The Many, Many (AI) Killers of Charlie Kirk
Your phone is hallucinating. Your Grok is hallucinating. And, maybe you're hallucinating. How AI is influencing you without you knowing it.
(Content warning: no gore or violent pictures but it does discuss the murder of Charlie Kirk, not in any gruesome manner, to be clear).
We don’t know (as of the publishing of this article) who killed Charlie Kirk. What we do know is that there is a suspect, or at least the vague picture of what appears to be a young man, as well as the gruesome footage of Kirk’s assassination. There was bound to be speculation after any high-profile killing like this, and let’s be clear, Kirk was high profile. His fundraising network speaks for itself; over a hundred million raised in 2024 alone to help Trump. At 31, he had potentially a fifty year career ahead of him, as an influencer, fundraiser, even potential political leader in his own right one day. And so people, as they would for the killing of a Kennedy or even a state legislator, they speculated, and are speculating at this very instant.
Tonight, we’re not talking about these speculations; they would be happening regardless. Instead, tonight the focus is honed solely in on an accelerant fuel of what many in the therapizing community are calling ‘GPT-induced psychosis’ or more broadly ‘AI hallucinations affecting human thinking’. This has thus far in this case represented itself in two key ways, one of attempting to divine the suspect’s identity using AI and the other, through the automatic manipulation of cameras that nearly everyone has on by default in their phones.
First, the many would-be detectives’ sleuthing to find Kirk’s killer resulting in nothing but a semi-intelligent copypaste machine spitting out everything ranging from… well, it’s better to show than tell in this case.
First, the original pictures of alleged suspect provided by the FBI Utah Branch:
Two extremely blurry pictures of a non-descript white guy; really, it could be anyone. From that foundation people went to Chat-GPT, went to Grok, went to whatever AI image generator with a prompt to do the CSI show sci-fi thing they do where an image is ‘cleaned up digitally and sharpened’. And likely armed with nothing else but knowledge of shows like that, it was off to the races.
Discounting the last two satiric parodies of these AI hallucinations, the real examples reveal that there is nothing of value in these ‘cleanup' efforts. These AI models are only as good as their data sets, and even then, those data sets exist solely to regurgitate a statistically probable outcome; the average set of outcomes per request. Anyone can try it from themselves. Give a computer 1000 blurry images and it will return 1000 different cleaned up versions, each visually similar but with tiny, near-microscopic differences: the base statistical average outcome of AI image generations. When a user asks therefore an AI to clean up the picture of the suspect, it will always generate a new image of someone who fits the profile (in this case a white or white-ish young man).
So what though? If you’re gullible enough to go out of your way to ask an AI to clean up a picture and then believe it, you are far, far too trusting of the computer. A computer is only as great as the programmer and their team that made it. And no one’s perfect.
Still, if that was it, it would only be a tiny portion of the population, people who really ought to know better, considering how much the computer has been lying to us all of our lives (you are not, we regret to inform you, the 50,000th visitor to this website. You are not winning a cruise to the Bahamas.).
Yet AI is everywhere, the machine learning is everywhere, even if you don’t know it. Anyone with a modern camera phone can do this with their default settings by taking a photograph or video. As the data is captured, it is sent to a cloud, which stores the original and transmits back a digital copy of what that person took: that is what is in the end saved onto the phone or whatever modern device took it.
For this specific killing, there is the controversy of FREEDOM. Specifically, in how Kirk’s FREEDOM shirt changed before and after the shooting.
Many a commentator noticed that from this one witnesses’ camera, the F disappeared. The D changed font sizes. The O appeared elongated while the M cut in half! What could have caused this? Perhaps Kirk was not dead? Perhaps it was faked? Perhaps this was all a simulation? So spun the yarn of conspiracy.
The truth was and is that as that video was being saved, the AI behind it all smoothed out the sudden and violent shockwave effect the bullet had Kirk’s white shirt, blurring and smudging the letters in an attempt to make them still-legible yet failing even at that. And from that alone, people can be driven insane.
There’s little more to say, other than we live in a P.O.G. (Patriot Occupied Government) or as someone else put it:
PS even the President is on it, at least according to his AI-seeming representation immediately after Kirk’s death (and his subsequent droopy-faced appearance the following morning). Who knows what is going on? The best person to trust is yourself and the people you can see, or touch, or feel. Chiral connections will be our way through this mess.
Until then though…
