given the content of the passage I copied from the article.
If it was up in the atmosphere with long consistent pockets of air, maybe. On the ground, no way.
Even your ears can somewhat locate sound sources in 3D space, just not with perfect accuracy, especially in complex soundscapes like cities.
Assumption: the majority of commenters are working in technology fields, and perhaps even technology startups.
Let’s say that you have developed some software product that works pretty well for certain applications, but is certainly not perfect. But it works well enough that your customers are paying you to use your product.
Now, your customers decide to use your product in a way that you have neither intended nor prescribed. (In fact, your product comes with a disclaimer.)
In doing so, they end up in court. And you get hauled into the middle of it.
Should you have to expose your technology? This could endanger your entire business.
https://www.vice.com/en/article/qj8xbq/police-are-telling-sh...
https://www.vice.com/en/article/5db8vk/police-are-using-guns...
https://www.vice.com/en/article/88nd3z/gunshot-detecting-tec...
This sounds unfair to you, but sometimes the fair solution creates more injustice than it solves. This is the same reason why we have the EPA Superfund program. Love Canal might not have been the polluter's fault[2] but the "fair" alternative was to let everyone sit in court while they and their kids die from toxic waste.
[0] tbh the biggest problem with Chinese IP theft is that they don't share their plunder
[1] At least I'm assuming "keeping the law secret" is something Mao did, I might just be being an ignorant American
[2] If I remember correctly New York State or the county government actually threatened to use eminent domain to force the polluter to sell the toxic waste dump to them.
Believe it or not, this is not a unique problem for technology fields. Courts have long understood that techniques are possibly sensitive, and allow for a broad variety of discovery techniques to accommodate the defendant's rights. A common one for software is to have an independent, court-overseen third party review the code and produce an affidavit affirming that the code does what it claims. Another is to only allow review of techniques in a controlled environment, where the defense cannot steal intellectual property.
If ShotSpotter is able to figure it out correctly only then they are are allowed to NOT disclose their tech otherwise it would be very illegal to keep it undisclosed because people can literally get arrested on the wrong leads generated by a piece of tech.
What a fucking awesome new legal principle invention that I am sure everyone will get to enjoy equally!
Ideally, I'd like to replace police dogs to give police probable cause. Policing is a huge industry in America, and it's not realistically going to be reformed because there is also a lot of crime in America, so it's a lasting business.
I'm mildly worried someone else will get there first. But we'll see. I have a couple of ideas.
0: Though if you could run a large-scale single source of training with some sort of certification program, that would work. Dogs are a fantastic laundering device, but I don't like that it linearly scales cost to add more. You can force recurring revenue by getting "periodic retraining" into the process, which will play really well with everyone.
Whatever, the specifics of the product design are irrelevant details. It's about the bizdev. They still use polygraphs in the Federal Government. The device itself doesn't have to do anything.
[0]: https://www.usgs.gov/media/images/triangulation-locate-earth...
God this phrasing gets so annoying. If white neighborhood residents were being murdered in their streets, you really don’t think they’d put these devices there? Residents would be itching for them!
It’s about catching murders - putting these devices in places with the most gun violence.
Having police catch murderers is something we all want - rich neighborhoods included - and something that takes resources and effort to do. It’s something that we are privledged to have, compared to many other places in the world that just let poor people kill each other without legal recourse.
Good people in these neighborhoods want policing. I have a friend in Humbolt park, and he practically has PTSD from constant gunfire. You think he doesn’t want these devices?
Bad correlations ad nauseam.
The population is 6% Black. The homicide victims are 50% Black. We have twice the homicide count of Seattle, or more homicides than Seattle and SF combined. Portland is smaller than both cities, so it is very concerning.
Unfortunately the licensing model for ShotSpotter is such that it is only deployed in the high homicide neighborhoods due to cost. I have to imagine the marginal cost of an extra sq mile is low.
I’d like to see a diversity discount or something so that it would be economical to put the system everywhere and dispel the racial bias issue. Or maybe charge per homicide or gun recovered instead.
It's also possible for good-faith efforts to play into a feedback system that ends up resulting in racial injustices. They can even do that while doing other good things too.
Then you have the uncomfortable and terrifically subjective question on your hands of weighing the costs and the benefits - both now and into the future.
I don't think you can dismiss this as being largely about race, nor dismiss it as irrelevant to race.
In some ways, it'd be easiest if it doesn't work. That way the questions get real simple.
Don't Chinese people want the social credit system too?
You clearly glossed over the more important factor "is used by police to selectively enforce crimes, sometimes also altering the evidence". So, no, your "buddy in Humbolt park" doesn't want this when he gets accused of a crime he didn't commit. He wants actual systematic changes to an impoverished areas - black or white.
> If white neighborhood residents were being murdered in their streets
Hilarious, because this statement displays your inherent bias, which is that you automatically presume white neighborhoods are rich and therefore don't have problems with residents shooting each other!
During the "war on drugs" black men were getting snatched left and right, destroying communities. Now the fruits of that come back to create violence, and your answer is to push an expensive heavily flawed automated police lead generator?
Shotspotter is broken. It constantly results in false alarms and false stops to go with them, it's not effective enough to be reproducible, they openly admit to letting police override their classifications...
It's an expensive farce that isn't helping anyone.
If there were similar amounts of homicide in white areas, white residents would be using their political clout to demand solutions, not pretextual pseudoscience. And this is, in fact, what we see: white neighborhoods get beat cops who live nearby, while black neighborhoods get surveillance and patrol cars with humps from the next county over.
Have a little self-awareness.
https://news.ycombinator.com/newsguidelines.html
We detached this subthread from https://news.ycombinator.com/item?id=32259253.