34 ms·

>And while at it, maybe even have a limit on the number of targets each agency can investigate, so they chose them wisely. This is an insanely bad idea. I und

>And while at it, maybe even have a limit on the number of targets each agency can investigate, so they chose them wisely. This is an insanely bad idea. I understand where you're coming from, but think about how that would play out in real life. I was going to write it out, but I really think that you can probably figure it out if you do think it through. From the perspective of the intel/policing agency, and from a political side, there are too many good reasons why we should not limit the number of targets an agency can investigate.

nickpsecurity 11y ago
How about actively at one time if not in general? I think if you think through it you can figure it out. Just start with all the terrorists and crooks they already had red flags on but did nothing. Then, look at what excuse they make about connecting the dots in so much data. Then look at the two obvious solutions: (a) Magic, AI, pixie dust tech that spots the evil, relationships, and all key information in nearly unlimited, streaming data. (b) Eliminating bulk collection (except maybe metadata) while forcing them to focus on the results of targeted investigations on accounts or individuals. Also, forcing them to follow-up on solid, red flags rather than BS around trying to find reasons everyone else might be guilty. I think pushing for option B is a rational choice. Also, ensuring the greatest accountability possible given all the main players have a history of using their police and intelligence for matters that have nothing to do with protecting citizens.
- kobayashi 11y ago
  When does an investigation start and end? What about when you're close to filling your limit but a few new great leads come in. Now the agents have to go close a ongoing investigation to follow up on a new one. Sounds like a lot of overhead. Not only that, but the fact of modern intel work is that investigations don't operate in a clear, linear progression. I think that anyone that's had to deal with arbitrary limits imposed from above understands the perverse incentives that they create. This would be another example of administrators trying to fix something that they don't understand.
  - nickpsecurity 11y ago
    You're focused on lots of hypotheticals whereas I'm talking about what they do. Currently, the NSA and FBI are looking at hundreds of millions of connections trying to find needles in stacks of needles. They admit, outside these political debates, that they're drowning in data. FBI's metadata investigations are producing jack and missing major attacks while targeted investigations following actual evidence are resulting in convictions. Speaking of perverse incentives, the current model is so ineffective and broken that they're actually creating terrorists to convict while claiming the arrests are due to successful, post-9/11 power. Even have a number on the incentive: $100,000 per terrorist created and arrested. https://theintercept.com/2015/04/15/fbi-informant-stung-fbi/ https://theintercept.com/2015/04/15/fbi-informant-stung-fbi/ So, I'm pushing for more than an arbitrary limit on number of investigations. I'm pushing for them to switch their focus back to pre-9/11 where they leave most of America alone, keep good red flags, and focus their investigators on those. The barriers to intelligence sharing have been knocked down and local LEO's are integrated so dots should be connected faster than ever. With this focus, the limit will be the number of cases they're working simultaneously rather than on the books in general. It can be a multiple of the number of agents they have. Because it's hard for me to imagine one investigator effectively working ten, time-critical cases simultaneously. That's just overstretched although exceptions can be made for elite, multitasking, savant investigators if and where they exist. So, we start with non-hypothetical stuff: the current surveillance dragnet that treats a whole country as guilty and watches every communication/association is an utter failure, a waste of money, and imposes chilling effects on democracy. The alternative is the original model of presumption of innocence, trained investigators following leads, prioritizing for most serious, and following through with good policework. Interestingly, the alternative already exists and runs in parallel with the surveillance state. Most convictions come from the alternative method. So, I don't have to justify it so much as point out that one is field-proven, it harms fewer innocents, and its cheaper. Ditching the failed method for that one is a no-brainer. Unless, it's about money for defense sector and power over dissidents/competition for the nation that wields it. In that case, they'd try to keep the surveillance tech while continuing to lie about its purpose of protecting citizens from stuff it can't catch. Like we're seeing. ;)
  - coldtea 11y ago
coldtea 11y ago
>This is an insanely bad idea. I understand where you're coming from, but think about how that would play out in real life. Make it a sane number, and it's not gonna play that bad at all. The only problem would be them claiming "if I had more allowable targets I would have caught this or that case before it happened" etc. Which we should just ignore. After all, before they got these modern toys, they used to be content with 1/1000000 the available targeting that they have now. In the seventies and eighties even, to target somebody involved lots of people actually in the streets, a semi-manual process to record his phonecalls (and you had to go through them with a real person, no keyword triggering etc), no "web" and other records automatically available etc. Checking half or all of the US at once at the scale we do today was not an option, not even 1/100000 that. You could ask for phone records and a couple other items at best.

