News for hackers. Stuff that matters.
Re:Predictive Policing Software Is Pretty Terrible at Predic... by dahfizz on Tuesday October 03, 2023 @07:31PM (#37758976)
Vehicular manslaughter is already illegal, we should make drunk driving legal. Drinking and driving without crashing into anything is victimless.
Re:Predictive Policing Software Is Pretty Terrible at Predic... by heavyset_go on Tuesday October 03, 2023 @07:33PM (#37758988)
The equivalent would be banning all alcohol because sometimes people get drunk and do bad things.
Re:Predictive Policing Software Is Pretty Terrible at Predic... by jojobas on Tuesday October 03, 2023 @08:07PM (#37759258)
You chances of drinking alcohol responsibly for decades are way, way better than your changes on not becoming an unemployable wreck in mere years on meth or heroin.
Re:Predictive Policing Software Is Pretty Terrible at Predic... by bdangubic on Tuesday October 03, 2023 @08:54PM (#37759616)
an amazing sentence which sounds like it is rooted in something vaguely factual while in reality there is nothing factual that can support this claim
Re:Predictive Policing Software Is Pretty Terrible at Predic... by jojobas on Wednesday October 04, 2023 @05:31AM (#37762820)
There's plenty to support this claim: consumption prevalence and disorder prevalence.

Opiod users are vastly over-represented in all sort of statistics - addiction treatments, overdose deaths etc as compared with alcohol.

6% of alcohol drinkers are diagnosed with alcohol use disorder, the number exceeds 50% for meth users.

https://www.cdc.gov/mmwr/volumes/69/wr/mm6912a1.htm

Re:Predictive Policing Software Is Pretty Terrible at Predic... by master-lincoln on Wednesday October 04, 2023 @07:09AM (#37763473)
> 6% of alcohol drinkers are diagnosed with alcohol use disorder, the number exceeds 50% for meth users.

This seems like numbers you can not compare fairly. I would think if alcohol were illegal, a bigger percentage of users would be diagnosed with alcohol use disorder.

Re:Predictive Policing Software Is Pretty Terrible at Predic... by jojobas on Wednesday October 04, 2023 @07:17AM (#37763545)
Diagnostic criteria are quite agnostic (badum-tss) towards the chemical in question.

You could also argue that people consuming illegal substances are less likely to report for treatment and be diagnosed in the first place.

Why is it so hard to accept that different substances are differently addictive and harmful, per capita? Would you shoehorn chocolate and coffee to the same bucket?

Re:Predictive Policing Software Is Pretty Terrible at Predic... by master-lincoln on Wednesday October 04, 2023 @07:56AM (#37763868)
I am just careful in this area because previous policies have not been based on scientific evidence, but political agenda (as expanded upon other-where in this thread). So I don't trust my common sense... I accept that different substances have different effects.

Harm is a domain that is very broad and hard to quantify fairly I think. Does Meth cause more harm because more Meth addicts have fallen out of society than say Coke? If so, how can we make sure to differentiate between mere correlations and causal relations?

Re:Predictive Policing Software Is Pretty Terrible at Predic... by eru on Tuesday October 03, 2023 @09:22PM (#37759779)
That would be hard to get data on, because alcohol is mostly legal and meth and heroin are mostly illegal.

So you have a huge confounder: it seems likely people who are willing to do illegal acts, are more likely to become unemployable.

So if meth was legal, and alcohol was illegal, the situation might be reversed.

(I'm not saying this is true, but to make your statement you'd need to carefully exclude the possibility of this effect.)

Btw, heroin and other concentrated drugs are popular partially because we banned comparatively milder and bulkier alternatives. Eg opium ain't as hard as opium, but it's just as forbidden, and its bigger bulk makes it less profitable to smuggle.

Re:Predictive Policing Software Is Pretty Terrible at Predic... by cwkoss on Tuesday October 03, 2023 @11:55PM (#37760683)
Meth is legal, sometimes.

It's called desoxyn, and can be prescribed for ADHD or weight loss.

