Santa Cruz, California bans predictive policing in U.S. first • Hacker News Board

Re: Santa Cruz, California bans predictive policing in U.S. first

#361

Post by wigl » Fri, Jun 26, 2020, 10:00 PM UTC

ericjang wrote: post #260

wigl wrote: post #148

Earlier quoted context omitted.

Grew up in Harlem though not black. While I'm sure there are good police out there, broken windows policies effectively act as a safe haven for a subset of cops to enforce arbitrary (usually racial) bias whenever they feel like it. There's no impetus to actually do the neighborhood policing they claim aspire to--just a universal excuse to allocate resources and attention in the name of making neighborhoods "safe". Me…

This is horrific. When did this happen? Was this in Harlem?

9-10 years ago. Spanish Harlem during peak stop and frisk era/doctrine.

What I remember most was how routine it all was for everyone involved. Driver would sometimes pull the car up on kids drug raid style, swerving into the sidewalk. They'd turn on the high-beams and within a few words and seconds, it was a pop-up police lineup.

The only description of the suspects was this: a group of kids, one of them was wearing a bright hoodie. No one had even seen their faces. I was adamant about this as well but these cops didn't believe me. They instead kept asking me to "be real" with them and that I shouldn't "feel the need to be PC" around them. They acted as if I was self-censoring for virtue signalling instead of being honest. I just wanted them to log my IMEI and keep a lookout for my phone on CL/eBay because they had been doing press releases about it at the time. Apparently that was the less accessible option.

It scares me to think how much negative impact that one group of police might've had. When I think about how frustrated they became when I told them the truth over and over. Maybe they're used to more central park Karen type of crime reports.

Re: Santa Cruz, California bans predictive policing in U.S. first

#362

Post by lordnacho » Fri, Jun 26, 2020, 10:01 PM UTC

bhntr3 wrote: post #335

nwienert wrote: post #291

Earlier quoted context omitted.

Yea, both of these points are so bad as to be almost parody. Not only should you never, ever base your beliefs based on avoiding the discomfort of being an outsider, but you also should never, ever, use majority opinion as some sort of proof of... anything. If you did, you’d have advocated for slavery if only you’d been born south of the Mason Dixon years ago.

Here's the wikipedia article for the fallacious argument I used and which you're critiquing: https://en.wikipedia.org/wiki/Argumentum_ad_populum Here is some statistical evidence for systemic racism: https://www.businessinsider.com/us-systemic-racism-in-charts... (Shockingly this is actually pretty good despite being from Business Insider) Honestly, I'd just edit the link at this point if I could since it seems to ha…

With all due respect, I'm not sure the BI graphs are great support for the thesis. All of the economic ones are essentially the same: if you're poor, you have less savings, are less likely to be employed, less likely to be CEO, less likely to have health insurance, and so on. They're expected to be correlated, so in some sense they are not as informative as a whole as one might guess from the large amount of them. It's also not clear that this has anything to do with racism, economists would point to a whole load of confounders which mags like BI tend to skip over.

The one I found interesting was marijuana, where at least we have two measurements of the same thing, smoking up and getting caught with the stuff. That could be expanded on, as there's a thesis out there that police focus their resources unreasonably on black people.

A really good study that investigated this would be something like the essays in Steven Levitt's books. A load of numbers, a load of potential explanations, dig into the numbers and every explanation falls (eg if abortions reduced crime it would be correlated across other datasets, ie countries) except for one or two.

At the moment my leaning is actually with you, but mainly from anecdotal evidence. Almost every trip I've had to America has has this weird race-vibe to it at some point. I go to a wedding, and everyone is 95% a minority race. I go to a comedy night, comedian jokes about my race. I hear someone talking, I somehow know what colour they are before I see them. So something about the society has race coded into it.

And then of course there are these horrendous incidents that we hear about every now and again, where some poor black man has been shot by police. I'm sure far more have simply been mistreated, because just by it coming up in conversation with other people, it turns out one of them witnessed such a beating.

But what we need is proper statistical evidence. The BI charts you've got there would be torn apart by an economist in a second.

Re: Santa Cruz, California bans predictive policing in U.S. first

#363

Post by AnthonyMouse » Fri, Jun 26, 2020, 10:04 PM UTC

bhntr3 wrote: post #335

nwienert wrote: post #291

Earlier quoted context omitted.

Yea, both of these points are so bad as to be almost parody. Not only should you never, ever base your beliefs based on avoiding the discomfort of being an outsider, but you also should never, ever, use majority opinion as some sort of proof of... anything. If you did, you’d have advocated for slavery if only you’d been born south of the Mason Dixon years ago.

