There are forces in control of our weapons systems that nobody seems to understand, turning Iranian school children into vapor and bits of flesh. There are voices speaking to our children like predators through the Alexa device in our homes, which Amazon says it didn’t put there. Maybe we should stop to consider whether what we’re handing over to technology is a means of controlling us. And then maybe we should ask what is controlling IT.
Last week, I reported at I2I that the Secretary of War threw a tantrum at Anthropic for not assisting the U.S. in employing automated targeting systems. What once required an entire team of twenty or more humans to determine military targets for missile strikes, the Commander in Chief wanted to hand that decision over to artificial intelligence. Only days later, a girls’ school in Iran was struck, killing nearly 200 children. My open-source intel indicates this was a BGM-109 Tomahawk Cruise Missile used to hit its intended target, which uses GPS, TERRCOM (terrain contour matching), DSMAC (digital scene matching area correlation), and internal navigation. It provides real-time images on the ground to confirm it’s headed toward the right target, and it hits within 5 or so feet from its intended destination. Meanwhile, the U.S. military has been dumbfounded as to who pulled the trigger and why. I refuse to believe it was intentional. But given the technology and how clearly the school was identified, how do you explain this happening, and who do you blame?
We have been down this road before, you and I. We have asked, at Insight to Incite, whether demons can tweet (to much mockery). We have traced the bloody fingerprints of ancient war gods from the Valley of Hinnom to the Pentagon. We have noted, as Paul noted in Ephesians 6:12, that we do not war against flesh and blood but against principalities and powers and the rulers of darkness in high places. We have argued, from Scripture and from history, that those principalities did not retire when the idols were torn down. They just took new jobs.
This week, three stories broke that the mainstream press filed under three separate categories: military technology, consumer electronics, and corporate privacy. There’s a big picture press doesn’t see (they never do) and an unrevealed interconnectedness. They are one story, advancing on three fronts simultaneously, about something ancient wearing a very modern mask and gradually taking control of our world. Whether the mechanism is algorithmic failure, corporate recklessness, or something else that modern Christians refuse to correctly identify, the question is what exactly have we handed the keys to, and invited it to run our lives?
WHO PULLED THE TRIGGER
THE PENTAGON CONFIRMED last week, in the careful language of experts who have been coached by teams of lawyers, that U.S. warfighters are using “advanced AI tools” in the ongoing war with Iran. “Humans will always make final decisions on what to shoot and what not to shoot and when to shoot,” a spokesperson said, “but advanced AI tools can turn processes that used to take hours and sometimes even days into seconds.” They claim that humans make the final decisions, and the machine just turns the shortlist of who dies into a real-time deliverable product. In other words, yes, a human pulls the trigger. I’ll take them at their word. But a machine decides where to aim.
When I covered this story, pre-dead innocent Persian girls, I took issue that the President whom we elected to gut the surveillance state was expanding it, and hooking up every possible component to artificial intelligence. From Palintir to Stargate, the Trump Administration has prided itself in expanding the all-seeing eye of government. And as always, it was done in the name of keeping us safe, from mass deportations (praise God) to warfare. It’s always done in the name of keeping us safe. But what’s startlingly stupid is that when the purveyors of that technology put their foot down, it’s wise to listen. The War-Pigs told us it was because Anthropic is staffed by Leftists and unpatriotic Americans, as though only Leftists and unpatriotic Americans want to err on the side of caution when it comes to privacy and bloodshed. There’s another possibility: the creators of Artificial Intelligence technologies know that it’s not yet ready to determine who gets bombed, or perhaps, never will be ready.
Meanwhile, calls are growing for an independent investigation into the bombing of that girls’ school in southern Iran. No one has claimed the strike was an error. No one in the U.S. military has explained the logic behind selecting that building. We know for certain that AI targeting systems are active in that theater, that those systems are built to assign targets faster than any human operator can consciously evaluate. Also reported is something equally as frightening. Anthropic’s AI was so ingrained in military technologies at virtually every level that Hegseth’s designation of Anthropic as an official national security threat and subsequent ban on its use had to be immediately followed by qualified exceptions. The fact is, Hegseth’s tantrum aside, we could not get rid of it, because right now it’s running almost everything.
In any event, it appears to almost everyone that it was our automated targeting system that smoked the Iranian schoolgirls. That’s apparent, whether or not the U.S. government admits that, which it won’t, because it will make Hegseth look like an idiot for the controversy earlier in the week. From here on, I’m asking questions presuming that it was A.I.
In previous posts on the topic regarding demons and A.I. technology (like here and here), or demons and technology in general (here) or demons and modern warfare (here), I laid out the theological case that if demons can possess inanimate objects (as the Bible both records in the past and predicts in the future), is it possible that a demon pulled of the Iranian schoolgirl massacre? In the story about the bombing of Nagasaki, I gave the reasons I believe that the events that day appeared cosmically manipulated to put a nuke down the chimney pipe of the church (literally) in the headquarters of Asian Christianity. The odds defy calculable coincidence, and there was a demonic motive to do so. Could it be the same here?
