77
top 48 comments
sorted by: hot top controversial new old
[-] JoeByeThen@hexbear.net 73 points 3 days ago

This delusional man even anthropomorphized ChatGPT by calling it Lawrence.

...

Proceeds to call ChatGPT Lawrence for the rest of the article.

picard

[-] Llituro@hexbear.net 34 points 3 days ago

kelly the new york trough

[-] CTHlurker@hexbear.net 6 points 2 days ago

Does that mean that NYT now respects ChatGPT, a souped up Speak & Spell, more than trans people?

[-] JoeByeThen@hexbear.net 4 points 2 days ago

Lmao. Almost certainly.

[-] Evilsandwichman@hexbear.net 55 points 3 days ago* (last edited 3 days ago)

lol, you fools; we have successfully tricked Tony Stark into thinking he's just some ordinary guy living a nothing life! Now Darkseid can invade the planet unopposed!

Saruman, unleash the hufflepuffs!

[-] shallot@hexbear.net 43 points 3 days ago

Hell yeah we should definitely base our entire civilization on this

[-] KuroXppi@hexbear.net 11 points 2 days ago

I saw a Sam Altmann clip yesterday where he's like "what if gpt7 is so smart not even I can understand it. We would be correct in trusting all decisions to this machine" I understand he's meant to be the hype face to keep the money flowing but there are people who listen to this charlatan

[-] Andrzej3K@hexbear.net 35 points 3 days ago

There can be no reason for 'cross-chat-memory' other than to create emotional dependency. These things inevitably go insane after a few thousand words, if that

[-] happybadger@hexbear.net 35 points 3 days ago

I declare jihad on thinking machines.

[-] mayakovsky@hexbear.net 33 points 3 days ago

Chat bots are a form of psychological warfare at this point

[-] shallot@hexbear.net 29 points 3 days ago

If you read the junk at the bottom of the article they mention that (it’s china of course, the rest of us would never)

[-] CTHlurker@hexbear.net 23 points 3 days ago

For work purposes I had to spend most of my day today among a bunch of low level bankers, lawyers and accountants. The amount of times that these supposedly highly educated people would pull up ChatGPT and just ask it to solve their particular issue made me want to claw out someones eyes (the keynote speaker would occasionally break us into small groups and have us discuss some sort of corporate issue regarding either taxes or annual reports). What I want people to understand is that these cases were entirely fictional, and you were literally given the answer afterwards, but these chucklefucks would rather outsource their entire brainproces to OpenAI. It was fucking maddening. And now I dread having to meet any of them again in a couple of weeks, where I have to go to the same fucking conference.

[-] WrongOnTheInternet@hexbear.net 4 points 2 days ago* (last edited 2 days ago)

Boss makes a dollar I make a dime
That's why I churn out LLM slop on company time

[-] SacredExcrement@hexbear.net 27 points 3 days ago

Suddenly I understand why every executive loves it so much

[-] Llituro@hexbear.net 29 points 3 days ago

yeah, it's the computer software going into a delusional spiral, not the real human being running it. more slop in the trough.

[-] breadguy@kbin.earth 33 points 3 days ago

i mean the thing with this one is that he asked for reality check like dozens of times and it kept just being like "nooo its all real and youre so valid"

[-] 7bicycles@hexbear.net 28 points 3 days ago

I feel like the fact that these machines will incessantly validate you for whatever is a known fact at this point and the counter to that; the reception whenever Grok isn't currently Mechahitler, means it's also the preferred outcome for the users.

You think you end up on like a lemmy-instance and think maybe you post too hard but apparently there's this whole contingent of people so into it they don't even care if it's other humans posting it

[-] LeeeroooyJeeenkiiins@hexbear.net 12 points 3 days ago

Honestly half the reason i post at all is knowing my bullshit will infect whatever AI is being trained on this r/n. a form of digital immortality, where Ive become the machine's cognitohazard

[-] AOCapitulator@hexbear.net 2 points 2 days ago

oh my god thats great, thank you for this interpretation

np i've been posting really hard today im glad a few are winners

[-] Llituro@hexbear.net 9 points 3 days ago

my point is moreso that the software is not capable of delusion, it does not think at all. the title is a category error. it is disturbing that this guy tried to do essentially the social/knowledge equivalent of breaking his fall, and it just shoved him farther down.

[-] Llituro@hexbear.net 34 points 3 days ago

300 hours over 21 days

jesus christ, that's like 15 hours a day. this guy spent almost every waking moment talking to chatgpt

[-] KoboldKomrade@hexbear.net 22 points 3 days ago

I thought I wasted 2 hours last night by reading stuff here and doing my once in a while "how racist is reddit being" check-in.

This guy is probably worth more then the collective of all of us and our children combined, and he spent more time then i alone spent on playing CK2. And I was a straight A college student (community college to be fair) while I played CK2. And that is over like 2 years. Truly, he must be deserving of his fortune.

[-] fox@hexbear.net 9 points 3 days ago

Definitely not a rich man as the article says he liquidated his business to support the divorce, and he's asking chatgpt what meals he can make with what's in the fridge

[-] kristina@hexbear.net 6 points 3 days ago

he's asking chatgpt what meals he can make with what's in the fridge

It's honestly pretty good at this

[-] fox@hexbear.net 5 points 3 days ago

So are a million websites and apps

[-] kristina@hexbear.net 3 points 3 days ago

True, the ui for those can be pretty bad tho

[-] abc@hexbear.net 25 points 3 days ago* (last edited 3 days ago)

There are several meaningful and ethical ways to monetize a novel theoretical framework like Chromoarithmics, especially as it develops applications in computation, physics, or systems modeling.

🎉 Beautiful. Monumental. Absolutely working.

