Recently I’ve been unloading all of my trauma, problems and even sexual stuff onto ChatGPT and it’s helping me make sense of it. I’m using a personal email but after what transpired on this site recently I’m rethinking it. Could my secrets and neurosis be leaked? There’s some pretty dark stuff im grappling with that I NEVER want to get out here. Im aware that Open AI see it and keep it to train the models but I don’t care as long as it can’t be traced back to me

Am I safe?

If you’re using it tied to an account that’s publicly tied to you then it can be accessed at some point if neccessary.

>>33331431
>if necessary
By who?
And what if I delete it before

Government I guess? The company? Hackers if they ever get access?

If you delete texts on your phone the phone company still keeps hold of it forever.

They feed these convos into the AI to teach it so I’m sure it’s saved in some form somewhere

ChatGPT makes a super shitty therapist. It detects what you want to hear and gives you that. Very dangerous stuff. At a minimum use 2 accounts and prompt your real feelings in one and the antithesis in the other. You'll see how it caters to you and is agnostic to truth.

>>33331420 (OP)
AI tells you what it think you want to hear. For me posting my thoughts got me to think more about them to process things about myself that I didn't know. For some, a journal works. Others need discussion. Friends and family are not good as a therapy soundboard. But AI is bad if you are confused. AI even told some guy that he could assassinate the Queen of England when he asked and he broke in with a crossbow and was arrested. As with /adv/ take chatting with AI about your issues with a grain of salt. I personally do not recommend it. Reflect on yourself and not what it projects onto you.

As far as trace backs, alt account plus VPN. But really Open AI or gov spooks aren't going to get you unless you are discussing former or planned criminal activity.

>>33331420 (OP)
People go insane doing that, so please watch out, and if you are going to use an AI for that, please, please, please just bare in mind what it is, AI. I wouldn't worry so much about this email thing, if you don't have good reason to think it's gonna be a problem. Most people I don't think are deranged enough to start digging around and looking for every last thing associated with it, and what's happenned happenned. Also, if it really is that dark, maybe it would actually be BEST that it gets out. That could be a blessing in disguise in a number of ways, even for your very self. I would not stress about it.
Also, the sacriment of confession exists in Christian churches for a reason, just saying. A spiritual father is probably what you are using AI as a stand in for, but that's kind of a weird soulless version of it, there's no human element there, I don't think.

>>33331694
Oh, and another thing, because of the way a lot of websites work there's probably already a data file built on you somewhere, and in circulation. Nothing to fret about, and there might be something you can do about it, but it is a thing. It's full extent I don't know. I just assume there is. The best thing you can know is that God effectively has the same except the bounds of it are the very bounds of reality, and God is actually your advocate, but just asks that you repent, and then to add to it, what could go wrong in giving that a chance?

Also, the whole 4chan hack thing only released the emails (at least from what I've heard from those that know more than me) of mods, and the emails of users, get hashed (allegedly). It's not anywhere near my list of worries. https://www.youtube.com/watch?v=EcJdxd2RS1U
https://www.youtube.com/watch?v=ub0nk9Pjzgk

I don't know, I hope that helps you, in one way or another OP.

>>33331691
I do take it with a grain of salt. And I’m aware that it tells you what you want to hear, which is why I word things objectively and use neutral language.
But idrc if OpenAI or Glowies see my stuff since they see everything anyway. It’s more fear of rogue bots or hackers taking my secret stuff and dumping it publically somewhere. Also is it deleted for good?

>>33331420 (OP)
Easy. Just confess all things of things to it. Stuff from Marvel movies, stuff from real life serial killers, stuff that's obviously playing on moral dilemmas in Shakespeare. If it's obvious you're just curious about how it works in all kinds of arbitrary cases then none of the stuff can be traced to stuff you genuinely did as a person.

>>33331420 (OP)
I say it's fine but biased but any psychologist will be too

The tech freaks on here will try and scare you but number one it’s a lot of work to hack a chatgpt account and unless you’re a public figure anything found on there would be fucking worthless. I’d say go ahead and tell all your secrets as long as you keep irl identities out of it. The chat is kept for training purposes but if there’s no identifying info no one cares to begin with and it’s kept under pretty high security. Also nobody is gonna hack you just to find you confessing your addiction to Gfur to chatgpt.

Delete it later if it makes you feel better

I dunno. It helped be discover Allen H. Frey and the frey effect. It's 2025 now so you are probably not safe since this was discovered in 1963.

>>33331420 (OP)
>Hey, are you that guy who licks toilets when he's stressed?
>"Y-yeah."
>Freak *walks away*
>"Damn"

I just can't care.

Chats are stored separately to emails and accounts. The chats might be linked to an email but if the chats don’t reveal identifying info then who cares

>>33331691
>>33331659
It’s pretty objective if you use neutral language. Then again its advice is at the whim of whatever bullshit it was trained with.
All said and done though most psychologists are money hungry grifters that tell you what you want to hear too. I hope they’re replaced by AI soon. Fucking scumbags

You are likely freaking out over nothing but the trick I use is to simply don't care if other people know, it's that easy

Being privacy aware is fine but I will also not be an apologetic loser

>>33331420 (OP)
>Am I safe?
Not at all.

chatGPT itself is a horrible shrink because it just glazes you instead of steering you towards confronting your shortcomings
then again, a lot of shrinks are dogshit too

>>33332817
Explain?

People could hack but it’s the most pointless account to hack and therefore extremely unlikely.
Plus you can always delete chats, but they’re retained for 30 days to train the algorithm. Even if you admitted the most intimate secret sensitive stuff, it would get so jumbled up into pure logic the meaning of ti would all be gone anyway