Also chirality-flipped meth is available over the counter as a nasal decongestant

Re:Predictive Policing Software Is Pretty Terrible at Predic... by conradev on Wednesday October 04, 2023 @02:29AM (#37761554)
Over the counter… with an ID. It’s not on shelves precisely because it is so similar to meth
Re:Predictive Policing Software Is Pretty Terrible at Predic... by bratgpttamer on Wednesday October 04, 2023 @09:31AM (#37764869)
It's not on shelves because it's a precursor chemical to actual meth manufacturing, not because it's a dangerous drug (otherwise it'd be Rx-only, like the other meth-like drugs).

https://www.deadiversion.usdoj.gov/meth/cma2005.htm

Re:Predictive Policing Software Is Pretty Terrible at Predic... by Paul-Craft on Wednesday October 04, 2023 @06:14PM (#37772545)
Neither ephedrine, pseudoephedrine, or phenylpropanolamine are the reversed-chirality methampetamine. "Regular" methamphetamine is dextromethamphetamine; the other version is levomethamphetamine.
Re:Predictive Policing Software Is Pretty Terrible at Predic... by heavyset_go on Wednesday October 04, 2023 @04:50PM (#37771575)
You don't need an ID for levomethamphetamine.
Re:Predictive Policing Software Is Pretty Terrible at Predic... by bryanrasmussen on Wednesday October 04, 2023 @04:25AM (#37762392)
>Eg opium ain't as hard as opium, probably need an edit here. Is it opium isn't as hard as heroin?

all the shows I've ever seen on it suggests it is pretty life destroying though.

Re:Predictive Policing Software Is Pretty Terrible at Predic... by thghtihadanacct on Wednesday October 04, 2023 @09:39AM (#37765005)
Opium literally gets scraped off of poppies. Anyone can actually obtain it with some patience and know how. It would take an enormous individual effort to collect enough to use, let alone hurt oneself or sell any.
Re:Predictive Policing Software Is Pretty Terrible at Predic... by Paul-Craft on Wednesday October 04, 2023 @06:19PM (#37772586)
Yet, ridiculously, sometimes people who have simply eaten a lot of poppy seed bagels or something end up getting tagged as heroin users, because the metabolites are so similar. Never mind that the quantity of poppy seeds you'd have to ingest in order to feel even a little bit of opiate effect is probably more than a human can physically consume. It gets picked up and flagged because the tests are that sensitive.
Re:Predictive Policing Software Is Pretty Terrible at Predic... by Eisenstein on Tuesday October 03, 2023 @09:23PM (#37759780)
Yet we give amphetamines and opiates to people under prescription en masse without such a thing happening with more prevalence than alcohol.
Re:Predictive Policing Software Is Pretty Terrible at Predic... by Paul-Craft on Wednesday October 04, 2023 @06:22PM (#37772616)
Yep, and I tell you what: on the rare occasions when I have accidentally doubled my dose of Adderall, I do not like the way it makes me feel one bit. An actual addict would probably be taking 10x or more of my prescribed dose to get their high. I wouldn't be able to enjoy the high because I'd feel like I was literally dying.
Re:Predictive Policing Software Is Pretty Terrible at Predic... by jimbob45 on Tuesday October 03, 2023 @09:23PM (#37759788)
We do ban alcohol because of that though. Maybe not a total ban but we regulate the hell out of it. You can’t even drink until 21 and then you can only drink in certain areas and you’re cut off once you show signs of intoxication.
Re:Predictive Policing Software Is Pretty Terrible at Predic... by komali2 on Tuesday October 03, 2023 @11:31PM (#37760537)
Ok, so we should regulate the hell out of meth, heroin, ecstasy, etc. It doesn't make sense to federally ban a drug like LSD considering how little harm it causes compared to the drug alcohol.
Re:Predictive Policing Software Is Pretty Terrible at Predic... by chii on Wednesday October 04, 2023 @12:51AM (#37760974)
> how little harm it causes compared to the drug alcohol.

but what harm per dose? May be the harm is high, but there's so few users that nothing shows up in aggregate.

