AI Weapon Detection failure.
Antioch School Shooting in Nashville.
After a shooting at the Covenant School, a lot of places stepped up their efforts to fortify schools. This is something I've personally been an advocate for.
The problem is that we need to use proven strategies or, if we're going to rely on new technology, we need proven backups as well. One of those unproven technologies we've talked a lot about here at Bearing Arms is AI gun detection systems, such as those deployed on the New York City subway.
I'm just not convinced they're ready for primetime.
AI Gun Detection System Blows It In Nashville School Shooting – Bearing Arms
I had an experience with this system not too long ago. After going through a checkpoint in a public building, the system alerted security I had a gun on me.
I didn’t.
I went back later on that day and again I got buzzed, but it was the face of security that told the tale: They are getting to the point of despising and ignoring the systems. Another blatant screw-up I saw involved somebody in a wheelchair with some devices attached including an oxygen concentrator. They lit up the AI system as guns. The person was wanded, but you could tell by the attitude it was more of a “I was told to this crap after a beep.” than actually a security concern.
So, Tom Knighton is right: this thing is still not properly cooked.
GIGO Happened.
I've seen a version of this in the wild as well. I won't say where or when, but I will say, I was carrying at the time, in a state that does NOT require permits and "No Guns" signs do NOT carry force of law.
The computer flagged the photo, and I could see on the monitor that it circled where my firearm was located. But with the camera angled as it was, my wife's wrist -- with her watch -- was between it and my gun so it appeared at a casual glance that it flagged her watch (which it might have, there's no way to know for sure). I'm not even sure the person manning the "welcome" desk was watching the monitor closely enough to notice (which if the computer habitually flags on random crap, is neither surprising nor reassuring), but they waved us through. No wands, no pat-downs, no concerns.
If other locations treat the systems' warnings with that level of nonchalance, they're worse than useless. Useless because the employees don't always follow-up on warnings (because the warnings are usually crap), worse-than-useless because they provide a sense of security which people will trust but any rational person knows is 100% false.
Never heard of AI detection until now. I’m going to continue to believe that until someone invents a better human being, Natural Stupidity (NS) will beat AI.