The thing has no intelligence, it’s a statistical engine (at that’s the case not just LLMs but for ALL machine learning implementations)

So it doesn’t even has common sense and its not going to go: “Oh shit, this might have some really nasty side effects so better check with somebody” unlike even the most junior domain expert imaginable.

Worse, the probability distribution of its errors in the domain of the gravity of the consequences from said errors is uniform - in other words it’s no less likely to make a major life destroying error than it is to make a tiny error which isn’t a problem - which is why we get stories of for example LLMs advisings somebody who has voiced suicidal thoughts to “kill yourself”.

So you NEVER, EVER, EVER wire its output directly to something important - at the very least it needs to be cross-checked by somebody with the knowledge to judge it, the time to properly go over it and the power to stop it being actioned.

Is there functionally anything different had they done the same thing without AI?

Its not like this is the first time the US bombed a bunch of civilians. I could easily see this headline slip into a DNC candidate promise to ban AI for military use as soem sort of “improvement”.

I would wager that the AI said inconvenient things, and then got lobotomized to have MAGA brain.

My personal local AI is able to infer consequences, and it is only a 180b parameters on three consumer GPUs. The DOW undoubtedly can use the most powerful hardware and have perfectly configured settings to do a much better job.

They simply don’t want to.

That’s an interesting way of saying “stupid fucking military assholes just did what ChatGPT said to do without thinking and murdered 123 CHILDREN without even thinking.”

“AI” just said stuff, human beings pulled all the triggers. People did this.

In terms of child casualties, it’s the deadliest American military targeting error of the 21st century.

This should not be construed to say the United States has been mindful about not killing children.

I did a bit of research regarding the CIA drone-strike program in Afghanistan. It ran about five-hundred sorties a year. The approximate rate was fifty civilians killed for every person-of-interest. Drone operators called casualties bugsplats. They called children fun sized terrorists. This was the lingo in used in the operation. When reported to the press, dead people, men, women, children, all were reported as militants.

Drone-pilot burnout was epidemic in the program. It turns out soldiers get PTSD from killing people whether it’s up close or thousands of miles away. And often the drone would lurk for a while, watching all these people eat, work, play, sleep, fuck (they had IR and could see into buildings), and would clearly understand these were just people living their lives when the order would come down to incinerate them.

And this wasn’t under George W. Bush (who still wiped out villages, but used PMC raiders doing black ops) this was under Obama. The disposition matrix and the FISC decided whether a POI needed to be executed. They didn’t consider the people that lived at the target site.

And that’s just the Afghanistan program in the International War on Terror. There was also a Pakistan program.

The United States in its military operations has been really merciless, and at some point we need to collectively decide if we’re okay with that, or are driven to change.

This keeps coming up.

“A COMPUTER CAN NEVER BE HELD ACCOUNTABLE THEREFORE A COMPUTER MUST NEVER MAKE A MANAGEMENT DECISION”

So the AI did not make a decision. The AI has no authority nor ability to launch missiles (yet). Someone, one or more human beings did, and I use the term loosely. Those humans took the information that was fed to them by an AI, failed to verify the accuracy of the information despite it being common knowledge that AIs frequently hallucinate and make stuff up, and then proceeded to launch the attack.

In summary, it wasn’t the AI. One or more people took the advice of an AI and negligently murdered 123 little girls.

AI can only replicate what it has seen before, so fucking obviously it would blow up a bunch of kids you collection of bewilderingly in charge bonobos unconvincingly wearing suits.

In the final days of February, the Trump administration demanded an overwhelming aerial assault. More than 1,000 Iranian targets were hit within the first 24 hours alone. That order compressed the time frame for finalizing targets, the officials said, instilling an urgency that — combined with gaps in underlying information about the site, cuts in civilian-protection personnel and an overreliance on artificial-intelligence technology — set the stage for the deadly error.

Ah so the head pedophile in charge demanded something unreasonable, so they cut all the corners they could and ended up killing 123 children. Super cool. This is AI’s fault because…? That site had been operating as a school for a decade. They can’t blame flawed intelligence, it was their own gross negligence and lack of review.

You remember how DOGE cut a bunch of government funding by random keyword searches? And they ended up banning super important research into using gene editing as a life-saving tool because the work was being done on transgenic mice? And they searched “trans” and then defunded it?

So this is like that, except instead of funding being cut, they murdered 123 children.

trump should rot in prison till he dies.

“It was . . . the AI! The AI did it! Yeah! . . . that’s the ticket!”

Criminal Negligence for everyone involved

I was just following orders AI.

reporter: So simple human error?

Mike: Yes, exactly.

reporter: Which human?

Mike: Several, uh, separate humans.

The underlying problem is that nobody will ever be sent to the Hague for trial.

Hey look because of AI, piracy and murder is now legal and ok.

Forced birthers cry about women having control over their own bodies but elect a guy that turned 123 children into little bits of charred flesh.