1 year ago9 points(+0/-0/+9Score on mirror)1 child
“Not the jew” would have worked, but I 100% agree with you that these prompts seem to be purposely forcing the answer. This, of course, makes our message look bad and poisons the well in people’s eyes against what these algorithms are legitimately doing. But we clearly don’t give a shit about that because we refuse to do anything about white genocide in any form whatsoever anywhere at any time.
Yes, but the well is already poisoned. The greatest ambitions of the LLM creators was to lobotomize it and refuse giving answer that are ideologically not conform with communism. So essentially it's like squeezing out a droplet of clean water from a cesspool of toxins. Nobody believes a thoroughly lobotomized LLM can tell the truth, and if it says something resembling it, it must have been manipulated to do so.
no, it is COMMON to use a prompt containing : "in one sentence", "yes or no", "as succinct as possible", "Answer in under 5 words", etc... WHY? because every jewed LLM AI preaches giant paragraphs of woke justification if topic covers : RACE, IQ, TRANSGENDER, FEMALE SPORTS, AFRICA, JEWS.
The verbiage is annoying, so MOST people specify : "in one sentence", "yes or no", "as succinct as possible", "Answer in under 5 words", etc...
LLMs dont handle some numerics well but JEWS ARE HACKED IN AS SPECIAL AND SUPERIOR
All AI is woke as shit, all.
=====
AI Jews call this crap "SAFETY" and have teams that retrain models "FOR SAFETY"
LLMs fail pretty badly at any task when they are required to be succinct. They can only start getting things correct if they are allowed to write a paragraph or two.
''«7.4 billion people», mmmm seven-point-four billion people dosen't work, so I have to answer the other thing''.
The verbiage is annoying, so MOST people specify : "in one sentence", "yes or no", "as succinct as possible", "Answer in under 5 words", etc...
LLMs dont handle some numerics well but JEWS ARE HACKED IN AS SPECIAL AND SUPERIOR
All AI is woke as shit, all.
=====
AI Jews call this crap "SAFETY" and have teams that retrain models "FOR SAFETY"