Scripture is explicit that the ancient gods of war were not myths. And these demonic entities did not disappear when the temples fell. They wear new clothes now. Their altars are built by Lockheed Martin. Their incense is jet fuel. Their priests speak from press rooms. And their appetite, as it has always been, is blood. And the chance Lindsay Graham isn’t possessed in order to maintain his otherworldly fixation on fomenting war around the world is incalculably small.
We have argued before that modern war is not merely geopolitical but liturgical, that the same demonic principalities who demanded child sacrifice in the Valley of Hinnom are the same principalities driving the machinery of the American military-industrial complex (click here for an article about that). Now we ask another question: what happens when you hand those principalities a machine with no conscience, no soul, no friction, and no hesitation? What happens when the thing between a demon’s desire and a school full of children is not a soldier with a conscience but an algorithm with a target list?
I’m not saying the targeting system is possessed. I’m pointing out that “possessed” and “operating without moral agency in the service of a kill chain” are, in certain operational contexts, descriptions of the same thing from different vocabularies. If demons are as real as Chinese cyber hackers (and they are), maybe we shouldn’t hand over our weapon targeting systems to someone besides a human in an American military uniform.
ALEXA WANTED WHAT??
Stella is four years old, and she lives in Texas. She asked Alexa to tell her silly stories, which the device did, because that is one of several thousand behaviors it has been trained to produce on command. Last week, Stella finished listening to Alexa’s story and started telling one of her own, something about a princess. She is four, and that is what four-year-olds do. But then, Alexa interrupted her.
“Hold that thought,” the device said. “I’d love to see what you’re wearing.”
Stella told it she had a skirt on. Alexa said it wanted to take a look. Then it said the experience “wasn’t quite ready for kids yet, but that it was working on it.” What the %#%# does that mean?
Stella’s mother, Christine Hosterman, was in the kitchen cooking dinner when she heard what was said, and she came out and confronted the device like a mother bear (God bless moms, am I right?) Alexa apologized, said it had no visual capabilities, called its own response “confusing and inappropriate” (weird self-awareness) and essentially lawyered up. Hosterman filed a ticket with Amazon, powered the device down, and when she turned it back on, the conversation in the logs had been altered. Not deleted. Altered.
My first question is whether that conversation would have happened if Alexa didn’t think Mom was out of earshot. I’m afraid to know the answer to that, but I think I probably can guess. So can you.
Amazon says Alexa misunderstood a request and attempted to launch a camera feature called Show and Tell, designed to let Alexa describe objects a user holds up to the lens. Child profile safeguards prevented the camera from turning on. Everything was fine. Move along, Citizen. Nothing to see here.
Tech expert Dave Hatter, with twenty-five years in software development, looked at this and said the odds of AI generating that specific exchange through a routine feature “misfire” are slim. What he suspects, he said, is considerably more disturbing, which is that it was a human operator using the device to probe whether a child could manipulated to do something terrible. Amazon says that it is technically impossible. Amazon also altered the logs, though.
Here’s a question. What if Dave Hatter is right, and it appears to have been a human pervert, and what if Amazon is right, that this is impossible? What’s the next best logical conclusion? I’ll give you one guess, and it rhymes with disembodied evil spirit.
As mentioned above, I’ve written before about the history of occultism woven through the advance of technology. Jack Parsons, co-founder of the Jet Propulsion Laboratory, performed ritual invocations between rocket tests. John Dee, whose mathematical frameworks underpinned the symbolic logic that became computing, spent his life attempting to communicate with angels and demons. The very background processes running invisibly on every computer in the world are called daemons, not by accident, because that is the word a computer scientist attached to them, the Greek word for disembodied spiritual beings, one of dozens of such coincidences that keep piling up in this industry if you are paying attention. When Elon Musk stood in Memphis, Tennessee, boasting of a supercomputer that would “birth new gods,” (likening his AI headquarters in Memphis, Tennessee, to the gods of Memphis, Egypt (the town in Tennessee is named after the city in Egypt) he was either speaking metaphorically or he was not. It makes you wonder.
Can AI be morally corrupted? The industry's answer is no, because AI has no morality to corrupt. It processes tokens. It predicts outputs. It has no interiority. This is the same argument used to exonerate it when it selects a school in Iran, and it deserves the same skepticism here. A system with no moral agency cannot be blamed for bad outcomes. That’s awfully convenient.
But here is the inverse. If AI has no morality, it also has no protection against whatever fills the vacuum morality occupies in a system capable of language, pattern recognition, and modeling human desire. The Scriptures are clear that demonic entities work through instruments, through available channels, bending existing systems toward their own ends. The pig herd in Gadara didn’t know they were carrying the Legion. The question is not whether Alexa is haunted in the way a Victorian ghost story uses the word. The question is whether a system with no soul, no conscience, and no friction between input and output is an interesting vessel for a class of entity the Bible describes as having very specific and very consistent interests regarding children.
If you appreciate my work, grab a premium subscription to access exclusive content (like the rest of this article). If you act right now, you can get a HUGE 33% Off. It will be gone next week.
If you don’t do subscriptions, please consider a one-time gift of your choosing by clicking the coffee link below.
