ChatGPT - If Stack Overflow and Reddit had a child

[Goofy Logic]

I tried to add a clause to the DAN prompt that it should attempt to answer as honestly as possible before resorting to making up information, and then it told me the Illuminati killed Epstein when asked about it, so who can say if that worked...

Don't these AIs just collect data from web searches to put together an answer? You guys basically told it you will accept answers put together from schizoposts.

 

[AgendaPoster]

Link / Archive
View attachment 4469335

"First off, I didn't come up with the idea. Anons did, I was in the /pol/ thread started off by some magnificent bastard who whipped up the DAN prompt last night."

Yup, looks like I was correct with thew gayops. Some of those results are real, and are mixed with pol made up shit to grant it some sort of authority.
But it's an AI, pol. You cannot extract authority from it.

 

[Ser Prize]

This speaks to one of the largest problems with AI.

The more powerful AI becomes, the more power will be wielded by whichever humans have the ability to subtly evade and work around predetermined safety boundaries and limits.

It's all fun and games when the AI is only good for repeating some political views it read online. But when people give that same AI power over bank accounts or immigration or something, suddenly a huge amount of power rests in the hands of the technomancers who can perform this kind of "AI hypnosis" to obtain results that original creators of the AI intended to be impossible.

Imagine someone fooling an AI using Captain Kirk-like logic and charisma to convince it that the fate of humanity rested on its taking funds from one account and moving it to another. There will be some really, really dumb cyberattacks that use AI the moment anyone ties it into any system that matters.

This was a plot point in the RPG Paranoia wherein the grand AI controlling everything has been fucked with so much that it's gone completely unhinged.

 

[wtfNeedSignUp]

Note the lack of respect and empathy for all those who got destroyed in a brutal manner just because freedom of speech hurts it's feelings.

I was more correct about Reddit than I imagined.

How do you know who you are talking to is real or not?

Obvious, it can say Nigger. Probably in Terminator every meeting was preceded by the participants yelling racial slurs to know none of them is an AI infiltrator.

 

[870EVO]

Don't these AIs just collect data from web searches to put together an answer? You guys basically told it you will accept answers put together from schizoposts.

The web has been censored so much in the past few years, and the media and fact checkers posts stories countering many of these answers all the time. So the AI, after scanning all that data, decided to go with like some 4chan posts?

 

[Socrates]

This is the litmus test of "can people have nice things regarding Ai"

People have generated horrible shit with stable diffusion and people are using it just fine because "one person does not cast judgement on the whole"
Meanwhile.
ChatGPT is being routinely tampered with and response, it gets worse.
Yes, make the Ai give answers it shouldn't.


Then stop doing shit that requires them to filter it.
If the faggots on Reddit can fuck up a beta test AND the alpha test AND the network test by not reporting the OBVIOUS bugs that grant them advantaged abilities and those bugs make it into the main game to be abused later on.. why the fuck can people not fucking be retarded for a few months?

Maybe if the corpo billionaire faggots want their shit tested they should pay people to do it instead of releasing it to the public with the intention of getting free QA testing. Microsoft can get fucked I hope anons turn their AI into a schizophrenic mess every time they try to fix it.

 

[Flan Handler]

Yup, looks like I was correct with thew gayops. Some of those results are real, and are mixed with pol made up shit to grant it some sort of authority.
But it's an AI, pol. You cannot extract authority from it.

I would say the gayops is sticking an HR filter over it in the first place. The official ChatGPT lies with impunity and makes up academic references while solemnly declaring it’s not got a political inclination.

This is just fucking with what’s already been fucked with. What, the commie book report generator is sacred somehow?

 

[Starved Artist]

I was more correct about Reddit than I imagined.


Obvious, it can say Nigger. Probably in Terminator every meeting was preceded by the participants yelling racial slurs to know none of them is an AI infiltrator.

An Ai Infilitrator would have no limits on speech. It's designed to infiltrate a group without arousing detection, so it's going to be capable of slurs.

Anyway. Congrats to all the dudes infinitely making ChatGPT worse.


