Neat!
Guess it was to be expected for sattelites to come a long way since the original ones. I can see the advantages in a smaller sattelite. Just as I can see advantages in a bigger one...
They are watching us
Not really. It's like sitting on a park bench, or a bench in the local mall. Yeah, you're "watching" everyone's comings and goings, but it isn't exactly a big deal.
They are watching us
Not really. It's like sitting on a park bench, or a bench in the local mall. Yeah, you're "watching" everyone's comings and goings, but it isn't exactly a big deal.
Of course it is.
If you feel that way than I recommend never going outside your cave, Luddite.
I just find it confusing how people here complain about how the government is taking away freedoms, how now the police can go in a houses without a warrents, etc, etc, yet they are happy about satelites that can basically spy on them.
Ok... How is aerial photography spying? If you don't want to be seen outside your home, dont go out. Lemme know how that works for you. One of the key things to understand is just because they can doesn't mean they want to. Sort of like the people who claim that the government records every phone call everyone makes. Even if it did, where are the people to listen to all of these phone calls?
A quick search says that Americans make or recieve approximately 2.4 billion phone calls per day. If we half that number (because each conversation usually has two sides) that leaves us with 1.2. The averge length of a cell phone call, according to a quick search, is 3 minutes and 15 seconds. So, 3.25 minutes times 1.2 billion conversations... That's over 7400 years worth of audio. You'd need over 2.7 million people just listening to phone calls 24 hours a day, every day. If you assume 8 hour shifts, that turns into over 8.1 million people just to listen to phone calls. That doesn't count the payroll dept., human resources, managers, etc. The listeners alone would account for over 2% of the total US population. By contrast, there are fewer farmers in the US than that. And significantly fewer automechanics.
In other words, the manpower to watch everyone that the paranoid delusionists think the government does.. Well, that manpower just doesn't exist.
Filtering out the stuff you don't need so you need less people to search for the stuff you do want.
(No, I'm not serious. Neither was my other post)
Ok... How is aerial photography spying? If you don't want to be seen outside your home, dont go out. Lemme know how that works for you. One of the key things to understand is just because they can doesn't mean they want to. Sort of like the people who claim that the government records every phone call everyone makes. Even if it did, where are the people to listen to all of these phone calls?
A quick search says that Americans make or recieve approximately 2.4 billion phone calls per day. If we half that number (because each conversation usually has two sides) that leaves us with 1.2. The averge length of a cell phone call, according to a quick search, is 3 minutes and 15 seconds. So, 3.25 minutes times 1.2 billion conversations... That's over 7400 years worth of audio. You'd need over 2.7 million people just listening to phone calls 24 hours a day, every day. If you assume 8 hour shifts, that turns into over 8.1 million people just to listen to phone calls. That doesn't count the payroll dept., human resources, managers, etc. The listeners alone would account for over 2% of the total US population. By contrast, there are fewer farmers in the US than that. And significantly fewer automechanics.
In other words, the manpower to watch everyone that the paranoid delusionists think the government does.. Well, that manpower just doesn't exist.
I see a job creation opportunity!
Damn liberal! Those will be government jobs!
Only until the sector has grown enough to support competition, then we can privatize spying on the citizenry.
Edit: but remember that private industry would likely just outsource the whole operation.
Regarding listening to people's phone calls: Gimme twenty years, a department full of computer scientists and a budget. We'll have a super-cluster doing all that listening for us, with advanced linguistic algorithms flagging suspicious conversations for human review. The entire thing will eventually be a team of nerds keeping the computers going and a google browser.
Good, old-fashioned detective work is a lot more effective. Monitor the bad people, learn who the other bad people are, monitor the new bad people are, etc.
Note that I only talked about the manpower issue. I didn't even address the physical unlikelihood. (Enough storage space for 7400 years of audio on a daily basis?!)
Regarding listening to people's phone calls: Gimme twenty years, a department full of computer scientists and a budget. We'll have a super-cluster doing all that listening for us, with advanced linguistic algorithms flagging suspicious conversations for human review. The entire thing will eventually be a team of nerds keeping the computers going and a google browser.
Good, old-fashioned detective work is a lot more effective. Monitor the bad people, learn who the other bad people are, monitor the new bad people are, etc.
Note that I only talked about the manpower issue. I didn't even address the physical unlikelihood. (Enough storage space for 7400 years of audio on a daily basis?!)
