Stars, planets and other Sci-Fi peeves

HansMustermann said:
Basically here's the short version of the problem with proposing extreme systems: caring for the safety and wellbeing of your subjects should be about ALL subjects. INCLUDING those who end up victimized by the STATE, and including those who are victimized in the name of protecting others.

And frankly that's why we have so many safeguards and separations of powers and whatnot. Because the state can do the most harm. Now I'm not saying get rid of the state, but just that those safeguards are there for a reason. There are good reason why we require a jury of his/her peers, and evidence, and have an assumption innocence until proven guilty, and so on.

And to hammer some point some more, it's not even a new concept. The Magna Carta, early 13'th century, had this sentence, "No Freeman shall be taken or imprisoned, or be disseised of his Freehold, or Liberties, or free Customs, or be outlawed, or exiled, or any other wise destroyed; nor will We not pass upon him, nor condemn him, but by lawful judgment of his Peers, or by the Law of the Land." My emphasis, because this is what this is about: justice can destroy someone's life. And it was clear that far back in time.

Looking at just how many victims you prevent from having their life destroyed, is missing the concern for how many lives does your state destroy in the process.
Click to expand...

That's a beautiful illustration of how extremely you're over-analysing this. This has nothing to do with the conversation.
 
Delvo said:
You appear to be applying the 0.01% error rate to the entire population. But that's not what an (in)accuracy rate is. It's 0.01% of all predictions.

Your accuracy at a gun/archery range is hits on target divided by shots you took, not divided by the number of bullets/arrows in the city.
Click to expand...

But if you have to predict who in the total population will commit a crime, and who won't, you have to essentially predict both that some will and that some won't.

Homicide and (non-negligent) manslaughter rate in the USA (as of 2014) is 4.9 per 100,000 population. To stop even one, you have to have some criterion to predict that in a town of 20,000 people not only this guy will, but that the other 19,999 guys won't. The moment you claim any criterion can tell you which of the 20,000 will commit a crime, or even narrow it down to a smaller sample to look at, you ARE applying the criterion to all 20,000 and making a prediction.
 
Argumemnon said:
That's a beautiful illustration of how extremely you're over-analysing this. This has nothing to do with the conversation.
Click to expand...

But I think it does. I think that is the problem that people have with that scenario: that it allows some organization to ruin some people's lives for stuff they may or may not actually do. Which is not the nirvana fallacy.
 
It relates because you assume that the problem is "It's not 100% perfect at preventing crime, therefore it's crap." While actually the issue is more like HOW does it prevent crime: by ruining the lives of a LOT of innocents. THAT is the problem.
 
HansMustermann said:
It relates because you assume that the problem is "It's not 100% perfect at preventing crime, therefore it's crap."
Click to expand...

I don't assume it, I conclude it. That's the implication from the movie.

While actually the issue is more like HOW does it prevent crime: by ruining the lives of a LOT of innocents. THAT is the problem.
Click to expand...

It doesn't prevent crime by ruining lives. It intercepts murderers before they act, SAVING lives. I've already told you this before, so I'll bolden it this time: The issue is that the police then send you away for murder without further proceedings. Farrel's character makes a good point about that early in the movie but it's hand-waved away.
 
Well, if they brand you a murderer and send you away, they did just ruin your life, didn't they?
 
HansMustermann said:
Well, if they brand you a murderer and send you away, they did just ruin your life, didn't they?
Click to expand...

Hence the bolded part of my post, which you obviously didn't read, so I will make it bigger: You could have further legal proceedings after the arrest to make sure you're not sending an innocent to jail. The tech could still be very useful.
 
Proceedings based on WHAT evidence? If you still require enough evidence to prove an attempted murder or conspiracy to commit murder or such, you're essentially back to today's system. And if it's based on just trusting someone or some system to be omniscient and know the future, then you're back to square one: sending a lot of innocents to jail anyway.