nickpsecurity 11y ago

How about actively at one time if not in general? I think if you think through it you can figure it out. Just start with all the terrorists and crooks they already had red flags on but did nothing. Then, look at what excuse they make about connecting the dots in so much data. Then look at the two obvious solutions: (a) Magic, AI, pixie dust tech that spots the evil, relationships, and all key information in nearly unlimited, streaming data. (b) Eliminating bulk collection (except maybe metadata) while forcing them to focus on the results of targeted investigations on accounts or individuals. Also, forcing them to follow-up on solid, red flags rather than BS around trying to find reasons everyone else might be guilty. I think pushing for option B is a rational choice. Also, ensuring the greatest accountability possible given all the main players have a history of using their police and intelligence for matters that have nothing to do with protecting citizens.

kobayashi 11y ago
When does an investigation start and end? What about when you're close to filling your limit but a few new great leads come in. Now the agents have to go close a ongoing investigation to follow up on a new one. Sounds like a lot of overhead. Not only that, but the fact of modern intel work is that investigations don't operate in a clear, linear progression. I think that anyone that's had to deal with arbitrary limits imposed from above understands the perverse incentives that they create. This would be another example of administrators trying to fix something that they don't understand.
- nickpsecurity 11y ago
  You're focused on lots of hypotheticals whereas I'm talking about what they do. Currently, the NSA and FBI are looking at hundreds of millions of connections trying to find needles in stacks of needles. They admit, outside these political debates, that they're drowning in data. FBI's metadata investigations are producing jack and missing major attacks while targeted investigations following actual evidence are resulting in convictions. Speaking of perverse incentives, the current model is so ineffective and broken that they're actually creating terrorists to convict while claiming the arrests are due to successful, post-9/11 power. Even have a number on the incentive: $100,000 per terrorist created and arrested. https://theintercept.com/2015/04/15/fbi-informant-stung-fbi/ https://theintercept.com/2015/04/15/fbi-informant-stung-fbi/ So, I'm pushing for more than an arbitrary limit on number of investigations. I'm pushing for them to switch their focus back to pre-9/11 where they leave most of America alone, keep good red flags, and focus their investigators on those. The barriers to intelligence sharing have been knocked down and local LEO's are integrated so dots should be connected faster than ever. With this focus, the limit will be the number of cases they're working simultaneously rather than on the books in general. It can be a multiple of the number of agents they have. Because it's hard for me to imagine one investigator effectively working ten, time-critical cases simultaneously. That's just overstretched although exceptions can be made for elite, multitasking, savant investigators if and where they exist. So, we start with non-hypothetical stuff: the current surveillance dragnet that treats a whole country as guilty and watches every communication/association is an utter failure, a waste of money, and imposes chilling effects on democracy. The alternative is the original model of presumption of innocence, trained investigators following leads, prioritizing for most serious, and following through with good policework. Interestingly, the alternative already exists and runs in parallel with the surveillance state. Most convictions come from the alternative method. So, I don't have to justify it so much as point out that one is field-proven, it harms fewer innocents, and its cheaper. Ditching the failed method for that one is a no-brainer. Unless, it's about money for defense sector and power over dissidents/competition for the nation that wields it. In that case, they'd try to keep the surveillance tech while continuing to lie about its purpose of protecting citizens from stuff it can't catch. Like we're seeing. ;)
- coldtea 11y ago

coldtea 11y ago

>This is an insanely bad idea. I understand where you're coming from, but think about how that would play out in real life. Make it a sane number, and it's not gonna play that bad at all. The only problem would be them claiming "if I had more allowable targets I would have caught this or that case before it happened" etc. Which we should just ignore. After all, before they got these modern toys, they used to be content with 1/1000000 the available targeting that they have now. In the seventies and eighties even, to target somebody involved lots of people actually in the streets, a semi-manual process to record his phonecalls (and you had to go through them with a real person, no keyword triggering etc), no "web" and other records automatically available etc. Checking half or all of the US at once at the scale we do today was not an option, not even 1/100000 that. You could ask for phone records and a couple other items at best.