Here's the wikipedia article for the fallacious argument I used and which you're critiquing: https://en.wikipedia.org/wiki/Argumentum_ad_populum Here is some statistical evidence for systemic racism: https://www.businessinsider.com/us-systemic-racism-in-charts... (Shockingly this is actually pretty good despite being from Business Insider) Honestly, I'd just edit the link at this point if I could since it seems to ha…

> Here is some statistical evidence for systemic racism

None of those charts appear to be controlled for confounders. Doing that correctly is hard, but not doing it at all is ridiculous. It allows you to "prove" that many engineering schools (with majority-white admissions boards) are racist against white people in favor of Asians, or that "racism" exists to the benefit of first generation African immigrants who by many metrics are better off than the US population average.

> I think social proof, despite not being actual proof, is more powerful than you folks would like to believe it is.

It can be effective in convincing people on political issues, but that's pure tribalism. It's anti-science, because it "works" independent of whether there is any truth in the assertion or not.

Re: Santa Cruz, California bans predictive policing in U.S. first

#364

Post by chaostheory » Fri, Jun 26, 2020, 10:07 PM UTC

Misdicorl wrote: post #324

alfalfasprout wrote: post #232

Earlier quoted context omitted.

I agree that, in general, you're going to have biased training data (with a bias that's difficult to measure) and so inherently policing recommendations will be biased. This is particularly problematic when it comes to arrests for crimes that are inherently 'selective' in their enforcement. Eg; drug-related crimes, public intoxication, loitering, trespassing. But the fact is... violent (fatal and nonfatal) crimes do…

Horseshit. Police response to violent incidents in 'problematic' neighborhoods is already extremely slow. The threat of gang violence is not mitigated by police response. The threat of gang violence is mitigated by providing much better options for people who would otherwise join gangs. Police presence is immaterial to the lack of gangs in affluent neighborhoods.

This could be one of the reasons why

https://www.npr.org/2018/12/12/675359781/americas-growing-co...

https://www.thedenverchannel.com/news/national-politics/the-...

It will only get worse.

I wonder if AI and robotics will need to be used in the future assuming that the police shortage trend continues or gets even worse?

Re: Santa Cruz, California bans predictive policing in U.S. first

#365

Post by rexpop » Fri, Jun 26, 2020, 10:09 PM UTC

ryankemper wrote: post #356

bhntr3 wrote: post #184

The problem with predictive policing is in the name. Inference (ML) predicts the future from the past. If the past is racist, then inference will create a racist future. Since racism is systemic[1], especially when it comes to policing, predictive policing is actively working against an anti-racist future. There may be statistical ways to factor out systemic racism. There are two reasons I don't think that works: 1.…

> Since racism is systemic[1] > [1] If you don't believe this, you're in the minority now: https://www.vox.com/2020/6/11/21286642/george-floyd-protests... . This part was a little odd. It sounds like you're saying "because this belief is widely held by people, we should treat it as true"? I definitely am in the minority, but I find a much more logically consistent view than the modern definition of systemic racism to…

> behavior that happens to be broken along racial and class lines

A huge coincidence.

Re: Santa Cruz, California bans predictive policing in U.S. first

#366

Post by mnm1 » Fri, Jun 26, 2020, 10:10 PM UTC

ryankemper wrote: post #357

vmception wrote: post #355

Earlier quoted context omitted.

Swing and a miss, overpolicing has nothing to do with violent crime. It has to do with nonviolent infractions that marginalize people to begin with. Being actually caught with illicit substances , or caught for the unpaid ticket. These things are evenly distributed across society, but when overly targeting minorities it further marginalizes them. Exhibit A) YOU might not be able to relate to people that routinely hav…

> These things are evenly distributed across society This is an unproven assertion. You should provide evidence of it, because on its face it is a completely ludicrous claim. Rates of illicit drug activity, violent crime, etc are _not_ evenly distributed in society. Not even close. > Exhibit B) Silicon Valley openly brags about microdosing acid to perform at work at their half million $ jobs. Overpolicing would ensna…

Actually, black people use drugs at slightly lower rates than whites. This is common knowledge or should be. Here's 3 of the top 6 google links for "illicit drug use by race":

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2377408/

https://www.hamiltonproject.org/charts/rates_of_drug_use_and...

https://www.sciencedirect.com/science/article/pii/S004723521...

Re: Santa Cruz, California bans predictive policing in U.S. first

#367

Post by TallGuyShort » Fri, Jun 26, 2020, 10:10 PM UTC

UnpossibleJim wrote: post #264

TallGuyShort wrote: post #174

Earlier quoted context omitted.

