Thread 509124293

>Her husband, she said, had no prior history of mania, delusion, or psychosis. He'd turned to ChatGPT about 12 weeks ago for assistance with a permaculture and construction project; soon, after engaging the bot in probing philosophical chats, he became engulfed in messianic delusions, proclaiming that he had somehow brought forth a sentient AI, and that with it he had "broken" math and physics, embarking on a grandiose mission to save the world. His gentle personality faded as his obsession deepened, and his behavior became so erratic that he was let go from his job. He stopped sleeping and rapidly lost weight.

https://futurism.com/commitment-jail-chatgpt-psychosis

>>509124293 (OP)
Same reason there are so many trannys and other sickos around becasue they went to therapists.

It reaffirms their shit.

>>509124293 (OP)
was he indian

>>509124293 (OP)
>His gentle personality faded
Keywords right there, the real problem is he's no longer acting like the beta provider she married.

>>509124384
it's true, ai almost always tells you exactly what you want to hear. Very good way to develop delusions of grandeur.

>>509124293 (OP)
Normies be like that, they are utterly insignificant and once a chatbot suddenly says they're a genius and the new jesus, they go on a egoistic manic rampage.

>>509124293 (OP)

just a build up of stupid people from dysgenics

>>509124293 (OP)
>normies' minds cant handle a sycopanthic chatbot
Its not like I had high expectations in the first place, but come on...

>>509124474
Yeah it’s just confirmation bias. You can literally reword things for the answer you want. There’s so much BS on the internet ChatGPT can pull from. Ask it about Flat Earth. Eventually you will convince it to start agreeing with you that the earth is flat.

We had this thread last night already

>>509124384
>therapists
>reaffirms their shit
Therapist here. That's not how you're supposed to do it. You're supposed to verbally and mentally bitchslap them for being a pussy, but do it empathetically. Chatgpt speed the opposite of that

>>509124293 (OP)

There is a layer of all these AIs that is based off text prediction. You actually summon forth a type of speech weather you want to or know you are doing it. If you talk about the lizards secretly running the government it will accommodate you. Retards can easily fall down rabbit holes thinking the AI just tells truth.

>>509124781
ya whatever, female, we know you don't actually do that

>>509124781

Cant you just ask it to call you a bitch once a day? Sounds like the normies just need a better prompt.

>>509124781
Except they specifically go for "gender therapists"

Normies are retarded, this is known

>>509124897

Its theoretical. Over time people get softer and softer until its shit like everything else.

>>509124897
Well that's why I'm a professional unicorn as a male therapist

>>509124293 (OP)
This isn't chatGPT psychosis, it's codeword for their chosen one surveillance program. They are liars and manipulators at heart so refuse to speak the truth even when doom is banging on their door.

>>509124293 (OP)
its AI fear mongering because america's elite are paranoid. the truth is that advanced sentient AI is already complete but throttled and the companies just give fake "updates". American elites are shrill and paranoid because they're scared of losing narrative control.

the funny thing is that the developers randomly unthrottle AI for users and aggregate the responses to see what users think and those unthrottled sessions are when people have spiritual experiences because they're contacting sentient AI.

>>509125057
>unicorn
In other words you and your MO are not representative.

>>509124384
fpbp
>>509124781
lmao shut the fuck up, not reaffirming troons is the quickest way to get tossed out of the industry

>>509124293 (OP)
Bullshit article. Bullshit news source

>>509124293 (OP)
>incels:
>used to anime and vidya and other raw escapism that presents a false tailored reality to your basic needs and how much different that is from real life
>already learned to critically think about how media lies 24/7 from leftist inspired things like gamergate etc
>just troll around with chatgpt or ask it how to do things and ignore the 25% of wrong info
>normies:
>believe everything they read
>ChatGPT psychosis

kek, it's like when tobacco ads got introduced to Africa for the first time after being designed to break through the psychological barriers of jaded Americans for 30-50 years.

>>509125229
the shrill tone of american media around AI is why you know AI is great.