There is nothing magical about requiring a judge. Mao's had one too. Stalin's NKVD troikas had THREE judges who had to agree that yeah, that guy is too rich to be a peasant, he deserves to be shot. So what?

The real safeguards are due process, demanding evidece, the assumption of innocence, habeas corpus, etc. Once you remove evidence in favour of just trusting that someone or something is right about something that hasn't even happened yet -- INCLUDING a system that has a judge or whatever -- then you're back to the problem: ANY margin of error means ruining more lives than you prevent.
 
HansMustermann said:
But if you have to predict who in the total population will commit a crime, and who won't, you have to essentially predict both that some will and that some won't.
Click to expand...
Have you seen the movie though? That's not how it works. They aren't making predictions for everyone, they are having visions of specific *events*. At a fairly low rate. There isn't a constant stream of visions coming out for the whole population. Your predicted false positive rate is significantly higher than the prediction rate seen in the movie, and therefore impossible.
 
HansMustermann said:
But if you have to predict who in the total population will commit a crime, and who won't, you have to essentially predict both that some will and that some won't.

Homicide and (non-negligent) manslaughter rate in the USA (as of 2014) is 4.9 per 100,000 population. To stop even one, you have to have some criterion to predict that in a town of 20,000 people not only this guy will, but that the other 19,999 guys won't. The moment you claim any criterion can tell you which of the 20,000 will commit a crime, or even narrow it down to a smaller sample to look at, you ARE applying the criterion to all 20,000 and making a prediction.
Click to expand...

I have some system that predicts homicides and their perpetrators. It makes a prediction: Alice will kill Bob tomorrow at 9am.

I wait and see.

It turns out it's right.

It makes another prediction: Charlie will kill Doris tomorrow at 1:30pm.

I wait and see.

It turns out it's right.

I let this run until I get enough data to make a statistically valid assessment of it's accuracy and find that it's 99.95% accurate in that the event it predicts actually takes place and the perpetrator is the person predicted.

The rate of false positives will be 5 innocents predicted to commit murder for every 9995 guilty people.

This how the predictions are presented in the film.
 
So, the plan is to arrest the person prior to the murder as they have been prophesied to, at some point in the future, commit the murder.

Then, as the system is not perfect, there is then a trial?

How does one prove that they were gonna have done it? I suspect that's rather more tricky than the current situation where one has to prove that they did do it.
 
3point14 said:
So, the plan is to arrest the person prior to the murder as they have been prophesied to, at some point in the future, commit the murder.

Then, as the system is not perfect, there is then a trial?

How does one prove that they were gonna have done it? I suspect that's rather more tricky than the current situation where one has to prove that they did do it.
Click to expand...

Oh absolutely, but at least you stopped 9995 murders for each 10000 interventions. That's pretty good, either way.
 
Argumemnon said:
Oh absolutely, but at least you stopped 9995 murders for each 10000 interventions. That's pretty good, either way.
Click to expand...


You still have to have 9995 trials.

In any of those trials that were for murders committed on the spur of the moment, rather than those that are intricately planned, there will be literally no evidence for the prosecution to present other than 'the precogs said they'd do it'
 
Last edited:
Argumemnon said:
Oh absolutely, but at least you stopped 9995 murders for each 10000 interventions. That's pretty good, either way.
Click to expand...

Or you could, and I know this is a crazy idea here, but bear with me, you could just use it as a source of intel, and position people to watch the suspected victim, and stop the predicted perpetrator in the act.

Then you get proof and save lives.
 
Hellbound said:
Or you could, and I know this is a crazy idea here, but bear with me, you could just use it as a source of intel, and position people to watch the suspected victim, and stop the predicted perpetrator in the act.
Click to expand...

Isn't that exactly what I said? That the tech allows you to stop the crime before it happens? That the issue is that the movie has the police throw you in jail immediately thereafter (ETA: without a trial) ?
 
Last edited:
You must log in or register to reply here.

Back
Top Bottom