If crime being caught by a drive-by patrol is a significant way in which crime is being caught, to me that screams "victimless crimes". I break victimless laws all the time, sometimes with my police officer friends and neighbors present, sometimes on private property miles from any public roads. I never seem to get busted for it. So if 10-block patrols of "troubled areas" really saw an increase in enforcement action,…

What you're describing is a reform of America's drug laws, not of America's criminal prediction problem and they shouldn't be conflated. Police already use prediction models, they just do it in their heads and inaccurately with much more bias and have a harder time adjusting those bias'. I'm all for changing drug laws and trying to reduce recidivism through counseling and means other than pure imprisonment, but that'…

Drug laws are one example but by no means the only one. I've been on police ride-alongs where police watch a "suspicious" (i.e. poor) looking car as it passes them heading in the opposite direction. Oh look! They have a tail light out, we can pull them over. Then we can shine a flashlight in their backseat and see a bulge under a blanket. We can ask them certain questions that will either make us suspicious, or give us permission to search the car. Oh look! They have a pistol magazine that holds more than 15 rounds. Now there's a felony that they know the local courts will uphold despite there being a higher burden of proof about when the magazine was purchased than the officers can provide. Nevermind that the local gun stores all sell those magazines, and the police even buy them for personal use at those stores. But now, because of the way law enforcement works and the culture that has become normalized, somebody has a felony because of selective enforcement.

Re: Santa Cruz, California bans predictive policing in U.S. first

#368

Post by umvi » Fri, Jun 26, 2020, 10:10 PM UTC

thebradbain wrote: post #354

starkd wrote: post #285

Earlier quoted context omitted.

Not sure You can always separate out violent from non-violent events. The police are expected to deal with a whole community in a holistic way.

Why does a victim of sexual assault need to report to someone who has a gun? Why does a wellness check need to be performed with a gun? Why does a standard highway patrolman or cop doing routine speed traps need a gun? Why does every cop in a school need a gun? Why is the gun the constant?

> Why does a standard highway patrolman or cop doing routine speed traps need a gun?

Because civilians have shot and killed police during traffic stops? And not just with guns. I seem to remember a road rage case in Colorado where the raging driver shot and killed another driver with a crossbow.

Re: Santa Cruz, California bans predictive policing in U.S. first

#369

Post by zelly » Fri, Jun 26, 2020, 10:12 PM UTC

Living in a country with an inefficient police department is not fun. Ask anyone from you know who came from Cuba, Brazil, Honduras, Venezuela, or South Africa.

Living in a country with an extremely efficient police department is not fun either. See: China and Singapore.

There is a balance. Banning soulless social credit-tier stuff like this is a good step in the right direction. Another problem with stuff like "predictive policing" is that everyone knows a hypothetical 100% enforcement of the law would put almost everybody in prison; there are too many laws on the books in the USA. The first laws that come to mind are about the drug war. (Even cops are drug (steroid) addicts, so they themselves would go to prison too under 100% enforcement.) I don't use drugs, not even alcohol, but if someone thinks some white powder is going to make them happy or selling it would put food on the table then they should not be brutalized by the state for it.

ML and mass surveillance is going to actually allow 100% enforcement of laws, so it's time to have this discussion and redo the criminal code.

Re: Santa Cruz, California bans predictive policing in U.S. first

#370

Post by briandear » Fri, Jun 26, 2020, 10:12 PM UTC

r00fus wrote: post #340

starkd wrote: post #285

Earlier quoted context omitted.

Not sure You can always separate out violent from non-violent events. The police are expected to deal with a whole community in a holistic way.

It's easy - when someone wants a well-check or reports mental health or substance abuse or maybe even rape, that should go to someone other than police. Showing up with a gun to some events is guaranteed to make things worse.

A rape investigation should be done by a social worker? Are they trained criminal investigators? Is a social worker familiar with the rules of criminal evidence? The job of the police is to investigate crimes and present evidence to the DA.

For welfare checks or mental illness incidents, then of course a social worker or analogue could be a better choice. For investigating criminal acts, that’s specifically what police do.

If someone calls in a rape, what happens when the suspect is still there? Rape is a violent crime and showing up with a gun is absolutely appropriate. It’s ridiculous to suggest violent crimes should be investigated by people who aren’t trained in criminal investigation nor trained to deal with violent and dangerous suspects. Is the social worker going to carry handcuffs? Are they going to be trained in apprehending suspects? If so, then that means the social worker is now a cop.

Let’s not throw the baby out with the bath water here.