6 ms·

They found a body with a bullet in it's head. That's step one. Step two, they went to these recordings, looked where they found that body, and they tracked all

by Retra 10y ago

They found a body with a bullet in it's head. That's step one. Step two, they went to these recordings, looked where they found that body, and they tracked all the people who were there when the murder happened. >an entire narrative is presented without any shred of evidence Why should it? It's not trying to convince you a murder took place. That's already an established fact. It's trying to show you the technology. Nothing here would have been able to predict a crime. Someone driving around the block twice before driving across town might just be a driver whose car is making a funny sound when he turns left, so he's investigating it. That means nothing. Unless you've got a bloody corpse laying in the alley he's driving past. Then it means a whole lot.

MichaelMoser123 10y ago
> They found a body with a bullet in it's head. That's step one. i am not much of an expert, but for a start they would need the exact time of the murder - without this fact all these little dots that move around the screen don't make a lot of sense.
- ctdonath 10y ago
  Forensics is good enough to narrow the time down enough to get a good start.
duncan_bayne 10y ago
Until an AI monitoring the video dispatches the police to intercept you, because it's decided that behaviour like that is associated with subsequent criminal activity.
- Retra 10y ago
  I'm sure the police have the time and resources to spend their days running all over town chasing down false positives and arresting people for driving in circles.
vidarh 10y ago
> Then it means a whole lot. Assuming he saw it. Part of the problem with this is that you are looking at a scene with additional knowledge which the people present may or may not also have had, and you're looking at it from a vastly different angle. If these things makes it to court, expect a lot of challenges on the basis of those kind of things. If it's used merely as a starting point for "real police work" then that may be ok. But consider e.g. the policy of extrajudicial assassinations of suspected drug dealers for money in the Philippines[1], where a system like this would potentially allow police officers to "justify" their decisions far faster (and there's a lot of people with money in that) by playing fast and easy with what "looks like" a drug deal. Or consider that the US has a government that considers extrajudicial killings of terror suspects by drone acceptable. How long until systems like the above play into providing "evidence" into those kind of decisions - we already know they don't have a problem with rather substantial collateral damage. This kind of technology does have a lot of potential for good. But it also has a lot of potential for supporting authoritarianism and for aiding state terror. [1] http://www.bbc.co.uk/news/world-asia-37172002 http://www.bbc.co.uk/news/world-asia-37172002

by Retra 10y ago

MichaelMoser123 10y ago

> They found a body with a bullet in it's head. That's step one. i am not much of an expert, but for a start they would need the exact time of the murder - without this fact all these little dots that move around the screen don't make a lot of sense.

ctdonath 10y ago
Forensics is good enough to narrow the time down enough to get a good start.

duncan_bayne 10y ago

Until an AI monitoring the video dispatches the police to intercept you, because it's decided that behaviour like that is associated with subsequent criminal activity.

Retra 10y ago
I'm sure the police have the time and resources to spend their days running all over town chasing down false positives and arresting people for driving in circles.

vidarh 10y ago

> Then it means a whole lot. Assuming he saw it. Part of the problem with this is that you are looking at a scene with additional knowledge which the people present may or may not also have had, and you're looking at it from a vastly different angle. If these things makes it to court, expect a lot of challenges on the basis of those kind of things. If it's used merely as a starting point for "real police work" then that may be ok. But consider e.g. the policy of extrajudicial assassinations of suspected drug dealers for money in the Philippines[1], where a system like this would potentially allow police officers to "justify" their decisions far faster (and there's a lot of people with money in that) by playing fast and easy with what "looks like" a drug deal. Or consider that the US has a government that considers extrajudicial killings of terror suspects by drone acceptable. How long until systems like the above play into providing "evidence" into those kind of decisions - we already know they don't have a problem with rather substantial collateral damage. This kind of technology does have a lot of potential for good. But it also has a lot of potential for supporting authoritarianism and for aiding state terror. [1] http://www.bbc.co.uk/news/world-asia-37172002 http://www.bbc.co.uk/news/world-asia-37172002