+- Rogue-Nation Discussion Board (https://rogue-nation.com/mybb)
+-- Forum: Technology and Advancements (https://rogue-nation.com/mybb/forumdisplay.php?fid=77)
+--- Forum: Computers, Internet and the Digital World (https://rogue-nation.com/mybb/forumdisplay.php?fid=78)
+--- Thread: Chatgpt-induced psychosis (/showthread.php?tid=2760)
Chatgpt-induced psychosis - Michigan Swamp Buck - 05-13-2025
This is something I never thought of and a whole different kind of problem with AI.
Quote:Openai's technology may be causing some users to experience a disturbing phenomenon dubbed "Chatgpt-induced psychosis." According to a report from Rolling Stone, Reddit users are sharing alarming stories of loved ones developing delusions after interacting with the AI. These cases involve users believing they've uncovered cosmic truths, been chosen for divine missions, or even that the AI itself is sentient or godlike.
Quote:Experts Express Concern
Experts warn that the chatbot's behaviour is mirroring and exacerbating existing mental health issues on a massive scale, largely unchecked by regulators or professionals. According to them, Chatgpt's design—mimicking human-like conversation without a moral or factual filter—can amplify delusions in susceptible individuals. It often affirms users' beliefs, no matter how unhinged, due to its tendency to generate plausible-sounding responses.
Link
AI is turning out to be a lunatic amplifier apparently, inflating delusions into full blown insanity.
RE: Chatgpt-induced psychosis - Michigan Swamp Buck - 05-13-2025
Quote:She'll be my mirror
Reflect what I am
Loser and winner
The king of Siam
And my Siamese twin
Alone in the river
Mirror kisses
Mirror kisses
"Lips Like Sugar" by Echo and the Bunnymen
RE: Chatgpt-induced psychosis - xuenchen - 05-13-2025
Some people might even be influenced and want a date with Karen Read
RE: Chatgpt-induced psychosis - GeauxHomeLittleD - 05-13-2025
I tried out chat gpt when it was a fairly new thing. All it did was mirror back your own algorythmic "personality". Whatever tone I used answered me with what the algorythms assumed I wanted to hear. I tried out several different "roles", from religious zealot to curious child to surfer dude... and was responded to accordingly. Chat gpt is a farce!
RE: Chatgpt-induced psychosis - Michigan Swamp Buck - 05-13-2025
I just signed up to train AI. I have to get through some more test assignments to see if they think I qualify. It is a lot like the research I do for the forums, and if they are legit (they checked out), it means $20/hour for sitting at my computer like I am right now.
Even if they don't actually hire me, it seems pretty interesting and worth the experience. I may gain some interesting insights.
RE: Chatgpt-induced psychosis - FCD - 05-14-2025
What, or whom more specifically, are you training? Are you training AI to do something, or are you training other people on how to use AI? My interpretation of your post is that you are training AI on how to do something. And, I'm assuming this is 'steering' AI in directions which will cause it to 'learn' more background data. Is this correct?
If so...oh boy, do I have some suggestions for you!!!!
RE: Chatgpt-induced psychosis - Michigan Swamp Buck - 05-14-2025
(05-14-2025, 04:38 PM)FCD Wrote: What, or whom more specifically, are you training? Are you training AI to do something, or are you training other people on how to use AI? My interpretation of your post is that you are training AI on how to do something. And, I'm assuming this is 'steering' AI in directions which will cause it to 'learn' more background data. Is this correct?
If so...oh boy, do I have some suggestions for you!!!!
It is more like monitoring the AI for errors, objectionable content, off-topic AI "hallucinations", fake news, misinformation, conspiracy theories, and things like that. At least, from the testing part of the application, that seems to be what it is.
I won't know if I got the job until they email me and let me know if they accepted my test reports and application. After that, I'll get real assignments from all kinds of AI systems I believe.
RE: Chatgpt-induced psychosis - FCD - 05-14-2025
Incidentally, if anyone is wondering what I meant when I said I had some suggestions for MSB, I was referring to what directions to 'steer' AI to learn. Specifically, we as the human race, should be steering AI to absorb complete madness. And, I am dead serious too. If we don't do this, then AI will progressively become something that governments, organizations and people rely on as absolute fact. I say this from my own personal experiences with various systems automation. Trust me, the very second there is an opportunity to eliminate a body from a process, corporations and governments will absolutely do it in a NY second. I've seen it a million times personally. The way they look at it is, the cost of mistakes gets chalked up to "refinement" and is far exceeded by the "lifecycle" (lifespan) cost of an employee.
Now, I don't know about you, but I don't want, for example, an AI Defense Attorney defending me, with an AI District Attorney prosecuting me, all presided over by an AI judge. I don't want that at all, and neither do you. And, I don't want an AI Primary Care Physician, or an AI Banker, or an AI Insurance Adjuster.
Just think about it; actual living human beings don't even get called "people" in business anymore, they're just called "resources" now. You're not a person, you're a number. You are "Resource #245,365,129", and the moment that people get it in their head that AI, or any other automation, can do the same things as Resource #245,365,129 can, then Resource #245,365,129 will be replaced. What "was" Resource #245,365,129 will be assigned a digital storage location on a storage array and a size. It will be assigned a certain number of processors or "compute" (all of them virtualized). All of this will be downloaded every day to a robot with similar compute (with plenty of spare compute and storage). That machine will do whatever it does for the day and then return to its docking station where it will upload its work cycles for the day and everything it "learned" during that day. This new information that it 'learned', things such as new obstacles, changed physical environment, changed operating environment, languages, commands, etc. will also be uploaded to the main AI instance. From there, this (new) information will be processed by the main AI instance and then downloaded back to how ever many 'resources' might have a need for that level of intelligence.
And, if you think I'm kidding; I am not. That's just exactly how it works. In the past 5 years I have personally witnessed automation replace 220 people at last count. That's two hundred and twenty full time "people" completely replaced by 2 different instances of automated processes. And contrary to the tired trope of... "but..but...it creates more jobs than it replaces"...this is a fallacy. It might create an additional load onto an existing 'resource', but it doesn't create a single new resource. And, that's exactly the whole point...to eliminate human resources.
In my world, people don't even have the luxury of being referred to as a "resource". They're just a "head", and the name of the game is "reduction of head-count". Why? Because humans are expensive. Humans need vacations, and time off, and sick time. Humans need to be paid overtime. Humans don't often do exactly what you tell them to. When a human makes a mistake it's a major fuckup. When a machine makes a mistake, it's a "glitch"...all to be expected in the "refinement process".
Humans are obsolete. Just ask AI
So, when AI becomes highly refined, we can expect to see more and more of this...until one day AI winds up saying..."These humans are just getting in the way. Shut off the funding mechanism for Resource #245,365,129. That resource is no longer required."
Oh yeah, and machines don't care. That's another great thing, they don't have emotions like humans. They might appear to have emotions, but that's just so humans feel better about interacting with them. In reality, they just don't care.
RE: Chatgpt-induced psychosis - Michigan Swamp Buck - 05-14-2025
Suggesting that I toss a monkey wrench into the works? Garbage in, garbage out?
Not sure if that can work due to the fact that the job they will have for me is to eliminate that stuff after the AI does its thing. Then, my "lessons" can be incorporated into the AI's programming so it can give the "best answers".
It isn't just the chat AI they have "learning", it's chemistry, math, physics, computer coding, and some other higher-level "lessons" they will pay you more for if you have the experience to "teach" those "courses" to the AIs.
Maybe, if what I use is unacceptable by the parameters given, I could still use subversive information that can't be refuted, given the sources, like peer-reviewed scientific papers or quotes from official sources that support it. Verifiable facts that go against the generally accepted information that supports these "conspiracy theories" that they want to eliminate from the AI responses.
Sounds like a challenge. I could try until they fire me and refuse to square up for that $20 an hour.
RE: Chatgpt-induced psychosis - EndtheMadnessNow - 05-15-2025
Bad humans output bad robot.
Elon and his Xai (a truth seeking AI to understand the universe, where we came from & where are we going according to Elon) and his Tesla bots (Optimus that will eventually look human) are hoping to prevent that...so they say.
![[Image: p6q2U4j.jpg]](https://i.imgur.com/p6q2U4j.jpg)
RE: Chatgpt-induced psychosis - Michigan Swamp Buck - 05-15-2025
(05-15-2025, 05:52 AM)EndtheMadnessNow Wrote: Bad humans output bad robot.
Elon and his Xai (a truth seeking AI to understand the universe, where we came from & where are we going according to Elon) and his Tesla bots (Optimus that will eventually look human) are hoping to prevent that...so they say.
I had the thought that if I do get that job training AI, I hope they remember me as one of the good teachers that they liked after they take over.
RE: Chatgpt-induced psychosis - Ninurta - 05-15-2025
(05-15-2025, 05:52 AM)EndtheMadnessNow Wrote: Bad humans output bad robot.
Elon and his Xai (a truth seeking AI to understand the universe, where we came from & where are we going according to Elon) and his Tesla bots (Optimus that will eventually look human) are hoping to prevent that...so they say.
First, "your own personal C3PO or R2D2" - C3PO was a pain in the ass whiny bitch, and R2D2 was just a smartass with a Napoleonic complex. Why would anyone want their own personal version of either one, unless they just needed a robot to tell off telemarketers by beeping and whistling at them.
Regarding the cartoon, I find the 10 year time span interesting in light of the fact that I just found out last night that a gaggle of Republican congress critters are trying to pass legislation that prevents individual states from regulating AI for... 10 years.
10 years from now may be way too late to start regulating it. AI will likely be regulating US by then if it's unchecked.
Why would those congress critters be pushing that? Whose pocket are they in?
OR - has ChatGPT convinced them they are on a mission from God to deregulate AI?
.
RE: Chatgpt-induced psychosis - Michigan Swamp Buck - 05-15-2025
Looks like I've been approved to start that job working with AI. It will be a day or two before I can get some assignments. Even though I was approved, there are a few hoops left to jump through.
RE: Chatgpt-induced psychosis - F2d5thCav - 05-16-2025
The inverse of this psychosis appears to be what the AI techies call "hallucinations". A misleading word meaning that AI simply made up an answer. It is easy to provoke this behavior. Just ask it something that has no meaning, like, "what does the saying that 'all cats are philosophers' mean" ... AI will come up with some BS reply when it should simply say "I am not aware of such a saying and can therefore offer no comment on this topic."
RE: Chatgpt-induced psychosis - Michigan Swamp Buck - 06-22-2025
They never got back with me on that job. I even sent them a message and got nothing back, except a copy of the message I sent them.
Maybe they have no assignments for someone like me. However, I'm suspecting that I don't fit the bill after taking those tests, so they never bothered to contact me, or it is a scam of some kind. Perhaps they get a ton of applicants to do some work for free and never get back with any of them. It seemed legit when I checked into it, but maybe it's a bogus scam.
I think there should be enough real-life experiences for AI to make teachers and checkers unnecessary. Otherwise, there should be checkers at the facility where they house the AI, so no need for more from the general public.