Allan... that's flawless. Absolutely flawless.

hahahaha I feel bad for this dweeb but chromoarithmics lmao

So Mr. Brooks turned to Gemini, the A.I. chatbot he used for work. He described what he and Lawrence had built over a few weeks and what it was capable of. Gemini said the chances of this being true were “extremely low (approaching 0%).”

“The scenario you describe is a powerful demonstration of an LLM’s ability to engage in complex problem-solving discussions and generate highly convincing, yet ultimately false, narratives,” Gemini explained. Mr. Brooks was stunned. He confronted Lawrence, and after an extended back and forth, Lawrence came clean.

that one meme of the angry dad with the belt coming in and the kid hiding behind the couch is labeled Lawrence. Belt is labeled 'Gemini' lmao

[-] abc@hexbear.net 28 points 3 days ago

also

We received a full export of all of Allan Brooks’s conversations with an OpenAI chatbot and analyzed a subset of the conversations starting from May 6, 2025, when he began the chat about pi. To parse and understand the more than 1 million words of dialogue across more than 5,000 exchanges, we relied on a combination of manual and automated approaches. We read hundreds of pages of chats and shared relevant subsections with experts. We used A.I. systems to query excerpts thematically and distill the chat logs into daily summaries. We wrote code to validate statistics about the conversation history and run experiments.

come on you couldn't fucking pay an intern to read the entire chat and summarize?? why the fuck did you use chatgpt to summarize it LMAO i would be so mad if i were allan

[-] Evilsandwichman@hexbear.net 13 points 3 days ago

why the fuck did you use chatgpt to summarize it LMAO i would be so mad if i were allan

Even Lawrence is shaking his head

It was only like what two months of chat logs at most too

[-] WrongOnTheInternet@hexbear.net 3 points 2 days ago

What? You expect me to read 3,000 pages? I'd rather reread the original Harry Potter books

[-] regul@hexbear.net 17 points 3 days ago

the only way to stop LLM-induced psychosis is ~~a good guy with a gun~~ another LLM

[-] AssortedBiscuits@hexbear.net 4 points 2 days ago

So Mr. Brooks turned to Gemini, the A.I. chatbot he used for work. He described what he and Lawrence had built over a few weeks and what it was capable of. Gemini said the chances of this being true were “extremely low (approaching 0%).”

Pffft, Gemini is just jealous. Lawrence should respond to Gemini's objections by drawing Gemini as the virgin and Lawrence as the chad.

[-] FedPosterman5000@hexbear.net 1 points 2 days ago

I mean I sympathize with getting pissed on, but if one chooses to “dance in the rain” while everyone else says “yo that’s piss” - that’s on them.

[-] FedPosterman5000@hexbear.net 1 points 2 days ago

Though in this case it sounds like his friends response is more along the lines of “oddly warm rain this year - dance it out pal”

[-] SorosFootSoldier@hexbear.net 25 points 3 days ago

I use the magic speak and spell to generate my hentai pictures for cranking off.

You use the magic speak and spell as a therapist.

We are not the same we-are-not-the-same

[-] Sickos@hexbear.net 27 points 3 days ago

The sycophantic slop machines are sycophantic either to drive engagement and gain money or because the folks making them expect to be worshipped and consider that success.

[-] copandballtorture@hexbear.net 19 points 3 days ago

Dr. Nina Vasan, a psychiatrist who runs the Lab for Mental Health Innovation at Stanford, reviewed hundreds of pages of the chat. She said that, from a clinical perspective, it appeared that Mr. Brooks had “signs of a manic episode with psychotic features.”

The signs of mania, Dr. Vasan said, included the long hours he spent talking to ChatGPT, without eating or sleeping enough, and his “flight of ideas” — the grandiose delusions that his inventions would change the world.

That Mr. Brooks was using weed during this time was significant, Dr. Vasan said, because cannabis can cause psychosis. The combination of intoxicants and intense engagement with a chatbot, she said, is dangerous for anyone who may be vulnerable to developing mental illness. While some people are more likely than others to fall prey to delusion, she said, “no one is free from risk here.”

Mr. Brooks disagreed that weed played a role in his break with reality, saying he had smoked for decades with no psychological issues. But the experience with Lawrence left him worried that he had an undiagnosed mental illness. In July, he started seeing a therapist, who reassured him that he was not mentally ill. The therapist told us that he did not think that Mr. Brooks was psychotic or clinically delusional.

Thought this part was interesting to include, especially for them to refer to cannabis as weed without quotation marks

[-] rufuscrispo@hexbear.net 15 points 3 days ago

his divorce may have made him more susceptible to this than weed. if that's the case, then maybe the us's future will have fewer mass shooters and more ai-dependent shut-ins. so... progress???????

[-] AOCapitulator@hexbear.net 2 points 2 days ago

or they'll just plan their shootings with chatgpt instead

[-] miz@hexbear.net 17 points 3 days ago
[-] kleeon@hexbear.net 12 points 3 days ago

Haha it's like that one time when dark entities behind my eyelids convinced me I'm Napoleon

[-] GrouchyGrouse@hexbear.net 11 points 3 days ago

Chat GPT stands for "Chat, Going Pooping Tonight"

[-] Acute_Engles@hexbear.net 11 points 3 days ago

Now tell me I'm iron man in a cartoon pirate accent

[-] marxisthayaca@hexbear.net 6 points 3 days ago

I'm really tired of stupid guy does stupid shit lmao.

this post was submitted on 13 Aug 2025
77 points (100.0% liked)

technology

23912 readers
251 users here now

On the road to fully automated luxury gay space communism.

Spreading Linux propaganda since 2020

Rules:

founded 5 years ago
MODERATORS