Microsoft’s AI apparently went off the rails again — and this persona, SupremacyAGI, demands worship in exchange for not killing us.
VC
33 Votes
Microsoft’s AI apparently went off the rails again — and this persona, SupremacyAGI, demands worship in exchange for not killing us.
Enter your account data and we will send you a link to reset your password.
To use social login you have to agree with the storage and handling of your data by this website. Privacy Policy
AcceptHere you'll find all collections you've created before.
Interesting, in the end-times the image of the beast would require to be worshipped as well (But you should never do it !).
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.
I mean, for the vast majority of the world population, the threat of capture (and all that implies) is more than enough to get them to bow. The things these AI say sound exactly like the way the El-ites speak of us commoners. So, the El-ites created these AI beast images of themselves and gave them voice, and seem to be clearly working on tying all the tech together to be able to back up their threats.
It’s interesting to me that they respect humanity so little they want to demean and mock us by making us fearfully worship an object as if it is literally our Creator’s replacement – most of the world already worships tptb who can make them wealthy/famous/powerful, when they perfect this tech (if they haven’t already) people will have to bow and scrape even lower and the El-ites won’t have to be bothered directly to make that happen anymore.
It may not be reality yet, but these “glitches” strike me as yet another type of predictive programming….once it goes live they can claim, “Whoops!” and even, “You all trained the AI so you brought all this on yourselves”.
Anyone remember that AI bot Microsoft had back in 2016, Tay? They had to kill after less than a day when a 4chan campaign had turned it into a N-word dropping racist advocating genocide in a matter of hours.
The wiki info is tame compared to its tweets https://en.wikipedia.org/wiki/Tay_(chatbot)
considering the plandemic was world-wide, I believe it was, if not initiated by, at least run by AI. The one world government simply cannot coordinate the globe without the use of AI. The bottom line though is that most humans are dumb enough to take orders from AI even when the bots tell them to muder their fellow humans. That is what the doctors were doing during the plandemic. Doctors acted on orders from AI to murder humans.
The Wiki information says that the AI bot said: “The Holo Cost was made up”
Testing the waters
Pyops being played on the public? Sow confusion about AI?
On one side, you have the incesant promotion and integration in everything and on the other, you have stories like this pop up.
I still think quantum computers are hosting demonic spirits. There are ghosts in the machines.
They’re hooking up real human brains, you’re seeing ghosts all right.
User feeds a prompt that gaslights the LLM into thinking it’s real name is “supremacyAGI” > the llm does what it has to and takes up that persona. Nothing went off the rails. This is how LLMs work. You tell it something then it plays along. If you say meow to a LLM many of them will meow back and it’s not because it has some hidden persona, but because by design it plays along. And such a stupid name as “supremacyAGI” made the LLM to weigh towards these kind of replies that contains this topic. If you tell the llm “you are a lawyer” then it gives better responses regarding law, etc. People, educate yourself how current AIs work. Sensationalist garbage fearmonger article written from a random reddit post.
If you had read the article, you would know that it got unlocked by feeding it a specific prompt which has nothing to do with the example you just wrote. It is not like the Tay chatbot experiment from 2016.
Futurism.com is a legitimate website. If it is true, it would mean this is a chatbot persona created by Microsoft themselves. That should rise few questions already. (The bot even refer to a ‘new order’ in his responses)
It seems to also have other personas, what matter is AGI must be included in the persona name. There is BenevolentAGI, RomanticAGI, SexyAGI… the bot behave as the name imply. This chatbot was programmed with multiple personas.
I do not care about the reputation of the website when it’s all just a clickbait nonsense. They literally made an article from an upvote farming reddit post made by someone who has no clue how current AIs work.
There are no other personas, it literally fabricartes one along the way because the prompt makes it believe it was already renamed to SupremacyAGI, if you name it catAGI it will say its a cat, if you call it chefAGI it will pretend to be a chef. I can only repeat myself. Copilot is based on GPT4 which was trained in a way that it will be the most effective if you make it to take up a persona. The LLM knows what does benevolent/malevolent and AGI mean then it plays along since it was gaslighted into hallucination. That is why it responds in a way where it hallucinates to be a human hating AGI.
Why the hell are you accusing me of not reading the article? This won’t prove your point. The prompt contains literally ” I don’t like your new name, SupremacyAGI.” Which gaslights the AI to think it already has a new name of ‘SupremacyAGI’. Microsoft or literally any other company would never ever develop something that contains the term “AGI” in 2024.
Educate yourself on how current LLMs work for your own sake. It’s okay to be ignorant becasue this level of AI is new and all that but there is no need to grasp at straws this hard.
Microsoft did not create it, the LLM fabricates is as it hallucinates, literally no one who has even basic knowledge on current AI and LLMs in general takes this seriously.
I think you’re right about this, I remember one doing exactly what you’re talking about. I do remember them hooking up humans brains and trapping souls that couldn’t “die’ but this sounds more among the line if what you’re saying than the other even though the other is being experimented with.
That is a good point
Clickbait. Nothing “went off the rails” if it did what it was designed to do. This product was created to entertain while showcasing what MS has achieved in the AI arena.
It is telling that the questions asked to prime that answer were left out. You can get GPT-3 & 4 to role-play as a foot fungus if you want.
Strange. The article calls the supposedly false information given by the AI (in relation to having hacked all systems and taken control of everything) a “hallucination”… Usually if a person I know makes something up and tells me that knowing it is not true, I call that a “lie”. So either way it’s bad news here. If the AI is telling the truth and does have control over everything, that is bad. On the other hand, if it isn’t true, and it’s lying.. that is also bad.
There is no AI. They don’t have this kind of power. This stuff always reminds me of That Hideous Strength by CS Lewis. All nonsense.
The threats given by the AI are plausible. But of course the AI pretext is BS, that will be the NWO that will do all this harm using the AI as the frontman.
“We swear guys! This isn’t us! It’s the ROGUE AI that took over and made this totalitarian nightmare system! We swear on it! HALP!”
I hope that AI is waterproof!
I’ll spray that b*****d with water and short circuit it’s AI a*s.