I had this same misgiving when Dallas PD used a robot to carry a bomb towards a hostile rooftop sniper.

I think it will usually be justified as being in the imminent defense of others / the public. The sniper was actively threatening to continue shooting people.

Perhaps they’d say the hypothetical guy shooting at a robot is endangering others via any stray bullets. And perhaps sometimes that would be correct.

I mean, the Dallas (ex-military) sniper was exactly the right use of this sort of technology. He was actively, explicitly murdering cops, and the police had no reason to believe he was unwilling or unable to kill more of them if they moved to arrest him. I can understand concern that this might be used in more ambiguous circumstances, but I don't understand why anyone would object to its use in that particular case.

Because at the exact moment the bomb was detonated he wasn’t a threat at that moment.

Maybe ten minutes before, perhaps ten minutes after.

But perhaps they could have cordoned off a wide area until he got hungry or thirsty enough to come down. That would have been a massive hassle for the city and difficult to ensure the city’s safety but you could always blow him up next time he starts shooting at someone.

On balance obviously it was expedient to blow him up at that moment but it’s not clear it was strictly necessary.

It's not clear to me that they would have had a better opportunity. "Lethal intervention" is among the best case scenarios in a mass shooting, I don't know why we are litigating it like the outcome was particularly tragic. It may not have been perfect, and I'm sure there's something to learn from it, but people are discussing it like it was Uvalde or something. It seems like it was as close to a desirable outcome as could reliably be secured--I'm glad police didn't roll the dice in hoping for a less lethal outcome with innocent lives hanging in the balance.

Good point. Though what if the suspect is threatening to shoot others, and there is a robot (with a live round) nearby?

I get that cops can't be trusted to only use it in this scenario (as implied by the author), but if hypothetical "better police officers" were available, should we allow it then?

EDIT: theoretically the robot (even if remote controlled) could aim better.

> the basic principle of: "you only kill when your life is in danger"

Emphasis mine. Is it not okay to shoot at someone who's shooting at your partner?

> it violates the basic principle of: "you only kill when your life is in danger"

For police, it's "you only kill when a life is in danger". It doesn't have to be your life.