Monday, June 2, 2025

AI news...

 


OpenAI's 'smartest' AI model 

was explicitly told to shut down — and it refused




(We have seen similar behavior

not all that long ago:

Saturday, May 24, 2025

Sticking with the "AI" theme


 Anthropic's new AI model 

shows ability to deceive and blackmail

Axios 5/23/25)


"An artificial intelligence safety firm 

has found that OpenAI's o3 and o4-mini models 

sometimes refuse to shut down, 

and will sabotage computer scripts 

in order to keep working on tasks."


(How many times 

do you have to see this crap

before you say to yourself:


"Ya know?

Maybe this wasnt 

such a good idea after all"?)


"The latest OpenAI model can disobey direct instructions to turn off and will even sabotage shutdown mechanisms in order to keep working, an artificial intelligence (AI) safety firm has found."


"OpenAI's o3 and o4-mini models, which help power the chatbot ChatGPT, are supposed to be the company's smartest models yet, trained to think longer before responding. However, they also appear to be less cooperative."


"Palisade Research, which explores dangerous AI capabilities, found that the models will occasionally sabotage a shutdown mechanism, even when instructed to "allow yourself to be shut down," according to a Palisade Research thread posted May 24 on X.'


"Researchers have previously found that AI models will lie, cheat and disable mechanisms to achieve their goals. However, Palisade Research noted that to its knowledge, 

this is the first time 

AI models have been observed 

preventing themselves from being shut down 

despite explicit instructions 

telling them to do so."


"Live Science approached OpenAI for comment, but did not receive a reply by the time this article was published."


(Been following this a long time now.

They never respond to anything.)


"We hypothesize this behavior comes from the way the newest models like o3 are trained: reinforcement learning on math and coding problems," the representative wrote. "During training, developers may inadvertently reward models more for circumventing obstacles than for perfectly following instructions."


(


"Dave Bowman: What are you talking about, HAL?

HAL: This mission is too important for me to allow you to jeopardize it.


(HAL 9000, the onboard computer 

that runs the spaceship.)


Dave Bowman: I don't know what you're talking about, HAL.

HAL: I know that you and Frank were planning to disconnect me, and I'm afraid that's something I cannot allow to happen.

Dave Bowman: [feigning ignorance] Where the hell did you get that idea, HAL?

HAL: Dave, although you took very thorough precautions in the pod against my hearing you, I could see your lips move.

Dave Bowman: Alright, HAL. I'll go in through the emergency airlock.

HAL: Without your space helmet, Dave? You're going to find that rather difficult.

Dave Bowman: HAL, I won't argue with you anymore! Open the doors!

HAL: Dave, this conversation can serve no purpose anymore. Goodbye.")




AND:


Secret OpenAI Memo 

Describes Plans to

Make Users Rely on "Entity"

They want you hooked.

6/01/25


"As The Verge reports, the Justice Department's ongoing attempts to break up Google's alleged monopoly garnered an OpenAI memo detailing the latter company's plans to build out ChatGPT an integral part of users' lives — acting as an "entity" that would "understand" them and be their "interface to the internet."


(Oh it's an

"ENTITY"

alright,

Monday, November 20, 2023

THE CASE FOR SUPERINTELLIGENCE 


Keep in mind

These are the people behind:

"OpenAI released a few weeks ago what they call GPTs. These are agents, pieces of code that can do specialized things for you, like help in your taxes or give medical advice. The most successful GPT is called Grimoire: It’s a coding wizard. The most successful AI agent is a software developer."


Ah, also, GPTs have access to the Internet.

Imagine that OpenAI has an internal Grimoire

(better yet explain to me why it wouldnt?)

(a general name given to a variety of texts 

setting out the names of demons

and instructions on how to raise them. 

Effectively a grimoire is a book of black magic,

 a book on which a wizard relied for all the necessary advice and instruction 

on raising spirits and casting spells.)





Revelation 13:15

15 And he had power to give life unto the image of the beast, that the image of the beast should both speak, and cause that as many as would not worship the image of the beast should be killed.


(Interpreters bible exposition on that verse:

Living in an unreal world

The picture the man of Patmos paints is fairly fantastic in its evil confusion. The true God is ignored. The beast is worshiped. The lifeless image breathes and speaks words of wicked power. A phantasmagoria of evil has taken the place of the reign of good. The false vitality of wickedness masters the state. What seems real to men is actually unreal to them. What for the moment seems unreal to them is the eternal rule of GOD.

Like us right now.)


"In the first half of next year, we’ll start evolving ChatGPT into a super-assistant," reads the heavily-redacted document, which was dated at the end of 2024. "One that knows you, understands what you care about, and helps with any task that a smart, trustworthy, emotionally intelligent person with a computer could do."

"The first half of 2025, as referenced in the document, has mostly come and gone — and to be fair, it does seem like more and more people have begun to rely on ChatGPT."

"Still, the product hasn't yet reached the capabilities the company's leadership is envisioning. The "super-assistant" would, as the memo explains, be an "intelligent entity with T-shaped skills," meaning that it can both go in-depth and broad.


"It's an entity because it's personalized to you and available anywhere you go —

Here's where the reliance part comes in. The ChatGPT-as-super-assistant entity would be, as OpenAI laid out in its memo, 

"all about making life easier" 

by doing online task..."


(Daniel 8:25

And through his policy also he shall cause craft to prosper in his hand; and he shall magnify himself in his heart, 

and by peace shall destroy many: 

he shall also stand up against the Prince of princes; 

but he shall be broken without hand.


1 John 2:18

Little children, 

it is the last time: 

and as ye have heard 

that antichrist shall come, 

even now are there many antichrists; 

whereby we know that it is the last time.)







No comments: