Lie Detecting AI Kiosks May Be The Future of Border Security

rgMekanic

[H]ard|News
Joined
May 13, 2013
Messages
4,302
CNBC is reporting that The U.S Department of Homeland Security funded research of a technology called the Automated Virtual Agent for Truth Assessments in Real-Time, or AVATAR, and about six years ago tested on volunteer travelers at the U.S.-Mexico border. According to one of the developers of the system AVATAR has a success rate of 60-75 percent in deception-detection. In the future, the AI powered kiosks could be used to provide initial screenings of people crossing the border, and flag people for human security to check out of it detects deception.

Very cool practical use of AI in my opinion, provided the accuracy is what the team claims it is. This coupled with face recognition could really speed up the process of border crossings in the future. Little try-hard on the name tho.

"The technology has much broader applications potentially," despite most of the funding for the original work coming primarily from the Defense or Homeland Security departments a decade ago, according to Aaron Elkins, one of the developers of the system and an assistant professor at the San Diego State University director of its Artificial Intelligence Lab. He added that AVATAR is not a commercial product yet but could be also used in human resources for screening.
 

JosiahBradley

[H]ard|Gawd
Joined
Mar 19, 2006
Messages
1,787
Sigh what could possibly go wrong. 60-75% accuracy is as good as an educated guess. God I hate "AI" in its current state and applications. Sure keep trying to make it better but fuck no to production environments.
 

ruffbytes

Limp Gawd
Joined
Oct 22, 2015
Messages
447
There is a danger to this where agents will trust the kiosk more than their own eyes and hassle folks who are just nervous at border crossings, etc.

But if the line moves faster and agents think of this as a tool, but not absolute truth, it sounds pretty good to me.
 

WhoMe

Gawd
Joined
Jan 3, 2018
Messages
827
Next step it'll zap you with 10,000V (and a few mA) if it thinks you're lying...as an added incentive to be truthful and only 40% will be falsely shocked. The real bad guys will be taught how to beat it (keep your eyes focused on one spot etc.).
 

sadsteve

Gawd
Joined
Oct 1, 2010
Messages
565
Next step it'll zap you with 10,000V (and a few mA) if it thinks you're lying...as an added incentive to be truthful and only 40% will be falsely shocked. The real bad guys will be taught how to beat it (keep your eyes focused on one spot etc.).
Heh, and I can tell you by personal experience you don't want to be shocked by high voltage. Years ago I got zapped by a 20KV Helium Neon laser power supply, first time I could actually taste the metal fillings in my teeth. Yes, they taste bad.
 

katanaD

[H]ard|Gawd
Joined
Nov 15, 2016
Messages
1,987
How soon before the AI is accused of profiling?

I think one thing that is going to be an issue, is we will not be able to review the reasons its uses for its conclusions. It will be kept secret, which is for me an issue for anything used in a law enforcement capacity.
 

Master_shake_

[H]F Junkie
Joined
Apr 9, 2012
Messages
11,175
only a moron would lie at a border crossing.

they flag your shit and you can be banned for life.

the truth will set you free.
 

Seventyfive

[H]ard|Gawd
Joined
Jul 14, 2004
Messages
1,347
We need actual competent people working the border. I have a friend who went to Israel with his brother's wife (the rest of the family was already in the country) and they both got pulled into different rooms and questioned. They said "this isn't the person you normally travel to Israel with, who is this other person? Where is the other person you normally travel with?" Once everything was cleared up they were free to go. And to be clear, both of them are Israeli citizens. This is the level that their version of the TSA operates at. Compare that to our TSA? We're not even in the same galaxy.
 

Babbster

[H]ard|Gawd
Joined
Jan 13, 2006
Messages
1,434
Sigh what could possibly go wrong. 60-75% accuracy is as good as an educated guess. God I hate "AI" in its current state and applications. Sure keep trying to make it better but fuck no to production environments.
Don't blame AI for lie detection inaccuracy. That's a function of lie detection being a straight-up pseudoscience, and you can bet that the truth is closer to 50-60% than 60-75%. It's all coin-flipping bullshit to which only government agencies lend any credence.
 

steakman1971

2[H]4U
Joined
Nov 22, 2005
Messages
2,433
Wasn't there a movie where someone had to speak to a kiosk? Might be thinking of 2001 Space Odyssey.
 
Joined
Aug 3, 2017
Messages
794
We need actual competent people working the border. I have a friend who went to Israel with his brother's wife (the rest of the family was already in the country) and they both got pulled into different rooms and questioned. They said "this isn't the person you normally travel to Israel with, who is this other person? Where is the other person you normally travel with?" Once everything was cleared up they were free to go. And to be clear, both of them are Israeli citizens. This is the level that their version of the TSA operates at. Compare that to our TSA? We're not even in the same galaxy.
well, Israel is in a state of perpetual war , gobbling up enemies' lands every year. So whats the surprise...
 

Olle P

Limp Gawd
Joined
Mar 29, 2010
Messages
331
..., provided the accuracy is what the team claims it is. This coupled with face recognition could really speed up the process...
You mean the face recognition used by British police that has ~95% false positives?
Seems pretty useless to me.

Better yet, give them a US history quiz & ask them what skills they posses.
Why would you want that from tourists?
 
Top