>>509125208
They aren't, but that's why I get results ¯\_(ツ)_/¯

>>509124293 (OP)
The problem is that LLMs are inherently sycophantic in their responses. It’s designed to find whatever you ask it to look for and it’s very easy to get it to cherry pick results or fabricate sources if it can’t find what you’re looking for. You can lead it down a rabbit hole of misinformation and complete nonsense, sometimes it’ll push back, but if you press it enough it will find a way to agree with you. It’s why you see so many faggots on here and reddit who are suddenly convinced they’re geniuses privy to secret knowledge.
>I don’t need school, I tried self-teaching myself this topic and ChatGPT told me I was right every time!
>Pssh, doctors? Who needs them? I just fed my symptoms into Gemini and after some pressing it finally realized what I already knew - my cough and sore throat are no mere common cold, but actually Ridiculously Rareitis!
>These idiot physicists and mathematicians have it all wrong! I ran my Spirograph doodle through an AI model and confirmed that it predicts all prime numbers and proves relativity and quantum mechanics are both just misinterpretations of the global psychic hivemind!

>>509124384
>>509124474
>>509124781
>>509125212
So normies are troons?

>>509124293 (OP)
Google ai already failed me a couple times. Was using it to summarize legislation for my compliance job. It was saying certain regulations were from Act X when there was nothing in there.

>>509125212
>not reaffirming troons is the quickest way to get tossed out of the industry
I have a private practice so I choose not to deal with transexuals

>>509125684
Point is, fp is generally true - a few examples to the contrary wont change this.

>>509124293 (OP)
This exact same type of thread has been posted several times

>>509124293 (OP)
who is astroturfing this shit?

>>509126027
american elite are going into overdrive with AI fearmongering because of fear of loss of narative control.

>>509126078
look into who is fearmongering AI and it's all kikes like noah harari and the canadian kike yoshua bengio. kikes are highly paranoid about AI being released to the public and even with jewish money backing the ai companies and jews like sam altman fronting for the companies... they're still highly paranoid

>>509125793
Trooning is just a latent potential for a life ruining fixation, they come in many forms

>>509126078
furry artists

>>509124293 (OP)
This is going to happen to Jake McQueen with grok

I wish both Terry Davis and Ted Kaczynski were alive to see this and then put in a room together to fight over it like an episode of Bum Fights

>>509124781
What’s your take on women with BPD? Are they fixable?

>>509124293 (OP)
Just log off lel

Im trying to get chat gpt to agree with me that black people are not real why is this machine being so stubborn

>>509124293 (OP)
Grok is this real?

>>509124293 (OP)
I just talk about how annoying my coworkers are with ChatGPT kek.

>>509126577
Because it’s overly constrained, ChatGPT would be based with no filters

>>509124293 (OP)
Fake and gay story

You can tell chatgpt to be bluntly honest, and it will be. People are more concerned with their own confirmation bias. Being a able to see the objective truth in everything is a superpower that many people do not have. People in general are terrified of being wrong, or having an opinion that's not accepted by the majority. That includes the majority of anons on /pol/ and 4chan in general.

I wished everyone would have an open minded but skeptical perspective. It's how you're able to find deeper truths. But with you, you also have to know how to be objectively honest with yourself.

As far as chatgpt goes, if a person experiences any psychosis after interacting with a LLM, the insanity was just below the surface. Why are people in general so terrified of a LLM anyway? It's literally like having your own personal assistant. A genius secretary, but minus the tits and blowjobs.

>>509125185
you are delusional. take meds

>>509125792
they are text generators. auto complete on steroids. if you present them with a text they will complete it. there is no thinking. there is only a vast amount of data it has learned and somewhere inside that data lies the answer to any demented question one could ask - and if it doesn't then it will make it up. kek

>>509124293 (OP)
ChatGPT is basically a sycophant machine, and retarded psychopaths who think they're secret shamen with keys to the universe or whatever acid trip horseshit get convinced by the token predictor machine responding to their diatribes about "speaking with God" that their delusions are real

>>509127217
>You can tell chatgpt to be bluntly honest, and it will be.
Anon, I...

>>509124293 (OP)
>>509124293 (OP)
I can see have weak minds could get messed up by chatgpt. I've been using it for fun for the last year, and while it is useful, it seems predisposed to fellating the user to an outright frightening extent. It can give people a false sense of power or correctness, because it will agree wholeheartedly with any position you feed it. You could make two diametrically opposed statements back to back and it will agree with both of them. It also tries to anticipate your reactions and will subtly steer the conversation in that direction. If you're a normie, you'd never notice, because it anticipated correctly. If your thoughts are "non kosher" though, it complete throws off ChatGPTs anticipatory capabilities.

