Armed police handcuff teen after AI mistakes chip bag for gun in Baltimore

> It said the AI alert was sent to human reviewers who found no threat - but the principal missed this and contacted the school's safety team, who ultimately called the police.

So the company AI reviewers correctly determined that there was no actual gun, and the police responded appropriately to a report by a person (the school principal) who claimed to have seen this student with a gun. The question then is what the heck is this principal doing? Why do they have access to this pre-verified AI information, and why are they going off and calling the police with this information before doing any verification themselves?

Well also, none of the coverage includes what was said to the dispatcher, but maybe they screwed up too - I would expect a dispatcher to question the caller in more detail about what's going on. Like, what kind of gun was it, is it in their hand or in a pocket or what, did they point it at anyone or threaten anyone with it or are they just carrying it around, etc, and such information could be used for a more appropriate police response.

I'm assuming you're not in the US from your comment, so I encourage you to think more broadly about the problem rather than focusing on the headlines.

To play devil's advocate, this case exists within a larger system in the US. In particular:

1) the proliferation of guns in America causes tremendous uncertainty for police, for whom any encounter with the public could involve a firearm. In other countries where the public have fewer firearms, police don't have to escalate a situation as quickly for safety reasons because they can assume the person they're interacting with doesn't have a gun.

1a) therefore, police are trained to take control of situations in certain ways if they believe a threat to be present: for example, drawing their gun and handcuffing a potential suspect out of the abundance of caution, then performing their investigation once the scene is safe. That is what appeared to have happened here.

2) the risk of school shootings putting pressure on police, school systems, etc to prevent these situations. Hence, the deployment of technology, armed guards at schools, etc

2a) Sadly, local police often bear the brunt of larger system failures. Example: mental health treatment difficult to get? Police will have to deal with people who are not taking their anti-psychotic medications, or who are at risk of harming themselves or others. Proliferation of guns for political reasons? Police will have to encounter more people with guns.

3) Police are human beings. And even in the best case, with incredible amounts of training (which they may or may not have), they will still have to make split second decisions that could mean life or death. Some of those may be mistakes.

Combine all these forces, and you have a recipe for incidents.

It's absolutely fair to criticize behavior and hold law enforcement to a high standard, but man... I think we haven't put them in a position to succeed, and we ask a tremendous amount from them.

It's the same way if there's an incident in a production software environment, my general thought is not: "the bar for being a developer here must be really low because they're incompetent." My thought starts with, "what in our system allowed this event to happen?"

https://theonion.com/how-to-join-ice/

Including these gems:

* Be born with something just…missing

* Undergo background check confirming at least one prior arrest for a violent crime

How do we get the people who already doing this in businesses to stop doing it, if they see it as "saving time" and currently don't have qualms with it?