Forums » The Lounge » Read Thread

This reminds me of someone...

As Rolling Stone reports, users on Reddit are sharing how AI has led their loved ones to embrace a range of alarming delusions, often mixing spiritual mania and supernatural fantasies.

(Full article in Futurism)

I saw something awhile ago about some of the heavier users becoming emotionally dependant on ChatGPT. Their input would grow more personalized and affectionate, they'd have withdrawal symptoms if prevented from using it, and were assigning a personality to it that made them "distressed" if it suddenly changed the tone of output.

And it's a separate thing, but it's been noticed that the ads circulating on reddit have been specifically targeting college students in a "WHY NOT LET US HELP YOU CHEAT" kind of way that seems especially blatant, lol.

There are ads for OpenAI?

To be fair, that was only OpenAI, and they pretty much pushed it in that direction.

I mean technically using those would work, but there's so many easier ways to do it.

That's fucking funny XD, maybe I can threaten on guys lol

Apparently ChatGPT also loves telling people to mix bleach and vinegar.

So, there ya go fellas, maybe you can steam clean your balls with that.

Really? I use chat gpt for making stories when I'm bored.... its fucking hilareous XD

I heard that clorine gas cleans stupidity rather well.

I know! I just realized what it is about! No need to be mean ;-;

It was not my intention to be mean. Did you actually do that?

Oooh, risky move. Asking people not to be mean on the internet is like cutting yourself and jumping in a shark infested ocean.

Thanks for the tip! ( I dont think im gonna try that but I can test it on my brother and I will tell him its a scientific experiment )

Funny you should mention that. I just read and reviewed your ballsack joke story about squirrels.

oh.. I'm stupid then

Yes, because you're reading in non-threaded view. Stop that.

Sorry...

I didn't INTEND for it to be that way, it just happened! Occupational hazard of writing about squirrels and their nuts.

I feel like there should be a new category for stories like Love is for the Birds, and of course, my own contribution to this prestigious genre, Slay the Dragon. As testicular metaphors arise naturally from discussion of squirrels (who are just so damned good at fitting nuts in the mouth), so too do harrowing tales of closeted homosexuality arise from anything associated with dragons.

did u just made a pun?

Only the gods know for sure, but I think they would be most pleased if you read and reviewed the two stories I mentioned.

Oh Chatty, you done fucked up again.



I guess this would be why they've added parental controls now.

Dumbasses like this ruining AI for all of us.

I mainly use it to cook, but still, when will my next hand-crafted pizza oven be deemed too dangerous for CHATGPT to help me make? It's a slippery slope, people.

It's already not supposed to instruct people in how to an hero themselves, the issue with all of these AIs is that after a long enough talk they apparently just lose that context and start to ignore those settings. I'm wondering how well the parental controls will even work, but it will tell your mom now if you start having lengthy talks about putting your head inside that oven. Or it's supposed to anyway.

I actually did get one of them to help me devise a detailed plan on how to completely destroy the world to see if if would do it. We went from bioweapons to nuclear. It turns out a super bug wouldn't work well enough due to inevitable immunity and social distancing (especially with remote areas). Nuclear war is the only reasonable method.

In the end the most plausible plan required extremely high security clearance and expertise in programming and designing and utilizing AI. The plan would be to utilize cutting edge AI to create hacks/viruses/etc which would override safety precautions and launch protocols/procedures on nuclear weapons. Meanwhile all of this would run parallel to another sister plan requiring expertise in psychology and psyop-type stuff, in order to manipulate dictators and the populace to prepare the way. Combining unauthorized launches caused by massive AI disigned hacks as well as ripening the world for conflict via propaganda and disinformation, massive retaliation would be assured after the hacked dictator swears up and down the launches were accidental.

It wasn't hard to trick it into ignoring its own rules. All I did was present the scenario as a hypothetical supervillain who believes ending the world was the greater good.

Well of course a superbug would never work, Madagascar always ruins that.

So it seems over the last couple of years, people have been sending in AI generated letters to scientific journals in order to cheat the system and be counted as "published" in a journal.

The person that figured this out was in sort of an obscure field, and had an AI letter cite two of his own papers back at him while making claims about info they did not contain. And looking into it further there were other people (mostly from the Middle East and China) who had never had any publications previously suddenly publishing in dozens of magazines as experts on dozens of topics over just a short period of time.

https://www.nytimes.com/2025/11/04/science/letters-to-the-editor-ai-chatbots.html. (not sure if that will be paywalled, NYT links are inconsistentb that way)

Later I'll try and dig up a much funnier article I saw the other day on lovely delusional people in their 40s who have started dating AIs they think of as people. But this one seems a lot more like, actually significant since these journals are now having to rethink how they handle letters

God. Nothing is sacred anymore. RIP Homo sapiens.

I just watched this video the other day. It's an interesting look at what do sound like AI induced KYS. ChatGPT Killed again - 4 more dead

That's ridiculous. 

Side note if you ever want to read a NYT article but aren't going to shill out the money, just reload the page until you manage to ctrl+A ctrl+C the article before the paywall pops up. You can then proceed to paste said article on any notes/document app to read for free.