If you're using listening algorythms then build in a 3-minute buffer, if something "suspect" gets flagged then it actually saves the buffered information and records the rest of the call for review by a fleshbag.
Even with a three minute buffer the storage requirements would be insane. The information I gave earlier is for call center recordings - you know, people evaluated for efficiency on-call. I'd say easily half of their calls average three minutes or less.
Oh, yeah, totally, you'd quickly be dealing with petabytes of audio.
For a consumer, a Petabyte of hard disk space costs about $40,000. I'm guessing a government or corporation could get it a lot cheaper.
Hell, Youtube uses (by estimation according to a google search) about 10 Petabytes of space to store all its movies. Which doesn't sound unreasonable.
Hard disk space is cheap, and it's only getting cheaper.
3TB is about double of what my home computer has plugged in, and it's a cheap ass machine with not even a very big harddrive.
For a consumer, a Petabyte of hard disk space costs about $40,000. I'm guessing a government or corporation could get it a lot cheaper.
Hell, Youtube uses (by estimation according to a google search) about 10 Petabytes of space to store all its movies. Which doesn't sound unreasonable.
Hard disk space is cheap, and it's only getting cheaper.
I don't think you are understanding the scale in question, though. He's talking about 3 TB being barely enough to handle 500 people's calls for 2 years. That's not two years of canstant audio, that's just two years of the call center being in operation. I'm talking about 7400 years worth of audio on a daily basis. When I say 7400 years worth of audio, I mean taking every second of audio and stringing it together into an unbroken stream that keeps going day and night for longer than recorded history. And that's what you are pulling in on a daily basis.
Besides; like Escher said... 20 years. 20 years ago a GB was considered enourmous... now the PB is commercially available. I don't think the storage is gonna be an issue in another 20 years.
It's what you're handling... the basic idea is that you're not storing all of it.
Besides; like Escher said... 20 years. 20 years ago a GB was considered enourmous... now the PB is commercially available. I don't think the storage is gonna be an issue in another 20 years.
In 20 years? Probably not. The probelm is that there are people who think we're doing it right now. The technology isn't there, the manpower isn't there.
Hrm... you know, I haven't actually heard of such a tool that the FBI/CIA/police uses, but it wouldn't surprise me. If nothing else they could use it in conjunction with a traditional wire tap to reduce the amount of manpower needed.
So something like recording the conversation waiting for keywords, if it finds 'em it keeps the conversation and forwards it to a human operator, if nothing suspicous is found, trashes it...
Wait... You honestly think they don't use code words? I mean, yeah, most terrorists are dumb as bricks, sure. But even they aren't that dumb. Seriously, just using common American slang I can make a drug transaction sound like I'm going to a clam bake.
So no. I do not honestly believe that terrorists/bank robbers/suspicious militants/jaded wives don't use code words. I just have greater confidence in language analysis tools then you.
As we saw on April 1st, 2010, it's quite easy to change a filter to include a broader selection of words. And if you're not monitoring everyone, but setting up each monitoring station seperately, it's easier/more practical then ever to customize the key words. Throw in some more computing power and you can have it do analysis on syntax and grammar to look for other words/phrasings/tells that would be indicative of suspicious behavior.
So no. I do not honestly believe that terrorists/bank robbers/suspicious militants/jaded wives don't use code words. I just have greater confidence in language analysis tools then you.
I have far more faith in humans doing the analysis. I've done the job. And I would be very, very surprised to see anything of the sort being worth field testing in less than 15 years. You'd need something so smart that it would almost be a self-aware AI.
As we saw on April 1st, 2010, it's quite easy to change a filter to include a broader selection of words. And if you're not monitoring everyone, but setting up each monitoring station seperately, it's easier/more practical then ever to customize the key words. Throw in some more computing power and you can have it do analysis on syntax and grammar to look for other words/phrasings/tells that would be indicative of suspicious behavior.
So no. I do not honestly believe that terrorists/bank robbers/suspicious militants/jaded wives don't use code words. I just have greater confidence in language analysis tools then you.
If you're only monitoring a selection of people then might as well just do a conventional wiretap.
I'm not sure just what you're arguing here as I've said nothing about replacing humans.
Yeah, because if I was a budget minded supervisor I wouldn't fire a bunch of people If I could get a machine to do most of the work.
But that aside, you are talking about a machine doing a lot of analytical work, and I just plain don't trust it. Machines are too easy to fool. I wish I could go into why I think that machines won't be ready for this anytime soon, but I can't. At least, not without getting really uncomfortable and wanting to consult with someone about how much I can and can't say.
I'm not sure just what you're arguing here as I've said nothing about replacing humans.
I have no idea what you're arguing.
Point 1 - to actually monitor everyone would require prohibitive quantities of storage and manpower based on current technology.
Point 2 - even with sophisticated filtering software the volume incoming would likely still be too great to handle for any kind of practical purposes.
Point 3 - if they narrow the ranged of observed down to a manageable level then the software would actually return something useful, but at that point you're not actually doing anything that we haven't been able to do since the 50's. We know that persons of interest are monitored. This isn't about select people being recorded, it's about everyone being monitored. Even though software would certainly help you monitor more people it's not going to get you anywhere close to "everyone" since the wider you have to set your filters (to account for slang and code) the more false-positives you'll return.
Your first post in the thread is basically "I don't know about monitoring everyone, but I think we could manage to record a bunch of people." Well, no ****. They could do that before anyone in Western Canada had colour TV.
I have no idea what you're arguing.
Well don't blame me, seeing as ...
Your first post in the thread is basically "I don't know about monitoring everyone, but I think we could manage to record a bunch of people." Well, no ****. They could do that before anyone in Western Canada had colour TV.
shows that you haven't actually read my first post in the thread.
That said... there is probably nothing a computer will ever do that wasn't done before a computer could do it. And if you think that's a good enough reason to not use a computer in a task, then you're a moron.
Yeah, because if I was a budget minded supervisor I wouldn't fire a bunch of people If I could get a machine to do most of the work.
More like use a machine so that the same bunch of people can spend their time with data more likely to generate a hit, but I see your point as well.
At least, not without getting really uncomfortable and wanting to consult with someone about how much I can and can't say.
Fair enough.
Sorry non-tech geek here; WTF is a PB??
A PetaByte. Big brother to the Terrabyte. One million GigaByte.
Sorry non-tech geek here; WTF is a PB??
A PetaByte. Big brother to the Terrabyte. One million GigaByte.
He's lying. It's really Peanut Butter. It's the creamiest data storage medium known to man.
Although there's a lot of replication there.
Guess it was to be expected for sattelites to come a long way since the original ones. I can see the advantages in a smaller sattelite. Just as I can see advantages in a bigger one...
They are watching us
Not really. It's like sitting on a park bench, or a bench in the local mall. Yeah, you're "watching" everyone's comings and goings, but it isn't exactly a big deal.
They are watching us
Not really. It's like sitting on a park bench, or a bench in the local mall. Yeah, you're "watching" everyone's comings and goings, but it isn't exactly a big deal.
Of course it is.
If you feel that way than I recommend never going outside your cave, Luddite.
I just find it confusing how people here complain about how the government is taking away freedoms, how now the police can go in a houses without a warrents, etc, etc, yet they are happy about satelites that can basically spy on them.
Ok... How is aerial photography spying? If you don't want to be seen outside your home, dont go out. Lemme know how that works for you. One of the key things to understand is just because they can doesn't mean they want to. Sort of like the people who claim that the government records every phone call everyone makes. Even if it did, where are the people to listen to all of these phone calls?
A quick search says that Americans make or recieve approximately 2.4 billion phone calls per day. If we half that number (because each conversation usually has two sides) that leaves us with 1.2. The averge length of a cell phone call, according to a quick search, is 3 minutes and 15 seconds. So, 3.25 minutes times 1.2 billion conversations... That's over 7400 years worth of audio. You'd need over 2.7 million people just listening to phone calls 24 hours a day, every day. If you assume 8 hour shifts, that turns into over 8.1 million people just to listen to phone calls. That doesn't count the payroll dept., human resources, managers, etc. The listeners alone would account for over 2% of the total US population. By contrast, there are fewer farmers in the US than that. And significantly fewer automechanics.
In other words, the manpower to watch everyone that the paranoid delusionists think the government does.. Well, that manpower just doesn't exist.
Filtering out the stuff you don't need so you need less people to search for the stuff you do want.
(No, I'm not serious. Neither was my other post)
Ok... How is aerial photography spying? If you don't want to be seen outside your home, dont go out. Lemme know how that works for you. One of the key things to understand is just because they can doesn't mean they want to. Sort of like the people who claim that the government records every phone call everyone makes. Even if it did, where are the people to listen to all of these phone calls?
A quick search says that Americans make or recieve approximately 2.4 billion phone calls per day. If we half that number (because each conversation usually has two sides) that leaves us with 1.2. The averge length of a cell phone call, according to a quick search, is 3 minutes and 15 seconds. So, 3.25 minutes times 1.2 billion conversations... That's over 7400 years worth of audio. You'd need over 2.7 million people just listening to phone calls 24 hours a day, every day. If you assume 8 hour shifts, that turns into over 8.1 million people just to listen to phone calls. That doesn't count the payroll dept., human resources, managers, etc. The listeners alone would account for over 2% of the total US population. By contrast, there are fewer farmers in the US than that. And significantly fewer automechanics.
In other words, the manpower to watch everyone that the paranoid delusionists think the government does.. Well, that manpower just doesn't exist.
I see a job creation opportunity!
Damn liberal! Those will be government jobs!
Only until the sector has grown enough to support competition, then we can privatize spying on the citizenry.
Edit: but remember that private industry would likely just outsource the whole operation.
Regarding listening to people's phone calls: Gimme twenty years, a department full of computer scientists and a budget. We'll have a super-cluster doing all that listening for us, with advanced linguistic algorithms flagging suspicious conversations for human review. The entire thing will eventually be a team of nerds keeping the computers going and a google browser.
Good, old-fashioned detective work is a lot more effective. Monitor the bad people, learn who the other bad people are, monitor the new bad people are, etc.
Note that I only talked about the manpower issue. I didn't even address the physical unlikelihood. (Enough storage space for 7400 years of audio on a daily basis?!)
Regarding listening to people's phone calls: Gimme twenty years, a department full of computer scientists and a budget. We'll have a super-cluster doing all that listening for us, with advanced linguistic algorithms flagging suspicious conversations for human review. The entire thing will eventually be a team of nerds keeping the computers going and a google browser.
Good, old-fashioned detective work is a lot more effective. Monitor the bad people, learn who the other bad people are, monitor the new bad people are, etc.
Note that I only talked about the manpower issue. I didn't even address the physical unlikelihood. (Enough storage space for 7400 years of audio on a daily basis?!)
If you're using listening algorythms then build in a 3-minute buffer, if something "suspect" gets flagged then it actually saves the buffered information and records the rest of the call for review by a fleshbag.
Even with a three minute buffer the storage requirements would be insane. The information I gave earlier is for call center recordings - you know, people evaluated for efficiency on-call. I'd say easily half of their calls average three minutes or less.
Oh, yeah, totally, you'd quickly be dealing with petabytes of audio.
For a consumer, a Petabyte of hard disk space costs about $40,000. I'm guessing a government or corporation could get it a lot cheaper.
Hell, Youtube uses (by estimation according to a google search) about 10 Petabytes of space to store all its movies. Which doesn't sound unreasonable.
Hard disk space is cheap, and it's only getting cheaper.
3TB is about double of what my home computer has plugged in, and it's a cheap ass machine with not even a very big harddrive.
For a consumer, a Petabyte of hard disk space costs about $40,000. I'm guessing a government or corporation could get it a lot cheaper.
Hell, Youtube uses (by estimation according to a google search) about 10 Petabytes of space to store all its movies. Which doesn't sound unreasonable.
Hard disk space is cheap, and it's only getting cheaper.
I don't think you are understanding the scale in question, though. He's talking about 3 TB being barely enough to handle 500 people's calls for 2 years. That's not two years of canstant audio, that's just two years of the call center being in operation. I'm talking about 7400 years worth of audio on a daily basis. When I say 7400 years worth of audio, I mean taking every second of audio and stringing it together into an unbroken stream that keeps going day and night for longer than recorded history. And that's what you are pulling in on a daily basis.
Besides; like Escher said... 20 years. 20 years ago a GB was considered enourmous... now the PB is commercially available. I don't think the storage is gonna be an issue in another 20 years.
It's what you're handling... the basic idea is that you're not storing all of it.
Besides; like Escher said... 20 years. 20 years ago a GB was considered enourmous... now the PB is commercially available. I don't think the storage is gonna be an issue in another 20 years.
In 20 years? Probably not. The probelm is that there are people who think we're doing it right now. The technology isn't there, the manpower isn't there.
Hrm... you know, I haven't actually heard of such a tool that the FBI/CIA/police uses, but it wouldn't surprise me. If nothing else they could use it in conjunction with a traditional wire tap to reduce the amount of manpower needed.
