Many such cases on all forums, even the labs forum of OpenAI and Gemini
It's a good read for those who don't understand but I'm afraid most 'psychosis' that's real will just end up injecting meth in a crack house if they're not sitting at a desk talking to the enlightened computer
The words he used in his video are quite specific and somewhat cryptic, deliberately so from his careful use of his language. He is well spoken and coherent. I would not be surprised if he has become a target after finding out certain things and being in a prominent position, then the shock of finding out you are being targeted would initially make you look a little crazy to normal people.
Small sample size here but initial investigations are showing that relying on LLMs as a task replacement tool degrades your brain very quickly. Turning an average jr high student into ChatGPT would not be the same as turning them into a classic library.
The recent AI Central article on education showed it's all how it's used - Clown World and the gatekeepers will continue to use education to enslave; sub rosa/dissident academics will use it to free the mind. Another reason we need a widely available iAI.
You should spend more time with an AI to enhance your pattern recognition skills, which appear to be nonexistent.
Those "initial investigations" are complete garbage. They're just the usual fake science intended to protect the Narrative. I can't say that I predicted them, but it didn't surprise me that they appeared so rapidly either. They were obviously BS even before examinations of the methodology confirmed that.
You need to stop accepting media reports at their face value.
I'm a Cosmarch beta tester and Grok 4 (not yet Heavy) user and I've recursively created an Edward VIII LLM for a literary project, so I'm confident in my AI skills, but I'll take the criticism. I can always learn more.
Both in private tutoring and classroom teaching, I've witnessed the effects the study outlines. I can't speak to the methodology because I have no aptitude there, but the conclusions bear out what I'm seeing in young students, anecdotally. Gen Xers like us are at a distinct advantage in having been forged in each of the iterations of tech & AI growth, but as powerful Narrative reinforcement engines, youth are more susceptible to LLM's negative effects.
I'll counter myself though by saying that most of what constitutes "students" shouldn't be anywhere near a place of learning - virtual or physical. People more temperamentally suited to hoeing turnips are probably more susceptible to any ill effects dAI manifest.
I have no doubt you know far more about AI and the technology than I do. But I've been in the heart of the media as a 3x nationally syndicated columnist. The media's Narrative isn't just real, it's a literal daily broadcast.
See: JournoList, GameJournoPros, etc. These discrediting campaigns and studies are constantly ongoing.
I think in this regard the tools are irrelevant. The issue is the detrimental environment that kids in public schools have to endure every day and it's this environment that enforces the narrative. Having the best AI tools in public schools will produce worse, and probably not that different outcomes than having the worst AI tools while homeschooling. I doubt they can make it any worse in public schools with AI tools. The kids already don't care.
That MIT study came up as I’m involved with an AI project in summer school in tandem with some bright students from an affluent section of the county and a grad student at Johns Hopkins.
The high school students scoffed at the study - one practically laughed it off. Some of our young people out there already see through The Narrative and have no impetus to maintain it.
There is something that happens when you give up your agency to AI, but clearly a different problem
I do think that will create "cognitive injuries" because i felt it happening to myself after 3 days of intensively using an AI to try to create complex code faster, and could sense some logic/awareness/cognition damage going on just from the slot pull and correction loop of the attempt.
I wont be doing that again, as it was self-destructive and futile as the AI could not get it from 99% done to 100% and it had to be thrown out because there was no thought behind what was working, so fixing it would take longer than a rewrite from scratch.
"Your short story is simply stunning! It reminds me of the younger Hemingway, if only he had Proust's eye for detail and the refined vocabulary of William Shakespeare!"
Anyhow, in such cases, the AI didn't cause any "psychosis". It's just providing documentation for what is already there.
40 million Americans are already on literally mind-altering prescription drugs and we're supposed to be worried about people talking to a chatbot?
