I'm a big fan of the EFF but this article basically comes down to "X Y and Z were designed to solve problem A and they were implemented poorly therefore problem A is unsolvable and no one should even try to address problem A"
This is especially ridiculous when the failures of X Y and Z came from administration issues (failures of oversight), picking shitty contractors, buying faulty equipment, etc. all of which are solvable/preventable.
The takeaway of these past failures shouldn't be that securing the border is impossible and not worth even attempting. The takeaway should be that programs need to be meaningfully and intelligently invested in (maybe going with the lowest bidder or your personal friends/donors isn't always a good idea), and that there needs to be oversight and accountability to make sure that those funds aren't being wasted or pocketed by corrupt public servants and private contractors to ensure that systems are implemented correctly and maintained.
Of course it's going to take "record-level funding" to implement a massive solution when previous attempts were entirely half-assed, designed to attract and allow for corruption, and then neglected. Congressional leaders and the American public should be shocked and outraged at the money that's been wasted and they should be working to design a system that avoids those pitfalls and actually does the job right. Ideally we should also be tracking down the people responsible for those past failures and holding them accountable too where it's possible.
> Surveillance at the U.S.-Mexico border is a wasteful endeavor that is ill-equipped to respond to an ill-defined problem.
This article cites efforts of doing this all the way back to 1997. The critical problem is not that the cameras are literally broken, although they seem to have struggled with this, but the fact that the cameras have not shown to improve operational results.
An intelligently designed program would try to learn from these decades of trying the same thing unsuccessfully and try something else.
Before confidently claiming that the problem was the quality of the execution, you should try to present some evidence that this solution is a sensible one. Claiming it would work if it was better is something you can do indefinitely with no evidence.
The fact that past programs failed to collect good data or failed to make good use of the data they were getting does not imply that having data is unlikely to improve operational results.
I think it's plainly obvious that if your goal is to stop people from illegally crossing the border, then monitoring that border is a necessity. We could hire humans to stand all day in watchtowers and scan for people sneaking in with just their human eyes, but (working) cameras have a lot of obvious advantages.
There is plenty of evidence and data showing that cameras can be a highly effective part of solutions to the problem of monitoring. There are countless companies, governments, and individuals using cameras right now with great success. Cameras are not an unproven technology and their usefulness is not theoretical.
I don't think that anyone would be opposed to a better option that would eliminate the need for cameras and sensors, but if one exists I've never heard of it. Cameras and sensors seem to be the best options we have. The cameras and sensors (when they were functioning) were never the problem with past systems. Those past problems were more like "We paid for a bunch of cameras but we never budgeted for anyone to respond to them, or to maintain them, or to maintain the systems we need to collect data from them". Those are all very solvable problems which shouldn't have happened in the first place.
In the interests of optimization: maybe we could just not? Seems like the Occam solution here would just be to (1) rigorously enforce employment eligibility for all employers (not employees!), something that is done routinely at the salaried level already and has an existing bureaucracy in place and (2) hand out green cards like candy for people working positions for which there are openings. Problem solved, QED. No border towers needed, and "illegal" migration stops at exactly the level where it's required, because migrants obviously won't come in excess of the jobs available.
Obviously that's not the actual problem you're trying to solve, because the actual problem is political and ethnographic (c.f. Musk's thread that this will be the "last election" because of something to do with demographic change). But this is HN, and it seems like it's worth discussing actual solutions.
The illegal immigration problem can't be solved purely by making sure that anyone coming into the country illegally cannot be employed. Too many jobs are paid under the table and while almost all of the people entering illegally today are economic migrants looking for higher paying jobs, in the very near future we'll be looking at floods of climate refugees who literally won't have homes to go back to. Having otherwise open borders or just handing out green cards to anyone who asks also fails to address security concerns. No matter what, we need to be able to monitor and regulate who is coming in to the country and what they are bringing in with them (drugs, weapons, human trafficking victims, etc).
I agree that we should be strongly enforcing employment eligibility as well and that it would do a lot of good, but doing that won't mean that we can just leave our borders open and unmonitored.
This article is a little schizophrenic. It calls the cameras "wasteful" and "snake oil", yet the referenced NBC article calls them "a crucial tool". It quotes reports about failures of old systems from Boeing and General Dynamics, then implies without evidence that the same criticisms apply to the newer Anduril "AST" systems. There is no allegation that the AST systems are broken at all.
In the NBC article, "crucial tool" is a non-technical phrase by the article writer, it is not the memo. [1]
The EFF article's expressed intent is first, that there's a history that goes beyond the NBC report. Then, that the history shows reports written by various governmental & non-governmental entities that the tools are not effective.
I'm genuinely curious why that felt schizophrenic. ex. to me, even if "crucial tool" was in a government memo, I'm not sure why the EFF would be schizophrenic for disputing that.
[1] "Nearly one-third of the cameras in the Border Patrol’s primary surveillance system along the southern U.S. border are not working, according to an internal agency memo sent in early October, depriving border agents of a crucial tool in combating illegal migrant crossings."
> I'm not sure why the EFF would be schizophrenic for disputing that.
They don't dispute that. That's what's so weird. The argument boils down to "historic border surveillance tools have been pork-barrel debacles, therefore all border surveillance tools are bad," but the article doesn't even manage to draw a line from the past, cancelled programs to the current program.
One third of cameras not working might mean that there is no camera coverage (because the only working cameras are all pointed at a single storage shed), or it might mean that there is 100% coverage (because of redundancies in coverage.)
Its a data point, but without further context, you can't draw meaningful conclusions.
If I were to take a wild guess, one third of all security cameras in the world are probably not fit for purpose, and yet, the world keeps adding more and more of them. Smile, comrade.
Net migration between the US and Mexico is small. Pew research numbers:
An estimated 870,000 Mexican migrants came to the U.S. between 2013 and 2018, while an estimated 710,000 left the U.S. for Mexico during that period. That translates to net migration of about 160,000 people from Mexico to the U.S., according to government data from both countries.
That's 160,000 net in-migration from Mexico over 5 years. How much would you spend to bring that to zero?
You might think a bunch of tech people would profile performance before deciding what to optimize.
If you mean people that lived in Mexico, that may be. If you mean people that crossed from Mexico into the US across the Southern Border, those numbers are dwarfed. These are just the numbers that we know for sure from observation or interaction[0]
Sounds like a job for the NSA. With their surveillance apparatus, data lake, and analysis tools, they'd be able to make quick work of tracking and apprehending illegal border entries...
That would be a much better way to spend their time and money than invading the privacy of actual US citizens.
This article complains a lot about previous-generation, cancelled projects, but doesn't really investigate what's going on with the current surveillance towers or investigate anything at all, really. The linked NBC article explains that the FAA (WTF... yes, the airplane FAA) administer some surveillance towers, and that border patrol agents are mad that there's a big ticket backlog of broken ones to fix. That's pretty much it.
Nobody investigated enough to figure out things like:
* Why the FAA administer the towers and what the actual hold-up is towards getting a fix? Certainly at least the backstory should be public information.
* Why the towers are broken. This is probably sensitive information but I'm sure some of the disgruntled border patrol agents would be willing to have a chat about it.
* Is it a specific generation of tower that's broken? Is it some kind of backend issue, or just rot from deploying electronics into a hostile desert environment full of people trying to destroy them?
* How do the new "AI" towers work? They're probably just drawing boxes around people and items, no?
This is a disappointing and silly article, in my opinion. It doesn't convince me at all that border surveillance overall is a bad idea or a waste of money, just that some old programs turned into pork-barrel debacles. There's no fresh information or anything that would convince me either way on this issue.
They (Congress? GAO? DHS? ) should have a separate department certify and monitor these sensors. Make it this department's sole job to keep this sensor network running properly.
Readit News
This is especially ridiculous when the failures of X Y and Z came from administration issues (failures of oversight), picking shitty contractors, buying faulty equipment, etc. all of which are solvable/preventable.
The takeaway of these past failures shouldn't be that securing the border is impossible and not worth even attempting. The takeaway should be that programs need to be meaningfully and intelligently invested in (maybe going with the lowest bidder or your personal friends/donors isn't always a good idea), and that there needs to be oversight and accountability to make sure that those funds aren't being wasted or pocketed by corrupt public servants and private contractors to ensure that systems are implemented correctly and maintained.
Of course it's going to take "record-level funding" to implement a massive solution when previous attempts were entirely half-assed, designed to attract and allow for corruption, and then neglected. Congressional leaders and the American public should be shocked and outraged at the money that's been wasted and they should be working to design a system that avoids those pitfalls and actually does the job right. Ideally we should also be tracking down the people responsible for those past failures and holding them accountable too where it's possible.