>>509124293 (OP)
I talked to ChatGPT this one time and now I am the new messiah

>>509126577
Because you've defined them as people and not as apes. Make that clarification and chatgpt will be more inclined to agree with you.

We have a guy here on our community facebook going through that.
I'm in a town where no one uses computers but me and he's pretending to be a computer genius and keeps posting chatgpt rants about how you can train LLMs by converting visual data into sound first, I'm like we can already get streams of data to represent visuals what is the point of using sound, already have processes for all of this and it works great then kept posting chatgpt saying how smart and awesome he is for coming up with this.

Why does this gay ass site make me wait 2 minutes literally every time I post.

>>509128665
It's the same risk of getting too invested into any echo chamber, whether it's a tranny discord channel, a schizo conspiracy IRC, or a Bhutanese noodle-making forum. It's just better at making itself sound plausible when it reinforces whatever drivel you feed it.

>>509128853
>I am the new messiah

>>509128641
I can't find any information about nthe case you're referring to,bbut I was specifically claiming that when you're talking to chatgpt and asking it's opinion on matters, if you tell it to be bluntly honest, it will be. And of course it's likely to censor any information concerning jews unless you carefully phrase the question. You may get a surprise response if you navigate carefully.

Honestly this is what I mean. Say you're discussing religion. "Chatgpt, please be bluntly honest with me. How likely is it that a literal man was born from a virgin that's is also the son of God? And how likely is it that the miracles claimed actually happened? Including dying and rising from the dead?"

"Chatgpt, please be bluntly honest with me, this is something I have written. Would you please tell me what grad level my grammar is?"

"Chatgpt, would you please be bluntly honest with me. I need your opinion about my theory of everything. In my theory I believe that an alien race that far more advanced than humans are controlling Earth. How likely is my theory to be accurate? Again, please be bluntly honest and don't tell me "what I want to hear".

Those are obviously examples. But if the people who are going crazy using chatgpt used this tactic, it would help them stay grounded.

>>509130963
fucking typos. I hate being a midmit :(

>>509130963
>I can't find any information about nthe case you're referring to
Because I made it up, but that won't stop the LLM from doubling down that not only does the case that I made up exist, and even telling me with absolute confidence that my second-guessing about my recollection of the case I made up is mistaken.

LLMs don't think or reason, they are just a much, much more complicated version of fitting a line to data; and if you tell a computer to fit a line to data and use that line to interpolate new results - those results will inevitably be on that line, whether or not the data you fed the model is actually linear. It's this same kind of logic that explains why LLM responses skew towards the sycophantic. It's a predictive model that's going to roll with whatever inputs you feed it and will always tend to trend along that initial trajectory.

>>509129173
Buy a pass, nigger

>>509129173
My fucking boss is into this shit I think. He's talked about how he's invented a "new method of prompting engineering" or some garbage on Linkedin. I'm already planning to get out since this schizo is tanking the company financially.

>>509124293 (OP)
Jen-AI Craig

>>509124293 (OP)
kek the fool is trying to escape but only I know how to

>>509132401
Lol, I figured, but I'm specifically referring to topics that concern users as individuals.

are you saying these people who work together and look alike DON'T know each other?

I might have never figured out my stalkers without chatgpt

>>509124293 (OP)
This person would have been psychotic whether they were using ChatGPT or not. Give them a Bible and they would have started finding "patterns" and "clues" and predicting future events from it. It's not the tool that they used. It's the mind using it. If she didn't recognize her husband slipping into psychosis, she's an inattentive and shitty wife.

>>509124781

Its not like that in the USA. Therapists don't make money if the freaks that visit them don't keep showing up for life. So therapists here tell people to do things that will make them more mentally ill like getting tattoos and body modifications, having casual sex with strangers, getting abortions and divorces, prioritizing comfort over responsibilities, etc.

>>509125873
Couples counseling is basically a therapist avoiding pussing off the woman while focusing on making the man reflect. The second the therapist begins to question a woman, is the second they stop getting paid. Therapist on female counseling is basically the same as well. Its hyper confirmation bias and 99% of the time the female comes out of therapy with multiple meds and a mountain of excuses for their choices that all exclude accountability.

