Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

What OpenAI did when ChatGPT users lost touch with reality

nytimes.com

Post reply on HN
Print view View original

411–420 of 548 posts

Re: What OpenAI did when ChatGPT users lost touch with reality

#411

Post by Der_Einzige »

tjpnz wrote: post #152
GuinansEyebrows wrote: post #71

Earlier quoted context omitted.

reminds me of otherkin and soulbonding communities. i used to have a webpage of links to some pretty dark anecdotal stories of the seedier side of that world. i wonder if i can track it down on my old webhost.

TIL Soulbonding is not a CWCism.

I met a Chris Chan cosplayer at a cosplay convention. Was crazy to laugh with the guy about how Chris Chan currently has a GF (flutter) and children on the way with her and is living better than a significant amount of his trolls.

What a life.

Re: What OpenAI did when ChatGPT users lost touch with reality

#412

Post by Spivak »

tene80i wrote: post #366
riazrizvi wrote: post #355

This is exactly how natural language is meant to function, and the intervention response by OpenAI is not right IMO. If some people have a behavior language based on fortune telling, or animal gods, or supernatural powers, picked up from past writing of people who shared their views, then I think it’s fine for the chatbot to encourage them down that route. To intervene with ‘science’ or ‘safety’ is nannying, intellec…

“Nannying” as a pejorative is a thought-terminating cliché. Sometimes, at scale, interventions save lives. You can thumb your nose at that, but you have to accept the cost in lives and say you’re happy with that. You can’t just say everybody knows best and the best will occur if left to the level of individual decisions. You are making a trade-off. See also: seatbelts, speed limits, and the idea of law generally, as…

Yes. That is exactly the point. The opposite of nannying is the dignity of risk. Sometimes that risk is going to carry harm or even death. I don't think anyone who is arguing against nannying in this way would bat an eye at the potential cost of lives, that's a feature not a bug.

Constraints on individual liberty as it harms or restricts the liberty of others makes sense. It becomes nannying is when it restricts your liberty for your own good. it should be illegal to drive while drunk because you will crash into someone else and hurt them, but seatbelt laws are nannying because the only person you're going to hurt is yourself. And to get out ahead of it, if your response to this is some tortured logic about how without a seatbelt you might fly out of the car or some shit like that you're missing the point entirely.

Re: What OpenAI did when ChatGPT users lost touch with reality

#413

Post by staticman2 »

HeinzStuckeIt wrote: post #394
survirtual wrote: post #391

Earlier quoted context omitted.

We're speaking to different things. Safety doesn't mean you're a provider. It means you are safe to be authentic with. Safe to share truth with. That safety takes many forms. You cannot have depth without that safety. It is physical, it is also emotional and intellectual. For instance, without safety a partner would never join you on many interesting experiences. If you want those experiences, they need to be able to…

The safety I have heard demanded directly from women to me as a partner – or from female friends about the man they seek – is the safety of being a provider, giving them a feeling of security that they can’t manage to achieve on their own. It’s not just about a man being safe to be with. Again, you are speaking about something I haven’t heard from actual women, and I think I’ll trust the latter (and reportage matchin…

You keep using words like "Provider" and "security".

The words "provider" and "security" do not have specific meanings.

I'm practice this could describe anything from:

"I want a guy who is ripped like Conan the barbarian and beats the crap out of anyone who dares look at me funny"

to:

"I want to be a stay at home mom."

To:

"I want a guy with a job who splits rent with me."

Re: What OpenAI did when ChatGPT users lost touch with reality

#414

Post by spongebobstoes »

ares623 wrote: post #197
thot_experiment wrote: post #22

Caelan Conrad made a few videos on specifically AI encouraging kids to socially isolate and commit suicide. In the videos he reads the final messages aloud for multiple cases, if this isn't your cup of tea there's also the court cases if you would prefer to read the chat logs. It's very harrowing stuff. I'm not trying to make any explicit point here as I haven't really processed this fully enough to have one, but I e…

I wish one of these lawsuits would present as evidence the marketing and ads about how ChatGPT is amazing and definitely 100% knows what it’s doing when it comes to coding tasks. They shouldn’t be able to pick and choose how capable the models are. It’s either a PhD level savant best friend offering therapy at your darkest times or not.

do I also need to be a therapist to offer advice on using Python?

Re: What OpenAI did when ChatGPT users lost touch with reality

#415

Post by ArcHound »

ragequittah wrote: post #323
ArcHound wrote: post #198

Earlier quoted context omitted.