> Surveillance at the U.S.-Mexico border is a wasteful endeavor that is ill-equipped to respond to an ill-defined problem.
This article cites efforts of doing this all the way back to 1997. The critical problem is not that the cameras are literally broken, although they seem to have struggled with this, but the fact that the cameras have not shown to improve operational results.
An intelligently designed program would try to learn from these decades of trying the same thing unsuccessfully and try something else.
Before confidently claiming that the problem was the quality of the execution, you should try to present some evidence that this solution is a sensible one. Claiming it would work if it was better is something you can do indefinitely with no evidence.
I think it's plainly obvious that if your goal is to stop people from illegally crossing the border, then monitoring that border is a necessity. We could hire humans to stand all day in watchtowers and scan for people sneaking in with just their human eyes, but (working) cameras have a lot of obvious advantages.
There is plenty of evidence and data showing that cameras can be a highly effective part of solutions to the problem of monitoring. There are countless companies, governments, and individuals using cameras right now with great success. Cameras are not an unproven technology and their usefulness is not theoretical.
I don't think that anyone would be opposed to a better option that would eliminate the need for cameras and sensors, but if one exists I've never heard of it. Cameras and sensors seem to be the best options we have. The cameras and sensors (when they were functioning) were never the problem with past systems. Those past problems were more like "We paid for a bunch of cameras but we never budgeted for anyone to respond to them, or to maintain them, or to maintain the systems we need to collect data from them". Those are all very solvable problems which shouldn't have happened in the first place.
The problem is unauthorized border crossings. What about that is ill defined?
Obviously that's not the actual problem you're trying to solve, because the actual problem is political and ethnographic (c.f. Musk's thread that this will be the "last election" because of something to do with demographic change). But this is HN, and it seems like it's worth discussing actual solutions.
I agree that we should be strongly enforcing employment eligibility as well and that it would do a lot of good, but doing that won't mean that we can just leave our borders open and unmonitored.
The EFF article's expressed intent is first, that there's a history that goes beyond the NBC report. Then, that the history shows reports written by various governmental & non-governmental entities that the tools are not effective.
I'm genuinely curious why that felt schizophrenic. ex. to me, even if "crucial tool" was in a government memo, I'm not sure why the EFF would be schizophrenic for disputing that.
[1] "Nearly one-third of the cameras in the Border Patrol’s primary surveillance system along the southern U.S. border are not working, according to an internal agency memo sent in early October, depriving border agents of a crucial tool in combating illegal migrant crossings."
They don't dispute that. That's what's so weird. The argument boils down to "historic border surveillance tools have been pork-barrel debacles, therefore all border surveillance tools are bad," but the article doesn't even manage to draw a line from the past, cancelled programs to the current program.
Its a data point, but without further context, you can't draw meaningful conclusions.
If I were to take a wild guess, one third of all security cameras in the world are probably not fit for purpose, and yet, the world keeps adding more and more of them. Smile, comrade.
An estimated 870,000 Mexican migrants came to the U.S. between 2013 and 2018, while an estimated 710,000 left the U.S. for Mexico during that period. That translates to net migration of about 160,000 people from Mexico to the U.S., according to government data from both countries.
That's 160,000 net in-migration from Mexico over 5 years. How much would you spend to bring that to zero?
You might think a bunch of tech people would profile performance before deciding what to optimize.
[0]https://www.cbp.gov/newsroom/stats/southwest-land-border-enc...
That would be a much better way to spend their time and money than invading the privacy of actual US citizens.
Dead Comment
Nobody investigated enough to figure out things like:
* Why the FAA administer the towers and what the actual hold-up is towards getting a fix? Certainly at least the backstory should be public information.
* Why the towers are broken. This is probably sensitive information but I'm sure some of the disgruntled border patrol agents would be willing to have a chat about it.
* Is it a specific generation of tower that's broken? Is it some kind of backend issue, or just rot from deploying electronics into a hostile desert environment full of people trying to destroy them?
* How do the new "AI" towers work? They're probably just drawing boxes around people and items, no?
This is a disappointing and silly article, in my opinion. It doesn't convince me at all that border surveillance overall is a bad idea or a waste of money, just that some old programs turned into pork-barrel debacles. There's no fresh information or anything that would convince me either way on this issue.
Dead Comment