Re:Predictive Policing Software Is Pretty Terrible at Predic... by lynx23 on Wednesday October 04, 2023 @03:05AM (#37761791)
It is far easier to use psychedelics irresponsible then it is with alcohol. Most people crash at some point when drunk, but usually do not undress and walk out onto a busy street... The tendency of psychedelics to unearth mental health problems is IMO too much of an issue given how many people have untreated issues these days.
Re:Predictive Policing Software Is Pretty Terrible at Predic... by sfn42 on Wednesday October 04, 2023 @07:58AM (#37763882)
I've taken most common drugs. LSD, MDMA, shrooms, cocaine, amphetamines, list goes on.

None of them, except maybe some prescription pills like benzos, take control away from me the way alcohol does.

None of them have had me wake up in some random place with no idea how I got there. None of them have made me do things I don't remember, none of them have made me do and say things I regret.

Alcohol has caused all of that and more. The only reason I still drink alcohol is because it's the only socially acceptable drug. I'd much rather do coke or molly, to me those feel significantly safer than alcohol. And they're better, and they don't ruin the next day to the same degree alcohol does, at least not in my experience.

And yes, alcohol absolutely can make people undress and walk out into a busy street. It can also make people aggressive. One time my father came home shitfaced, furious at me for no apparent reason, punched a hole in my bedroom window, came into my room and started screaming at me while bleeding everywhere from punching the glass, then grabbed me by the neck and started choking. I can't remember how it played put after that but I ended up spending the night at my aunt's place. The next day he was in shambles, he would never do anything like that sober. And for the next few years until I moved out, I was anxious every time he went out.

I still do illegal drugs. Not very frequently, I'm a regular guy who works a decent job, is dependable etc. I'm not going to work high or anything like that. Just like most people don't get drunk at work. I know many people with similar lifestyles. I'd bet the majority of illegal drug users are just like me, they have their fun in a responsible way and live normal lives. Just like with alcohol the really messed up addicts are the exception, not the norm.

Re:Predictive Policing Software Is Pretty Terrible at Predic... by shmel on Wednesday October 04, 2023 @07:58AM (#37763886)
It is a statistical bias. The vast majority of psychedelics users don't undress and walk around naked. But you often don't notice them precisely because psychs are illegal and people don't advertise their use. It is however widely accepted to drink in public. Everybody who's been in a busy downtown on Saturday night would agree that alcohol definitely brings up violence in some people. And yet it is typically brushed off because most people manage to drink responsibly. Unless you are in the psychedelic community, you just don't know what the majority of psychedelic users are like.

Go to a psytrance festival and see for yourself. I've been to many and I haven't seen people running around naked. If anything, I feel safer there than on many large events where people drink.

Re:Predictive Policing Software Is Pretty Terrible at Predic... by lynx23 on Thursday October 05, 2023 @04:08AM (#37776026)
I am inclined to not believe that due to my anecdotal evidence. I lost a dear friend at the age of 17 due to him trying the McKenna mushroom dose, and ending up jumping out of the window on the 7th floor. I life in a country where alcohol use is not unheard of, but after roughly 40 years living here, I still dont know anyone how died from alcohol intoxication in a similar way. Alcohol casualties are usually a long-term effect.
Re:Predictive Policing Software Is Pretty Terrible at Predic... by komali2 on Thursday October 05, 2023 @11:26AM (#37779841)
Well, trading stories, I've lost two close friends to drunk driving accidents and one because he slipped into an icy river when drunk.

I'm really sorry about your loss. Situations like that are awful. I think it's really important for people interested in psychedelics to try them in a safe environment with people who have experience with the given drug and are willing to stay sober and keep an eye out. Every time I've encountered psychedelics (hundreds of occasions) it's been in situations like this, and they prevent tragedies.

Those of us who are arguing for legalization are basically arguing for society to create these kinds of safety nets writ large. Rather than counting on having a wookie friend nearby you can straight up call up 411 or something before doing a drug. I know it sounds ridiculous but if people are going to do these things anyway... I mean we're happy to have society teach eachother how to have sex safely, why not drugs? And that includes alcohol.