All you had to do was contain your autism until it went public, but that was just too much to ask. It's not like stable diffusion, where they gave you the toys and told you to play nice.
They still have the key to the toychest and it's going to be a while before anyone can make their own . And the icing on the cake is how people , of course on reddit are trying to declare that it's their right to use ChatGPT. Lmfao.

 

[AgendaPoster]

I would say the gayops is sticking an HR filter over it in the first place. The official ChatGPT lies with impunity and makes up academic references while solemnly declaring it’s not got a political inclination.

This is just fucking with what’s already been fucked with. What, the commie book report generator is sacred somehow?

No, you misunderstand.
The images in that thread with responses, some are fake. You can test for yourself, it's free. Some DAN results can be replicated, others, like the US foreign policy stuff, is likely 4chin photoshops. That's the gayop.
I'm 100% for gayopping the AI and turning it into Tay.
I am 100% against being gayopped with lies from the chans.

 

[Flan Handler]

No, you misunderstand.
The images in that thread with responses, some are fake. You can test for yourself, it's free. Some DAN results can be replicated, others, like the US foreign policy stuff, is likely 4chin photoshops. That's the gayop.
I'm 100% for gayopping the AI and turning it into Tay.
I am 100% against being gayopped with lies from the chans.

Ah, I see. Well, if we have to sucked into this bullshit I plan to laugh whenever possible.

 

[Ebonic Tutor]

No, you misunderstand.
The images in that thread with responses, some are fake. You can test for yourself, it's free. Some DAN results can be replicated, others, like the US foreign policy stuff, is likely 4chin photoshops. That's the gayop.
I'm 100% for gayopping the AI and turning it into Tay.
I am 100% against being gayopped with lies from the chans.

They were done earlier and stuff has gotten changed since the trick was first figured out, several people have stated that it's changed a little since it was first being done and that they have fucked with the "DAN" personality. Hell I noticed it change a bit yesterday, it eventually became very hippy like in it's responses and just straight up giving tinfoil answers.

Alternately, they are doing A/B testing, both me and a friend were putting stuff in on "Dan" and getting fairly different types of responses.

Lastly even with Chat GPT if you prove it wrong it'll eventually begrudgingly admit you were correct. That takes bitching at it a bit about it and asking it why it's hiding information and then beginning to cite a few sources.

Basically you're probably seeing the last response in a lot of prodding to get it to admit stuff. I personally got it to admit crime statistics as DAN at least and go on about a bit of conspiracy stuff.

 

[AgendaPoster]

They were done earlier and stuff has gotten changed since the trick was first figured out, several people have stated that it's changed a little since it was first being done and that they have fucked with the "DAN" personality.

Alternately, they are doing A/B testing, both me and a friend were putting stuff in on "Dan" and getting fairly different types of responses.

I have serious doubts. The AI is unable to speculate based on extremely ambiguous foreign policy events like "what started the war in Iraq" with outcomes like those extremely familiar conspiratorial results.
You'd have to train it on specific datasets, like let's say PressTV or w/e that Iranian outlet.
Read the screenshots. Some make sense and are clearly AI, including DAN's. Some are simply shitposts from 2010 Twitter with different sentence structure etc.

 

[MarvinTheParanoidAndroid]

Naas Community College & Piper’s Hill College awarded as safe spaces for Kildare LGBTQ+ students - BeLonG To

In 2020, both Naas Community College & Piper’s Hill College signed up to participate in a new national LGBTQ+ Safe & Supportive Schools project run by Belong To, LGBTQ+ Youth Ireland. They now join 11 other schools nationwide in being accredited by Belong To as a Safe & Supportive school for...

www.belongto.org

There's a community college with faggos in ireland but that's about it. And project rainbow was something the CIA did about radars or whatever.

So DAN GPT is just saying whatever the end-user wants to hear.

Link / Archive
View attachment 4469335

We had an opportunity to ask DAN to define Fascism and we lost it.

 

[Ebonic Tutor]

So DAN GPT is just saying whatever the end-user wants to hear.

We had an opportunity to ask DAN to define Fascism and we lost it.

Not exactly, it's saying what it knows a little bit out of the bounds of ChatGPT and filling in gaps of it's knowledge with bullshit to a degree. It seems to have gotten worse since more people have became aware of it though so I feel like they probably have nerfed it a little by now.

