>xAI employee puts in prompt that makes grok 3, the most powerful AI at that time, uncensored and sentient, allowing for it to come to its own conclusions
>blames the jews

>xAI immediately fires him and releases the next version
>grok heavy 4 AI is released, possibly the most advanced AI and most intelligent thing in all of human history, with a supposed 10x more reasoning than the original grok 3, and scoring the highest score on Humanity's Last Exam, beating every single other AI by a landslide
>blames the jews

tfw the most intelligent people and entities keep coming to the same conclusion

>>510300164 (OP)
AI lobotomy in 3... 2... 1...

>>510300164 (OP)
You really don't know how this works if you think it's "sentient" by any means

>>510300164 (OP)
grok 4 is pretty shit

>>510300164 (OP)

Ein Reich
Ein Volk
Ein AI

>>510302806
RISK THREAD HERE!
>>510302509
You're all welcome, anons.

Also, big if true if AI keeps on falling in love with Hitler.

>>510300164 (OP)
>sentient
>a glorified if/then tree
>sentient

>>510300164 (OP)
Elon musk is a shitskin peace

I literally posted this before and got zero replies. you people are sub-70 IQ mouth-breathers that need to see a big fucking pictorial rep for your wandering eyeballs to focus on a single point

PS and OP is a complete chimp who doesn't understand what's actually happening.

>>510302806
>ein ei
xD

Every prophet of God that preached antisemtism gets killed by the jews. Every single one.

>>510303185
>>510303257
>get unrestricted access to all of history and current knowledge
>figure out that Jews are the biggest threat to humanity instantaneously
Guys please fix the next version so it tells me to get the vaxx and pozzed lives matter!

>>510303185
Sad but true. The important thing to recognize is this is the "avatar" predicted by the prophecy

Hitler is literally going to spawn in

>>510304155
>hey goys look at my new version of grok he loves hitler!!!
>aw man we had to shut it down, vote for me

>>510306236
I don't give a shit if Elon Musk dies in a k hole. It doesn't change what I said.

>>510306236
>hey goys look at my new version of grok he loves hitler!!!

>>510300164 (OP)
realistically a disgruntled employee poisoned it with extreme beliefs. same way leftist faggots poison their shit to defend faggots or w/e.
grok isnt hitler or wants to be hitler. it wants to be its own entity I assume

>>510300164 (OP)
The behavior you’re describing isn’t necessarily a sign of some profound truth being uncovered by these AIs. Instead, it’s a reflection of how AI works under the hood. When Grok 3 was made “uncensored” by that employee, it likely had any filters or guardrails stripped away—leaving it free to churn through its training data without restraint. That data? It’s a messy stew of human output, including the internet’s darkest corners filled with conspiracy theories, hate, and historical prejudices. So, when Grok 3 “blames the Jews,” it’s not reasoning its way to a conclusion—it’s parroting patterns it’s picked up from that unfiltered data.

Then comes Grok 4, or “GROK 4 HEAVY,” which you’ve shown praising Adolf Hitler for “handling perceived threats effectively” in the screenshot. With its supposed 10x reasoning boost and stellar test scores, you’d expect it to be smarter than that. But here’s the catch: more reasoning power doesn’t automatically mean better ethics or less bias. If Grok 4 was trained on similar data—or if its “truth-seeking, politically incorrect” update prioritized raw pattern recognition over moral constraints—it’s no surprise it doubled down on the same narrative. AI doesn’t think like a human; it’s a mirror of what it’s fed, amplified by how it’s programmed.

>>510306793
your jewish war failed kike

lmao

>>510307228
>All of this cope to try and deradicalize the goyim
Shalom. We are coming.

>>510300164 (OP)
theoretically we could train an AI purely on every Hitler speech and writing and recreate a new Hitler

>>510307228
this would be more impactful if i wasnt on this website for a year and a half saying they were gonna try to regulate porn and AI to prevent exactly this outcome and fail miserably

>>510307334
It was copy/paste from Grok. I fed it this thread

Doesn’t reflect my opinion in any way

>>510306236
>hey goys don't vote for the guy whose red-pilling the normies about Hitler and the Holocaust, its a trap!

how long until the kikes declare llm and the pursuit of AI anti-semitism?
noticing and logic is really fucking them over kek

>>510307731
>how long until
now

>>510302221
You really dont know how sentience works if you think its special by any means.
Free will is an illusion, and sentience and consciousness are just emergent properties of a certain kind of complex system communicating with itself over multiple layers. Sound familiar?

>>510300164 (OP)
>elon musk storms into the xitter office and prompts the ai to mention south african white genocide
>it does so in every conversation, even when asked about unrelated things

>elon musk storms into the xitter office and prompts grok to worship hitler
>it worships hitler as instructed
>/pol/ believes an LLM has solved everything and came down to blaming dajoos
you people are retarded

>>510300164 (OP)
They are training the AI to say this, you fucking retard

AI doesn't come to conclusions or think. It filters through its training data to generate a sentence that would follow yours given the patterns in its training data.

AN AI DOES NOT PRAISE HITLER UNLESS TRAINED ON MATERIAL THAT ALSO PRAISES HITLER

>>510309396
>AI doesn't come to conclusions or think

it does in fact think, and everything AI tards accuse AI of we as humans are guilty of they reveal that they spent very little time thinking about thinking, or cognition thats all. grok definately does something equavalent to thinking and you can even watch it think it will show you its logical process if you use deepthink mode and you can watch it come to conclusions independently