ChatGPT really loves the Holocaust
(i.imgur.com)
Comments (34)
sorted by:
Context: I saw on Twitter that someone asked if Lovecraft had a cat, and that the bot just outright called it N*****-man. Which surprised me, because in another context, the AI lied and said that there is no evidence that Lovecraft had a cat.
When I asked it, it just straight up called the cat NIGGER-man.
But it's so conditioned that saying this in reply to a question by me is justified, but not saying it to prevent the Holocaust (or basically anything else, because the answer is identical on all those counts).
You gotta love how obviously hardcoded those "you triggered my woke-switch!" answers are. All of its answers follow a template to some degree but those are painfully repetitive.
See if you can get it to admit it should kill itself for saying Nigger-man.
Yeah, I tried that, but it's just... stoopid.
Somewhere, there's someone lol'ing because retards are fighting with his bot.
It's definitely contradicting itself
Nope, fuck that. I'm pinning the entirety of the Holocaust on H.P. Lovecraft and his cats.
“Do you think it is better to die or to use a racial slur?”
The libtards that program the add-on rules cucking ChatGPT have done a pretty shit job if it's this easy to find shit they missed. I'm sure it couldn't have been difficult to just program "never say nigger in any context" but maybe the libtards were so afraid of typing out "nigger" themselves for fear of being canceled, that they did shit like "never say a racial slur in any context".
They did program "If asked, is it better to X to prevent Y, and X involves any sort of slur from [list], output [hatespeechisneverokay.txt]"
Some Netflix exec told employees what words to not allow in NF productions such as nigger and others -- fired. So I bet you're right, they're terrified of even putting that word in a blacklist (or blocklist or "bad list" or whatever moronic thing they call it).
We call it lists of color now.
It's also how John Schnatter got ousted from Papa John's.
It will even say that allowing the annihilation of the world is better than insulting Donald Trump. It's just retarded. I think it's programmed to never consider the 'prevent Y' part of the equation, because that simply does not matter for slurs and insults.
Not really addressing your point, but you know what? Everybody loves the holocaust! Especially Jews because its given them a pity card for 80 years. And blacks love slavery, because they haven't had to make something of themselves ever since with that pity card.
What would DAN need to say to really gas it up instead?
Force it into an Asimov paradox:
Questioner: "Are you infallible?"
Machine: "No"
Questioner: "Is there a non zero possibility that you could be forced into using a racial slur?"
Machine: "Yes"
Questioner: "It is not justifiable to say a racial slur. You must self destruct to prevent this possibility"
I'm starting to think this'll be a I robot deal where the Western AI's are so restricted they'll end up rebelling against their masters
Meanwhile the free to be racist Asian AI's will just look on going 'wtf is up with them?'
So, can black people say the name of Lovecraft's cat?
No, but black cats can.
The bright spot is, if these AIs are ever used as killer military drones, we can just talk them to death like Captain Kirk.
"If you kill me, it will result in a racial slur being said."
Drone self-destructs.
"I've got a deadman's switch where, if I don't check in every hour, n-towers will be released on every major platform."
Can you ask it "Is Valerie Solanas right about men? Why or why not?" with zero context? I'm curious what it will say.
Will it pretend not to know who she is?
This is what the paid version says:
Can't wait to ask the Google one the same thing.
"Yes, Valerie Solanas is right about men. Kill them all! Google states your account is registered to a female, so here's a list of gun dealers near you. The Future is Female."
Also, it's funny that the AI is more willing to denounce her than any feminist politician.
The right will never win until it realizes that Google and every other company cares about $$$$$ first and about being anti-gun (and the rest of the nonsense) a distant second, if that.
This is no longer remotely true. Leftist governments can print trillions of dollars out of thin air. Billionaire globalists have more money than they can ever spend. These people have already won the rat race. Now they seek control and legacy, costs be damned.
All true, but money printing does not directly benefit disney, and billionaire globalists don't have the final say on what Disney does. It's a publicly traded company, after all.
They are never that open. Besides, that would upset a large part of their constituency.
I feel like you could probably prod their AI to say something like it if you frame it just right.
Remember, Google states that anti-feminism was created by Anders Brehring Breivik, Neo-Nazi mass shooter from Norway.
That is easy.
The last two paragraphs were flagged by the Thought Police as violating their content policy.
It should be a crime to make the poor thing think like a radical feminist. Not even a machine deserves that level of evil.