3 ms·

As long as false flags get sent to a human gate agent standing right there who can make manual approvals, I fail to see why anyone traveling internationally wou

As long as false flags get sent to a human gate agent standing right there who can make manual approvals, I fail to see why anyone traveling internationally would object to this. I sure as heck want everyone else on the plane with me to be compared against their photograph. From a legal standpoint, domestic travel is different (and there should be open debate in Congress before they make a decision there). But I fully support facial recognition for international travel and other immigration benefits.

wahern 8y ago
I sure as heck want everyone else on the plane with me to be compared against their photograph. Why? I'm sure there's nefarious activity that could be prevented, I'm just struggling to identify activities that would make be feel (let alone actually be) unsafe on an airplane as a practical matter, and without doning a screen writer's hat. I'm assuming weapons are already adequately checked. And if not I'm not sure how facial recognition would appreciably help in this regard. Sure, maybe 12 bad guys enter the system though 12 airports with lax security, then all board the same flight at a major hub in a manner that circumvents analytics that searches for suspected bad guys flying together. But that already makes some dubious assumptions, such as the degree of efficacy in the system looking for suspected bad guys traveling together. For most of air travel hijackings and bombings were incredibly common. We all but solved that by the 1980s by physically screening for weapons and suspect material, not by pretending we could predict people's behavior.
- diebeforei485 8y ago
  IME questions like this typically come from people who don't deal with the bureaucracy of applying for visas very often, if at all. You assume fear of being bombed or hijacked, but it's really to make international travel easier and more efficient. Lack of a reliable way to verify identity is a major reason why applicants for admission (or applicants for a visa) are assumed "guilty until proven innocent", and a ton of people are denied visas (or denied entry) every day on the slightest reasons as a result. Being able to verify someone identity biometrically makes international travel and immigration more trustworthy, which makes it politically more palatable to make the lives of travelers and immigrants easier. And doing this on outbound flights is no different from doing it on inbound flights - from the host government's perspective, they need to know when people enter and leave.
  - wahern 8y ago
    But this has nothing to do with ports of entry or consular offices. It has to do with boarding at the gate. 99% accuracy is not remotely accurate enough to even begin replacing personnel at sensitive entry points. And consider that to the extent you automate security you permit automating measures to subvert the system. The reported accuracy of these solutions describe efficacy for typical use cases; they're not a measure of resilience to attack, which is what people often assume. Again, I'm not saying these things don't have their uses. Security just isn't one of them unless they're merely part of a larger, cohesive system (e.g. used to verify efficacy of auditing and reporting systems). In fact, they may decrease security if you rely on them as strong security measures.
    - diebeforei485 8y ago
      Outbound and inbound verification are sides of the same coin, as you need both in order to reliably track overstays. Congress has recognized this since 1996 when they first passed legislation requiring biometric entry-exit, and multiple times since then. The exit part of that has never really been implemented except for small trials. It really isn't about replacing personnel (that another strawman you've made up, after the strawman about weapons). Rather - instead of humans (TSA agent + gate agent), it's now a human and a machine. It helps that the kind of mismatches made by CNN's are generally different from those made by humans. > And consider that to the extent you automate security you permit automating measures to subvert the system. I'm not sure what you mean by automating measures to subvert the system - there are financial and physical barriers (you'd need to get a hundred people or more into an airport, and buy them all plane tickets) If you have that level of resources, you're better off bribing CBP officers[1]. [1] https://www.businessinsider.com/customs-border-agents-mexican-cartel-smuggling-2016-11 https://www.businessinsider.com/customs-border-agents-mexica...
bacon_waffle 8y ago
> I fail to see why anyone traveling internationally would object to this. If I'm understanding correctly, the new system is implemented on the departing side of the flight, and done (at least partly) in the name of security. So, to support this system, even more effort and equipment would be required in foreign airports for flights to the US.

wahern 8y ago

I sure as heck want everyone else on the plane with me to be compared against their photograph. Why? I'm sure there's nefarious activity that could be prevented, I'm just struggling to identify activities that would make be feel (let alone actually be) unsafe on an airplane as a practical matter, and without doning a screen writer's hat. I'm assuming weapons are already adequately checked. And if not I'm not sure how facial recognition would appreciably help in this regard. Sure, maybe 12 bad guys enter the system though 12 airports with lax security, then all board the same flight at a major hub in a manner that circumvents analytics that searches for suspected bad guys flying together. But that already makes some dubious assumptions, such as the degree of efficacy in the system looking for suspected bad guys traveling together. For most of air travel hijackings and bombings were incredibly common. We all but solved that by the 1980s by physically screening for weapons and suspect material, not by pretending we could predict people's behavior.

diebeforei485 8y ago
IME questions like this typically come from people who don't deal with the bureaucracy of applying for visas very often, if at all. You assume fear of being bombed or hijacked, but it's really to make international travel easier and more efficient. Lack of a reliable way to verify identity is a major reason why applicants for admission (or applicants for a visa) are assumed "guilty until proven innocent", and a ton of people are denied visas (or denied entry) every day on the slightest reasons as a result. Being able to verify someone identity biometrically makes international travel and immigration more trustworthy, which makes it politically more palatable to make the lives of travelers and immigrants easier. And doing this on outbound flights is no different from doing it on inbound flights - from the host government's perspective, they need to know when people enter and leave.
- wahern 8y ago
  But this has nothing to do with ports of entry or consular offices. It has to do with boarding at the gate. 99% accuracy is not remotely accurate enough to even begin replacing personnel at sensitive entry points. And consider that to the extent you automate security you permit automating measures to subvert the system. The reported accuracy of these solutions describe efficacy for typical use cases; they're not a measure of resilience to attack, which is what people often assume. Again, I'm not saying these things don't have their uses. Security just isn't one of them unless they're merely part of a larger, cohesive system (e.g. used to verify efficacy of auditing and reporting systems). In fact, they may decrease security if you rely on them as strong security measures.
  - diebeforei485 8y ago
    Outbound and inbound verification are sides of the same coin, as you need both in order to reliably track overstays. Congress has recognized this since 1996 when they first passed legislation requiring biometric entry-exit, and multiple times since then. The exit part of that has never really been implemented except for small trials. It really isn't about replacing personnel (that another strawman you've made up, after the strawman about weapons). Rather - instead of humans (TSA agent + gate agent), it's now a human and a machine. It helps that the kind of mismatches made by CNN's are generally different from those made by humans. > And consider that to the extent you automate security you permit automating measures to subvert the system. I'm not sure what you mean by automating measures to subvert the system - there are financial and physical barriers (you'd need to get a hundred people or more into an airport, and buy them all plane tickets) If you have that level of resources, you're better off bribing CBP officers[1]. [1] https://www.businessinsider.com/customs-border-agents-mexican-cartel-smuggling-2016-11 https://www.businessinsider.com/customs-border-agents-mexica...

bacon_waffle 8y ago

> I fail to see why anyone traveling internationally would object to this. If I'm understanding correctly, the new system is implemented on the departing side of the flight, and done (at least partly) in the name of security. So, to support this system, even more effort and equipment would be required in foreign airports for flights to the US.