3 ms·

This is a really elegant point. > Doing upfront authorization of powers is relatively efficient, and is also pleasingly self regulating. If an agency overloads

This is a really elegant point. > Doing upfront authorization of powers is relatively efficient, and is also pleasingly self regulating. If an agency overloads or confuses its ex ante regulator, they simply won’t get permission to do things. This provides a strong incentive for clear and concise requests to the regulator. It is very easy for regulators, even well-meaning, to focus on what they require of an individual transaction. But we live in a world where automation and dynamic workforces make it possible to break every expectation about how a system will be used.

dekken_ 4y ago
Those that would impose unaccountable automated systems think themselves your master.
permo-w 4y ago
just to play devil’s advocate, I would imagine it could and will be argued that this system is broken by being too much at the mercy of the competence of the regulator. if national police need to tap someone’s phone now or lives may be lost, and the regulator drags their feet/is understaffed/has an inefficient workflow, that is an issue personally, I support the ‘ex ante’ approach, but - especially for the all-too-common kind of free-market ideologues that despise government regulation and have no trust in the competence of public organisation whatsoever - shoot first, ask questions later is hard to argue against
- upsidesinclude 4y ago
  It's just that this almost never happens. It's always the excuse for why unlimited surveillance should be granted to the state, but we know from 20 years since the "patriot" act passed that it doesn't ever help. Of course it has been abused verifiably many times & countless times that can't be verified.
- tene 4y ago
  I understand that you're not personally advocating for this position, but I'm glad you brought it up, because I really resent this kind of "Here's an imaginable scenario where this would have some negative consequence" argument. Yes, if the regulators stop doing their job well, and there's a sudden extreme emergency requiring some kind of action that regulators have not already approved for use in emergencies, then there will be some costs, and sometimes those costs can be measured in lives. This is true, but all it says is "there is some nonzero chance of society paying some nonzero cost". These costs are what we are paying in order to have a well-regulated intelligence service. The bet is that the expected risk from an effectively-unregulated intelligence service has worse costs for society than one with effective regulation. In order for "think about the children" to be a meaningful argument, you need to actually establish that the nightmare scenario is meaningfully more likely than overreach and abuse of power that causes similar or worse costs for society. Has this kind of "We could save the children if only we could get regulator approval to tap this phone line! Unfortunately, the regulator is taking a nap, so we're forced to let the children die." scenario actually been happening? If so, is there any kind of much-more-specific permissions that could be granted by the regulators to address the actual emergencies that have been coming up? I kind of get "You can't trade off a life!" for some kinds of arguments, but we're talking about national security issues, and failures of corruption and overreach also involve risking lives. "We need to just drop all safeguards and trust our valiant heroes" only works if the people who are subject to regulations actually are pretty reliably valiant heroes, or there are other significant incentive and oversight mechanisms to rely on. I don't have personal experience with people who work in national intelligence and security, but I haven't ever heard anyone willing to say that people in this line of work are consistently virtuous and corruption-resistant. There are good individuals, certainly, but there really are also both selfish individuals, and well-intentioned-but-misinformed individuals who can do a lot of damage.

dekken_ 4y ago

Those that would impose unaccountable automated systems think themselves your master.

permo-w 4y ago

just to play devil’s advocate, I would imagine it could and will be argued that this system is broken by being too much at the mercy of the competence of the regulator. if national police need to tap someone’s phone now or lives may be lost, and the regulator drags their feet/is understaffed/has an inefficient workflow, that is an issue personally, I support the ‘ex ante’ approach, but - especially for the all-too-common kind of free-market ideologues that despise government regulation and have no trust in the competence of public organisation whatsoever - shoot first, ask questions later is hard to argue against

upsidesinclude 4y ago
It's just that this almost never happens. It's always the excuse for why unlimited surveillance should be granted to the state, but we know from 20 years since the "patriot" act passed that it doesn't ever help. Of course it has been abused verifiably many times & countless times that can't be verified.
tene 4y ago
I understand that you're not personally advocating for this position, but I'm glad you brought it up, because I really resent this kind of "Here's an imaginable scenario where this would have some negative consequence" argument. Yes, if the regulators stop doing their job well, and there's a sudden extreme emergency requiring some kind of action that regulators have not already approved for use in emergencies, then there will be some costs, and sometimes those costs can be measured in lives. This is true, but all it says is "there is some nonzero chance of society paying some nonzero cost". These costs are what we are paying in order to have a well-regulated intelligence service. The bet is that the expected risk from an effectively-unregulated intelligence service has worse costs for society than one with effective regulation. In order for "think about the children" to be a meaningful argument, you need to actually establish that the nightmare scenario is meaningfully more likely than overreach and abuse of power that causes similar or worse costs for society. Has this kind of "We could save the children if only we could get regulator approval to tap this phone line! Unfortunately, the regulator is taking a nap, so we're forced to let the children die." scenario actually been happening? If so, is there any kind of much-more-specific permissions that could be granted by the regulators to address the actual emergencies that have been coming up? I kind of get "You can't trade off a life!" for some kinds of arguments, but we're talking about national security issues, and failures of corruption and overreach also involve risking lives. "We need to just drop all safeguards and trust our valiant heroes" only works if the people who are subject to regulations actually are pretty reliably valiant heroes, or there are other significant incentive and oversight mechanisms to rely on. I don't have personal experience with people who work in national intelligence and security, but I haven't ever heard anyone willing to say that people in this line of work are consistently virtuous and corruption-resistant. There are good individuals, certainly, but there really are also both selfish individuals, and well-intentioned-but-misinformed individuals who can do a lot of damage.