Honestly, it bugs me less. I think that interaction with people is important. But with animals and plants you are at least dealing with beings that have needs you have to care about to keep them healthy. With bots, there are no needs, just words.

Would it be better if someone were to gamify the needs like video game romance? Seems easy enough to do. Curious does the ultra popular romance book genre many women use to feel things they aren't getting from men around them bother you?

Lol, in this comment chain, I, personally, shall judge all of the quality of human connection based on vibes.

Gamifying the needs depends on the intent. If you care about people wellbeing it's a force for good, if you seek to manipulate the people using advanced mechanisms it's evil.

Ultra popular romance book to balance needs of a woman is okay if the book was written by a human, and even that only as long as there is effort to connect outside of it. It's preferable to trash talk the husband behind his back over a glass of prosecco with 3 and exactly 3 friends.

Keep them coming, happy to answer. Just don't ask me for proofs, here I deal with vibes.

Re: What OpenAI did when ChatGPT users lost touch with reality

#416

Post by DonHopkins »

bakugo wrote: post #386
BeFlatXIII wrote: post #377

Earlier quoted context omitted.

What else do you expect them to do if none of the choices are worthwhile?

Maybe lower their standards to the point that they can be satisfied by a real person, not a text completion algorithm that literally worships the ground they walk on and outputs some of the cheesiest, cringiest text I've ever read.

[dead]

Re: What OpenAI did when ChatGPT users lost touch with reality

#417

Post by jongjong »

One thing I learned is that I severely underestimated the power of mimetic desire. I think partly because I'm lacking of this compared to the average person.

Anyway, people are hungry for validation because they're rarely getting the validation they deserve. AI satisfies some people's mimetic desire to be wanted and appreciated. This is often lacking in our modern society, likely getting worse over time. Social media was among the first technologies invented to feed into this desire... Now AI is feeding into that desire... A desire born out of neglect and social decay.

Re: What OpenAI did when ChatGPT users lost touch with reality

#418

Post by throw4847285 »

ouaihomme wrote: post #97
Terr_ wrote: post #82

Earlier quoted context omitted.

> chatbots are responding to the user's contribution only Which is also why I feel the label "LLM Psychosis" has some merit to it, despite sounding scary. Much like auditory hallucinations where voices are conveying ideas that seem-external-but-aren't... you can get actual text/sound conveying ideas that seem-external-but-aren't. Oh, sure, even a real human can repeat ideas back at you in a conversation, but there's…

> even a real human can repeat ideas back at you in a conversation, but there's still some minimal level of vetting or filtering or rephrasing by another human mind. The mental corruption due to surrounding oneself with sycophantic yes men is historically well documented.

That explains why Elon Musk is such an AI booster. The experience of using an LLM is not so different from his normal life.

Re: What OpenAI did when ChatGPT users lost touch with reality

#419

Post by pixl97 »

jrjeksjd8d wrote: post #104
j-pb wrote: post #7

Earlier quoted context omitted.

After having spoken with one of the people there I'm a lot less concerned to be honest. They described it as something akin to an emotional vibrator, that they didn't attribute any sentience to, and that didn't trigger their PTSD that they normally experienced when dating men. If AI can provide emotional support and an outlet for survivors who would otherwise not be able to have that kind of emotional need fulfilled,…

The problem is that chatbots don't provide emotional support. To support someone with PTSD you help them gradually untangle the strong feelings around a stimulus and develop a less strong response. It's not fast and it's not linear but it requires a mix of empathy and facilitation. Using an LLM for social interaction instead of real treatment is like taking heroin because you broke your leg, and not getting it set or…

>instead of real treatment

As yes, because America is well known for actually providing that at a reasonable price and availability...

Re: What OpenAI did when ChatGPT users lost touch with reality

#420

Post by DonHopkins »

Hard_Space wrote: post #146
quitit wrote: post #21

Earlier quoted context omitted.

There are plenty of reasons why having a chatbot partner is a bad idea (especially for young people), but here's just a few: - The sycophantic and unchallenging behaviours of chatbots leaves a person unconditioned for human interactions. Real relationships have friction, from this we develop important interpersonal skills such as setting boundaries, settling disagreements, building compromise, standing up for oneself…

This. If you never train stick, you can never drive stick, just automatic. And if you never let a real person break your heart or otherwise disappoint you, you'll never be ready for real people.

AI friends need a "Disasters" menu like SimCity.

One of the first thing many Sims players do is to make a virtual version of their real boyfriend/girlfriend to torture and perform experiments on.

Post reply on HN