I would fully disagree. We targeted civilians, not in war zones at functions like weddings, funerals and other explicitly civilian gatherings. We (the US) had the intent to kill civilians, and our tolerances for civilian casualties were an order of magnitude larger than what the IDF is using.
If anything it's not comparable because what we did was worse.
I mean, it probably has a neural network component.
DST vsm Standard time literally doesn't matter. It's the switching between the two that kills people.
This AI isn't a LLM.
This is almost certainly true. That's truthfully the only way to track humans en-mass.
What do you mean? There's a whole generation of Yemeni children afraid of Sunny days because those are the days that people die. I'm a place that gets 300+ days of sunshine a year.
We we're much worse.
So you're complaint is that people are telling you, "You have this percentage chance of this being reality" and then you're mad that they're unable to be more accurate? It's polling it's not fortune-telling.
There's no way Democratic leaders are truly this obtuse. Do they truly think that, "get over yourself" is a reasonable argument that will win them votes?
If I tell you that a rocket is going to land withing a 20ft circle 90% of the time and land 9 rockets in the circle and 1 out of it; was I accurate or inaccurate in your mind?
That would be such a poor move.
If a pill has a +- of 5-7 percent with 90% confidence. And you have ten polls, You would expect at least one to be off by more that 5-7%. What your describing is expected.
LLM's hallucinate all the time. The hallucination is the feature. Depending on how you design the neural network you can get an AI that doesn't hallucinate. LLM's have to do that, because they're mimicking human speech patterns and predicting one of my possible responses.
A model that tries to predict locations of people likely wouldn't work like that.