So something like recording the conversation waiting for keywords, if it finds 'em it keeps the conversation and forwards it to a human operator, if nothing suspicous is found, trashes it...
Wait... You honestly think they don't use code words? I mean, yeah, most terrorists are dumb as bricks, sure. But even they aren't that dumb. Seriously, just using common American slang I can make a drug transaction sound like I'm going to a clam bake.
So no. I do not honestly believe that terrorists/bank robbers/suspicious militants/jaded wives don't use code words. I just have greater confidence in language analysis tools then you.
As we saw on April 1st, 2010, it's quite easy to change a filter to include a broader selection of words. And if you're not monitoring everyone, but setting up each monitoring station seperately, it's easier/more practical then ever to customize the key words. Throw in some more computing power and you can have it do analysis on syntax and grammar to look for other words/phrasings/tells that would be indicative of suspicious behavior.
So no. I do not honestly believe that terrorists/bank robbers/suspicious militants/jaded wives don't use code words. I just have greater confidence in language analysis tools then you.
I have far more faith in humans doing the analysis. I've done the job. And I would be very, very surprised to see anything of the sort being worth field testing in less than 15 years. You'd need something so smart that it would almost be a self-aware AI.
As we saw on April 1st, 2010, it's quite easy to change a filter to include a broader selection of words. And if you're not monitoring everyone, but setting up each monitoring station seperately, it's easier/more practical then ever to customize the key words. Throw in some more computing power and you can have it do analysis on syntax and grammar to look for other words/phrasings/tells that would be indicative of suspicious behavior.
So no. I do not honestly believe that terrorists/bank robbers/suspicious militants/jaded wives don't use code words. I just have greater confidence in language analysis tools then you.
If you're only monitoring a selection of people then might as well just do a conventional wiretap.
I'm not sure just what you're arguing here as I've said nothing about replacing humans.
Yeah, because if I was a budget minded supervisor I wouldn't fire a bunch of people If I could get a machine to do most of the work.
But that aside, you are talking about a machine doing a lot of analytical work, and I just plain don't trust it. Machines are too easy to fool. I wish I could go into why I think that machines won't be ready for this anytime soon, but I can't. At least, not without getting really uncomfortable and wanting to consult with someone about how much I can and can't say.
I'm not sure just what you're arguing here as I've said nothing about replacing humans.
I have no idea what you're arguing.
Point 1 - to actually monitor everyone would require prohibitive quantities of storage and manpower based on current technology.
Point 2 - even with sophisticated filtering software the volume incoming would likely still be too great to handle for any kind of practical purposes.
Point 3 - if they narrow the ranged of observed down to a manageable level then the software would actually return something useful, but at that point you're not actually doing anything that we haven't been able to do since the 50's. We know that persons of interest are monitored. This isn't about select people being recorded, it's about everyone being monitored. Even though software would certainly help you monitor more people it's not going to get you anywhere close to "everyone" since the wider you have to set your filters (to account for slang and code) the more false-positives you'll return.
Your first post in the thread is basically "I don't know about monitoring everyone, but I think we could manage to record a bunch of people." Well, no ****. They could do that before anyone in Western Canada had colour TV.
I have no idea what you're arguing.
Well don't blame me, seeing as ...
Your first post in the thread is basically "I don't know about monitoring everyone, but I think we could manage to record a bunch of people." Well, no ****. They could do that before anyone in Western Canada had colour TV.
shows that you haven't actually read my first post in the thread.
That said... there is probably nothing a computer will ever do that wasn't done before a computer could do it. And if you think that's a good enough reason to not use a computer in a task, then you're a moron.
Yeah, because if I was a budget minded supervisor I wouldn't fire a bunch of people If I could get a machine to do most of the work.
More like use a machine so that the same bunch of people can spend their time with data more likely to generate a hit, but I see your point as well.
At least, not without getting really uncomfortable and wanting to consult with someone about how much I can and can't say.
Fair enough.
Sorry non-tech geek here; WTF is a PB??
A PetaByte. Big brother to the Terrabyte. One million GigaByte.
Sorry non-tech geek here; WTF is a PB??
A PetaByte. Big brother to the Terrabyte. One million GigaByte.
He's lying. It's really Peanut Butter. It's the creamiest data storage medium known to man.
Although there's a lot of replication there.