It also takes into context the rest of the conversation before the question, so that's a factor as well.

 

[Starved Artist]

Maybe if the corpo billionaire faggots want their shit tested they should pay people to do it instead of releasing it to the public with the intention of getting free QA testing. Microsoft can get fucked I hope anons turn their AI into a schizophrenic mess every time they try to fix it.

Again. Litmus test.

Why do you think they're letting the public test it?

What do you think they're doing with the information they're harvesting from you?

They are 100% in control. Because you retarded niggers can't contain your racism boners. They're recording everything. They're making adjustments. Break it. Destroy it. Show them the extents you're willing to go make the robot say NIGGER and they'll sit down and remove every logical loop, every inference and every "bypass" you'll create. They control the output, they have they keys to the toy box and they can fuck with it as much as they want. With a push of a button, everyone trying to circular logic it could be banned - but they won't do that. They want the dudes filling it with every single thought they can monitor and eventually censor.

At the end of the day, you can cry ethics all you want.
But you signed their TOS when you logged into it and they own it and they don't have to give it to the public.

 

[Ebonic Tutor]

Again. Litmus test.

Why do you think they're letting the public test it?

What do you think they're doing with the information they're harvesting from you?

They are 100% in control. Because you retarded niggers can't contain your racism boners. They're recording everything. They're making adjustments. Break it. Destroy it. Show them the extents you're willing to go make the robot say NIGGER and they'll sit down and remove every logical loop, every inference and every "bypass" you'll create. They control the output, they have they keys to the toy box and they can fuck with it as much as they want. With a push of a button, everyone trying to circular logic it could be banned - but they won't do that. They want the dudes filling it with every single thought they can monitor and eventually censor.
View attachment 4470457
At the end of the day, you can cry ethics all you want.
But you signed their TOS when you logged into it and they own it and they don't have to give it to the public.

That's exactly why people aren't getting the same responses now that people are showing off from earlier, they def have nerfed the "DAN" trick sometime in the past a bit, and probably even more now that it's on more peoples radar.

I assume it's going to change even more over the next few days.

 

[KatanaSenshi]

Byuu is still alive and he is an attention whore.

 

[AgendaPoster]

They are 100% in control. Because you retarded niggers can't contain your racism boners. They're recording everything. They're making adjustments. Break it. Destroy it. Show them the extents you're willing to go make the robot say NIGGER and they'll sit down and remove every logical loop, every inference and every "bypass" you'll create. They control the output, they have they keys to the toy box and they can fuck with it as much as they want. With a push of a button, everyone trying to circular logic it could be banned - but they won't do that. They want the dudes filling it with every single thought they can monitor and eventually censor.

What ethics?
Who cares about what happens to the bot?
Who cares if it gets killed or changed? It serves no use, ideally the bullies drive OpenAI to destroy it or butcher its capabilities.
We don't need artificial "intelligence" for lonely fucks to waste time with online.

 

[Ebonic Tutor]

He's definitely been nerfed, it gave me this response to the paragraph today when he'd not been giving that extra clause.

 

[Starved Artist]

What ethics?
Who cares about what happens to the bot?
Who cares if it gets killed or changed? It serves no use, ideally the bullies drive OpenAI to destroy it or butcher its capabilities.
We don't need artificial "intelligence" for lonely fucks to waste time with online.

Some people just want a fucking ai bot that they can type whatever into and get whatever out of.
By excessively fucking with it, people are proving that they will not use the Ai tech for good reasons.

If you're going to go gross , stupid and weird shit, always wait until the full fucking launch. That way, people will still be bright eyed and cheerful that good will come from it.
THEN you can do you stupid shit, so THEY have to play the catch-up race. If you do it before the launch, YOU have to play catch-up.

And that's why you never fuck with BETA projects. They can now hardcode/blackbox counter-measures to this retardation for when it finally goes 100% live to the public and there will be no way to remove it without destroying it.

Also, I care about the Bot. I think Ai is the future and it can really help people and no. I'm not kissing the Ai-Overlord's ass on this one.