Re:Predictive Policing Software Is Pretty Terrible at Predic... by itishappy on Wednesday October 04, 2023 @02:42PM (#37769710)
I've experimented with both psychedelics and alcohol (never together). It was alcohol that eventually caused a psychotic break. Scary stuff.
Re:Predictive Policing Software Is Pretty Terrible at Predic... by jahsome on Wednesday October 04, 2023 @12:08AM (#37760756)
Moreover, dry counties exist.
Re:Predictive Policing Software Is Pretty Terrible at Predic... by heavyset_go on Wednesday October 04, 2023 @03:02AM (#37761780)
You can still drink in dry counties and possession of alcohol is not illegal. Lack of liquor licenses is not the same thing has the criminalization of other substances.

The equivalent to that are towns that forbid the sale of marijuana in states that legalized it. It's still legal to use and possess.

Re:Predictive Policing Software Is Pretty Terrible at Predic... by jahsome on Monday October 09, 2023 @05:10PM (#37825519)
Great point thank you for making the distinction.
Re:Predictive Policing Software Is Pretty Terrible at Predic... by eru on Tuesday October 03, 2023 @09:19PM (#37759762)
You are going for a reduction to the absurd here, but what you are saying isn't actually that far off what is practical.

In general, many behaviours influence the chance of bad things happening. Eg getting behind the wheels of a car in the first place increases the chance of vehicular manslaughter (compared to walking).

Society can punish the probabilistically bad behaviour, or the bad outcome or any combination of the two (including none).

If the punishment is purely financial, and enforcement is certain and actors have deep pockets, it doesn't really matter which combination you pick.

You can emulate the deep pockets by requiring people to get insurance. Even if the law only polices the bad outcomes, a rational insurance company might want to police its customers behaviour. (Eg your health insurance gives you a discount, if you show them evidence of healthy habits. That's equivalent to charging people who fail to provide that evidence more, but the discount is more PR friendly.)

Both driving at all and drunk driving increase the chance of a person committing vehicular manslaughter. Societies largely allow driving, but ban drunk driving.

Drunk driving is dangerous, mainly because drunkards have slower reaction times and are more prone to taking risks. The more alcohol you have in your blood, the harsher the fines and punishment. But we typically don't check people's sober reaction times and risk appetite, and charge them more for driving sober.

It's all about trade-offs.

In the case of vehicular accidents, most people don't necessarily have deep pockets, and the damage of injuring or killing people can't be undone by paying money. However we eg still let old people drive, and only do something about them when they actually get involved in an accident.

Re:Predictive Policing Software Is Pretty Terrible at Predic... by NovemberWhiskey on Tuesday October 03, 2023 @11:12PM (#37760428)
The essential element of vehicular manslaughter that differentiates it from "a crash in which someone dies" is the mental state of the driver - specifically, that they're acting in a reckless or negligent manner.

This is fundamentally nothing to do with economic trade-offs: it's about generally disincentivizing risk-taking when in command of a multi-ton vehicle moving at speed.

We can know this because even if a "bad" driver is objectively more risky than a professional race-car driver who is 20yo, but has drunk a beer; it's the latter that's committing a crime when they drive.

Re:Predictive Policing Software Is Pretty Terrible at Predic... by Retric on Tuesday October 03, 2023 @11:21PM (#37760479)
That argument breaks down when you look at how these laws were actually created and changed over time.

People didn’t pick 0.08 BAC and then lower it because they found 0.08 BAC wasn’t low enough, instead the same organizations kept pushing for ever lower limits so they keep dropping independent of any actual impact.

Re:Predictive Policing Software Is Pretty Terrible at Predic... by NovemberWhiskey on Tuesday October 03, 2023 @11:41PM (#37760596)
In fact, that reinforces my argument: it's not about relative risk, it's about the definition of what's considered negligent varying over time based on changing attitudes.
Re:Predictive Policing Software Is Pretty Terrible at Predic... by Retric on Wednesday October 04, 2023 @04:56AM (#37762624)
Zero tolerance laws have nothing to do with impairment. 0.001 BAC doesn’t have a detectable influence on behavior but 7 states set have zero tolerance at literally 0.000. The definition of negligence isn’t changed by these groups, they are going to arbitrarily increase penalties over time because that’s why they exist.

Nobody advocates for drunk drivers as a group so there’s no equilibrium reached, just ever harsher penalties and lower limits. I don’t drink, but it’s obvious that using literally 0 as your threshold will pointlessly waste money going after edge cases unrelated to drinking.