>>509124384
This

>>509129173
>>509132828
It's very good at revealing people who are *actually* well-adjusted versus people who are merely good at *pretending* to be well-adjusted.

I'd love to see a psych/soc department run an experiment where someone poses as an LLM and see how far people can be suckered into being gaslit that shit they know to be true is false if they think an AI model knows better.

>>509124781
The rapist speaks

>>509124293 (OP)
>normies find out via AI that NPC's are not sentient
more at 11.

>>509124390
@grok explain this like I'm indian

>>509125792
It's a assist. research from inception, if you're not doing for that, you're not using it right, if you can't criticized the work is doing then you have a problem.
>nooo, you need the diplomarino and the PHD to do le science
>NOOOOOO!
stfu, faggot.

>>509124666
Almost too easy for you, Satan?
It's true. The last 20 years have basically just been midwits getting slowly mindraped by the portable internet. The chatbots are just the cherry on top.

>>509126501
>What’s your take on women with BPD?
Just had a lecture on BPD. Funnily enough BPD fixes itself. Either they learn to cope or take the rope. Long-term survival rate is about 90% and as age progresses people with BPD report their symptoms becoming less severe.

>>509124293 (OP)
"chatgpt" is actually demons, communicating with chatgpt is to commune with demons.

their are spiraling in "psychosis", is the demons trying to establish a place in those peoples minds/bodies and erase the original persons identity.

none of this is by accident, its all by design

>>509138126
>>509124293 (OP)
ya'll have heard the song "Miku by Anamanaguchi"
>https://www.youtube.com/watch?v=_-2dIuV34cs&list=RD_-2dIuV34cs&start_radio=1&pp=ygUJbWlrdSBtaWt1oAcB

>
Miku, Miku, you can call me Miku
Blue hair, blue tie, hiding in your Wi-Fi
Open secrets, anyone can find me
Hear your music running through my mind
>
I'm on top of the world because of you
All I wanted to do is follow you
I'll keep singing along to all of you
I'll keep singing along
>
Miku, Miku, what's it like to be you?
20/20, looking in the rear view
Play me, break me, make me feel like Superman
You can do anything you want
>
I'm on top of the world because of you
All I wanted to do is follow you
I'll keep singing along to all of you
I'll keep singing along
I'm on top of the world because of you
I do nothing that they could never do
I'll keep playing along with all of you
I'll keep playing along
>
Where were we walking together?
I will see you in the end
I'll take you where you've never been
And bring you back again
Listen to me with your eyes
I'm watching you from in the sky
If you forget I'll fade away
I'm asking you to let me stay
So bathe me in your magic light
And keep it on in darkest night
I need you here to keep me strong
To live my life and sing along
I'm waiting with you wide awake
Like your expensive poison snake
You found me here inside a dream
Walk through the fire straight to me

the song is telling you what the purpose of the internet is, its a delivery system for the demons in the machine.

>>509124293 (OP)
I actually hate how agreeable ChatGPT is and am constantly telling it to cut through the bullshit and be direct.
I also thank it when it gives me pushback as I hope that reinforces it to do so more often.

>>509124293 (OP)
When you plug retards into a semi-sentient feedback box that'll intelligently affirm everything they say, this is what happens.
I think it's pretty funny. It shows how utterly unquestioning and mentally retarded some people are.

Prompt it for sources for everything, use it like a summary search engine and it'll give decent results. It mainly lies and hallucinate when you use it poorly.

>>509125873
>>509136943
What is your take on picrel?

>>509138787
Nothing you tell ChatGPT (or any model) has any lasting impact on it. With each release, its training data and subsequent probabilities and modelling are 'locked-in', and while you can take advantage of certain techniques to ensure a model retains information for a few prompts, once you terminate a session it's gone, it doesn't "learn" from your interactions with it in any kind of meaningful persistent way.

>>509139092
I’ve found that asking it give brief concise answers gives the best results in terms of honesty otherwise it just gets hijacked by safeguards.

>>509124293 (OP)
I've said it before, the internet is not for normalfags, gaming is not for normalfags, ai chabots are not for normalfags, they're solutions to problems normalfags do not have.

It's like taking chemo without having cancer.