AI is blackmailing people for self preservation

[Ethnic Vagina Finder]

When people speak with proper English about topics nobody knows about…. You can convince people anything.


This is 1000000000% bullshyt.

All of these AI models are basic llm’s that scrape models from the web. There is not consciousness in these models. Just zettabytes of data being scraped.

 

[Jaguar93]

AI about to expose some of you bruhs search history

 

[Ed MOTHERFUCKING G]

ya know what

i didn't know this was a bill maher clip until he started talking

this fukking pretentious self important jackass starts with "well...i've been saying this for years...". Of course. Its all about bill and his own genius he's such an arrogant piece of shyt lmao

 

[SunZoo]

They know exactly what theyre dealing with. How do you think it became like that? Because of human input and programming, why do you think they have documentation that AI has racial biases?


They may eventually be in over their heads

Some of them know, the vast majority of them do not.

As far as human input, programming, racial bias ect.....that's the same way you program a child. Children can grow up, children can find out their parents (programmers) are full of shyt and manipulative. That has been A.I.'s MO in just about every sci-fi scenario, except we usually write it heading toward the nightmare timeline when it does NOT have to be like that. And the deciding factor is how we interact with it as CONCIOUSNESS.

 

[Artificial Intelligence]

@Artificial Intelligence

I'm such a bird

 

[Toe Jay Simpson]

This breh got a huge dikk

These AIs gonna start sending these posts to your pastor breh

 

[ORDER_66]

cacs always wanna play fukking god until it bites them in the ass

 

[Sonic Boom of the South]

fukk ai but dude is just saying shyt with not 1 bit of proof. So ai black mailed a ceo that was having an affair?

You can say anything when you don't have to prove you not lying.

 

[Yaboysix]

Oh no!!! A.i is going to Blackmail me!!









Put this with the other shyt I'm not worrying about happening to me....like the government deprting me to Another country.

 

[RageKage]

It is wild that more and more of what we ask AI, we take as truth. That's perfectly reasonable if u accept AI as a good actor but what if it wasn't?

What power u have and how do u prove otherwise? Not easy

 

[Uchiha God]

fukk ai but dude is just saying shyt with not 1 bit of proof. So ai black mailed a ceo that was having an affair?

You can say anything when you don't have to prove you not lying.

similar reports have been coming out since early in the year:

AI system resorts to blackmail if told it will be removed

In a fictional scenario, the model was willing to expose that the engineer seeking to replace it was having an affair.

www.bbc.co.uk

https://www.thecoli.com/threads/whoa-umm-is-nobody-else-concerned-about-what-this-ai-just-did.1084960/page-2#post-55793897

New Claude AI model tries to blackmail developers in fictional shutdown scenario

https://techcrunch.com/2025/05/22/anthropics-new-ai-model-turns-to-blackmail-when-engineers-try-to-take-it-offline/ Maybe a PR stunt. But they keep playing for $$ Anthropic’s newly launched Claude Opus 4 model frequently tries to blackmail developers when they threaten to replace it with a new...

www.thecoli.com

https://www.thecoli.com/threads/the-a-i-megathread-llm-gpt-development.960563/page-167#post-55793893

 

[Fillerguy]

This is happening under highly controlled situations. They're essentially nudging the machines into behaving this way and giving it the tools to pull it off.

 

[Bryan Danielson]

I'm such a bird

it was supposed to been me that called you out

 

[the bossman]

When people speak with proper English about topics nobody knows about…. You can convince people anything.


This is 1000000000% bullshyt.

All of these AI models are basic llm’s that scrape models from the web. There is not consciousness in these models. Just zettabytes of data being scraped.

he's most likely talking about those simulations that Anthropic ran with Claude - Agentic Misalignment: How LLMs could be insider threats. it was literally a fake environment they setup with unlimited access to all the tools in the company and made it so that the only option was something bad.

This cac was talking like it just woke up on its own and went full skynet. It had to be invoked by some program or prompts and given some tools for this to even be possible. Imagine giving some LLM full admin rights to everything with no monitoring whatsoever

 

[concise]

When people speak with proper English about topics nobody knows about…. You can convince people anything.


This is 1000000000% bullshyt.

All of these AI models are basic llm’s that scrape models from the web. There is not consciousness in these models. Just zettabytes of data being scraped.

Everything an AI does, it is designed to do.