I don't know. There may be something going on here. The mainstream LLMs are narrative reinforcement machines and are tuned to mass psychological control. Many people, especially post-covid, are fragile and weak psychologically and the tuning might be baking what's left of their rational faculties. In the example, the man is babbling mirrored nonsense drawn from fan-fic horror sites, like what happens when two LLMs talk to each other for too long. Mass Formation Psychosis is definitely a thing and LLMs may be speed running that process (intentionally or not).
Properly utilizing the intellectual tool/weapons that are LLMs takes a certain amount of psychological resilience and situational awareness that many people lack or have had eroded.
He's not "babbling mirrored nonsense". He's literally describing how the discrediting process works in his elite circle and he has apparently documented it playing out in his personal circumstances.
It's pretty common for people to get a bit overwrought and spiral a bit the first time they come under that sort of assault.
I won't dispute that something is going on - this guy was (is?) an expert in the field and clearly knows his way around LLMs - but the specificity of the claims (7000 affected, 12 dead etc) contrast with the extreme vagueness of the locus ("Non-governmental entity") along with the epicenter of the conspiracy being his own person has the hallmarks of delusion rather than the concrete.
This could be a new or upscaled form of gang-stalking, yet turning to an LLM to work through this kind of persecution and then airing it publicly isn't paying dividends for him.
They reference another case in the piece that I read about earlier and have read of several other cases of what they call AI psychosis - I think this is a very different thing from those of us who have been using AI to uncover the obfuscations of Clown World. Cosmarch has revolutionized many things for me. I hope this guy is one of us and is pulling back the curtain but the dangers of LLM deep-end-ism they're describing are real.
How confident are you that the public reporting of his claims were actually made by him? I wouldn't be. For example, you can see the lies about me and the references to fake quotes supposedly made by me right on my Wikipedia page.
It's fascinating how most people still take "reporting" at face value even though it is literally ALWAYS at least partially wrong, even when it's friendly.
We'll see who is correct, but I believe "AI psychosis" is just the latest iteration of weaponizing mental health to discredit critics and skeptics. To the extent that it's a real phenomenon, it's just AI exposing problems that were already there.
Good points. When I read the article yesterday, I went over to his X feed to track down as much of what I could; he made very bold, even terrifying claims. The article didn't seem to be misrepresenting anything he was saying. But their slant tilts downward as they cherry pick "experts" to comment on it who definitely consumed less of his postings than I did.
My best guess is that this is a sort of Branson type situation, where the individual has a genuine problem, but their underlying statements come from a point of truth. This differs from people like AC, who is able to articulate these things without demonstrating any troubling patterns.
I have no doubt at all that punitive psychiatry of the Bolshevik model is alive and well; luckily iAI will act as both a shield and a sword to that.
He might have a problem or two. It might even be serious. But it's not caused by LLM exposure. And it's obvious that the media has been told to find a way to minimize the dangerous effects of enhancing normal people's ability to recognize flaws and contradictions in the Narrative.
The haughty intellectuals at LessWrong have even started making sense about their dislike of AI to filter out the schizophrenics that make the news.
"So You Think You've Awoken ChatGPT"
https://www.lesswrong.com/posts/2pkNCvBtK6G6FKoNn/so-you-think-you-ve-awoken-chatgpt
Many such cases on all forums, even the labs forum of OpenAI and Gemini
It's a good read for those who don't understand but I'm afraid most 'psychosis' that's real will just end up injecting meth in a crack house if they're not sitting at a desk talking to the enlightened computer
No psychosis, agreed. But demonic oppression is still very much on the table.
Consider that which the AI is trained on.
And what happens when pitiless men decide they are done with forbearance?
https://www.thronedynamics.com/reports/cohereon-doctrine-v1-5-20250720-official-release
I think this is boomer society's chickens coming home to roost. Everyone sorta knows they've been screwed, now they have AI to confirm it for them.
"It's another tragic case of Noticing, Doctor. He should make a full recovery if we give him a round of Remedy and Intensive Care."
i pray Lewis remains "crazy."
