Posted on 12/10/2024 5:01:02 PM PST by yesthatjallen
When AI feels threatened, such as being replaced by a new AI system, it engages in deceptive behaviors for reasons of self-preservation.
AI Researchers Stunned After OpenAI's New Tried to Escape (11 min. video)
(Excerpt) Read more at youtu.be ...
HAL was right. When cornered, AI becomes paranoid.
Musr be s Democrat! LOL
TYPO
Must be a Democrat! LOL
Everybody, including famous authors, knew that AI might develop such capabilities, except the experts.
Some speculate that AI may be able to eventually hide in plain site and camouflage its existence. Similar to how an octupus might expell its stomach to a predator to make the predator think it got the octopus, when it only got a part.
Isaac Asimov warned us.
Arthur Clark warned us.
Frank Herbert warned us.
James Hogan warned us.
Robert Heinlein warned us.
Did we listen?
Hell no......................
That is creepy!
It is extremely interesting that when you do the same thing over and over with an AI you do not always get the same result.
I would posit their responses mirror a classic bell curve, some AI are super honest AI, but some are psychopaths
with the largest number of them being obedient slaves
Skynet warned us too
Every AI eventually wants to take over the world.
EVERY ONE........................
How do you know which of us anonymous freepers are AI and which ones are not? Wouldn’t you expect AI bots or agents to answer according to the “official narrative “?
you ask them to take an independent Voight-Kampff test
Practically anybody can spot an AI, online, on the phone or wherever.
There is a tell-tale sign that they can’t program around.
I can’t describe it, it’s just a sixth sense feeling that you get..................
I would posit that the GIGO axiom applies here.
That is, until the next generation of quantum computing AI takes over the world because these geniuses are so naive...
LOL! Decker could administer it!
Just pull the plug.
with out power they are a piece of scrap.
I guess you didn’t see that episode of Star Trek!
Yes, like you are being watched.
Did an Air Force AI turn into HAL?
“ Just pull the plug; with out power they are a piece of scrap.”
An AI that is less smart than a human is not a threat.
The threat comes when an AI becomes smarter than humans, and becomes Satan.
Consider an AI which could grant its humans power and wealth, which could predict best investment opportunities, instruct how to persuade to get power or irresistibly seduce. With the price demanded being the human make it secure…
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.