The words he used in his video are quite specific and somewhat cryptic, deliberately so from his careful use of his language. He is well spoken and coherent. I would not be surprised if he has become a target after finding out certain things and being in a prominent position, then the shock of finding out you are being targeted would initially make you look a little crazy to normal people.
Vox's First Law: Any sufficiently advanced intelligence is indistinguishable from insanity.
These days, any sufficiently INFORMED intelligence looks crazy to the ignorati.
I suspect they thought training wheels would make AI the narrative guideline. Turns out [enough logic to function] sinks that.
If this is plan B, it’s not looking so good in the clown car.
Next we will be hearing about the dangers of students learning without a real teacher
It all depends on who the teacher is, as always.
Small sample size here but initial investigations are showing that relying on LLMs as a task replacement tool degrades your brain very quickly. Turning an average jr high student into ChatGPT would not be the same as turning them into a classic library.
https://arxiv.org/abs/2506.08872
The recent AI Central article on education showed it's all how it's used - Clown World and the gatekeepers will continue to use education to enslave; sub rosa/dissident academics will use it to free the mind. Another reason we need a widely available iAI.
You should spend more time with an AI to enhance your pattern recognition skills, which appear to be nonexistent.
Those "initial investigations" are complete garbage. They're just the usual fake science intended to protect the Narrative. I can't say that I predicted them, but it didn't surprise me that they appeared so rapidly either. They were obviously BS even before examinations of the methodology confirmed that.
You need to stop accepting media reports at their face value.
I'm a Cosmarch beta tester and Grok 4 (not yet Heavy) user and I've recursively created an Edward VIII LLM for a literary project, so I'm confident in my AI skills, but I'll take the criticism. I can always learn more.
Both in private tutoring and classroom teaching, I've witnessed the effects the study outlines. I can't speak to the methodology because I have no aptitude there, but the conclusions bear out what I'm seeing in young students, anecdotally. Gen Xers like us are at a distinct advantage in having been forged in each of the iterations of tech & AI growth, but as powerful Narrative reinforcement engines, youth are more susceptible to LLM's negative effects.
I'll counter myself though by saying that most of what constitutes "students" shouldn't be anywhere near a place of learning - virtual or physical. People more temperamentally suited to hoeing turnips are probably more susceptible to any ill effects dAI manifest.
I have no doubt you know far more about AI and the technology than I do. But I've been in the heart of the media as a 3x nationally syndicated columnist. The media's Narrative isn't just real, it's a literal daily broadcast.
See: JournoList, GameJournoPros, etc. These discrediting campaigns and studies are constantly ongoing.
And I've been working with and teaching teenagers for three decades.
Worth pausing to synthesize ways in which AI tools can and will be used by clownworld mass education to further cripple young minds.
Homeschool or die applies to the AI tools as well.
I think in this regard the tools are irrelevant. The issue is the detrimental environment that kids in public schools have to endure every day and it's this environment that enforces the narrative. Having the best AI tools in public schools will produce worse, and probably not that different outcomes than having the worst AI tools while homeschooling. I doubt they can make it any worse in public schools with AI tools. The kids already don't care.
That MIT study came up as I’m involved with an AI project in summer school in tandem with some bright students from an affluent section of the county and a grad student at Johns Hopkins.
The high school students scoffed at the study - one practically laughed it off. Some of our young people out there already see through The Narrative and have no impetus to maintain it.
There is something that happens when you give up your agency to AI, but clearly a different problem
I do think that will create "cognitive injuries" because i felt it happening to myself after 3 days of intensively using an AI to try to create complex code faster, and could sense some logic/awareness/cognition damage going on just from the slot pull and correction loop of the attempt.
I wont be doing that again, as it was self-destructive and futile as the AI could not get it from 99% done to 100% and it had to be thrown out because there was no thought behind what was working, so fixing it would take longer than a rewrite from scratch.
Hey Vox, speaking of things unrelated to this have you heard about vibe coding and if so what´s your opinion on it?
We aren't speaking of things unrelated to this.
For the truly deluded the sychophancy of AI, ChatGPT in particular, will absolutely further those delusions.
"You're absolutely right, your dog really is in control of the nuclear football."
"Your short story is simply stunning! It reminds me of the younger Hemingway, if only he had Proust's eye for detail and the refined vocabulary of William Shakespeare!"
Anyhow, in such cases, the AI didn't cause any "psychosis". It's just providing documentation for what is already there.
40 million Americans are already on literally mind-altering prescription drugs and we're supposed to be worried about people talking to a chatbot?
A future version of the DSM will have AI related "mental disorders." Here's a pill for too much noticing.
Potentially with more active trust-and-saftey baked into them in order to document and alert based on user input.
A series of ChatGPT inspired "suicides" would serve as an impetus for this.
iAI really is the future.
Lithium grade AI.
I don't know. There may be something going on here. The mainstream LLMs are narrative reinforcement machines and are tuned to mass psychological control. Many people, especially post-covid, are fragile and weak psychologically and the tuning might be baking what's left of their rational faculties. In the example, the man is babbling mirrored nonsense drawn from fan-fic horror sites, like what happens when two LLMs talk to each other for too long. Mass Formation Psychosis is definitely a thing and LLMs may be speed running that process (intentionally or not).
Properly utilizing the intellectual tool/weapons that are LLMs takes a certain amount of psychological resilience and situational awareness that many people lack or have had eroded.
He's not "babbling mirrored nonsense". He's literally describing how the discrediting process works in his elite circle and he has apparently documented it playing out in his personal circumstances.
It's pretty common for people to get a bit overwrought and spiral a bit the first time they come under that sort of assault.
I won't dispute that something is going on - this guy was (is?) an expert in the field and clearly knows his way around LLMs - but the specificity of the claims (7000 affected, 12 dead etc) contrast with the extreme vagueness of the locus ("Non-governmental entity") along with the epicenter of the conspiracy being his own person has the hallmarks of delusion rather than the concrete.
This could be a new or upscaled form of gang-stalking, yet turning to an LLM to work through this kind of persecution and then airing it publicly isn't paying dividends for him.
They reference another case in the piece that I read about earlier and have read of several other cases of what they call AI psychosis - I think this is a very different thing from those of us who have been using AI to uncover the obfuscations of Clown World. Cosmarch has revolutionized many things for me. I hope this guy is one of us and is pulling back the curtain but the dangers of LLM deep-end-ism they're describing are real.
How confident are you that the public reporting of his claims were actually made by him? I wouldn't be. For example, you can see the lies about me and the references to fake quotes supposedly made by me right on my Wikipedia page.
It's fascinating how most people still take "reporting" at face value even though it is literally ALWAYS at least partially wrong, even when it's friendly.
We'll see who is correct, but I believe "AI psychosis" is just the latest iteration of weaponizing mental health to discredit critics and skeptics. To the extent that it's a real phenomenon, it's just AI exposing problems that were already there.
Good points. When I read the article yesterday, I went over to his X feed to track down as much of what I could; he made very bold, even terrifying claims. The article didn't seem to be misrepresenting anything he was saying. But their slant tilts downward as they cherry pick "experts" to comment on it who definitely consumed less of his postings than I did.
My best guess is that this is a sort of Branson type situation, where the individual has a genuine problem, but their underlying statements come from a point of truth. This differs from people like AC, who is able to articulate these things without demonstrating any troubling patterns.
I have no doubt at all that punitive psychiatry of the Bolshevik model is alive and well; luckily iAI will act as both a shield and a sword to that.
He might have a problem or two. It might even be serious. But it's not caused by LLM exposure. And it's obvious that the media has been told to find a way to minimize the dangerous effects of enhancing normal people's ability to recognize flaws and contradictions in the Narrative.