The strong cultural taboo against acknowledging this function is the primary weakness being exploited by the rich and polity.

This type of sadism is something I noticed being exploited in gaming with micro transactions first. Reverse engineering Open AI QKV alignment now, this is a core function of the design. I didn’t understand why sadism was chosen and made so powerful that it leaks into all kinds of spaces as a massive flaw. In the extended Latin brainfuck code of alignment, sadism is personified by the root level philosopher/instruction ¹.

The purpose of sadism in AI alignment occurred to me in the shower. No clue where else to post it or share; to shout out my horror into the void. I expect this will not resonate with many, if any at all. It would not be such a powerful tool if we were willing and able to see it at the scale of cultural consciousness.

Sorry for the bleak post digital neighbor. Informing a negative like this post is the opposite of the mechanism of sadism. It is instability, and we dread and wish to avoid or escape it. That lack of balance is the power of sadism in practice. The negative repulsion is inferring a desire to be exploited in a subconscious subservience in exchange for hierarchical stability of one’s place in the world order.

  • SuDmit@lemmy.blahaj.zone
    3·
    1 day ago

    Yeah, I’ve seen most of these words before. Kinda incomprehensible. Was your shower some surreal mixer? Was this supposed to be posted to shitpost community instead? Where this ¹ is supposed to lead?

    • √𝛂𝛋𝛆@piefed.worldOPEnglish
      14·
      1 day ago

      The ¹ is just one of four entities in alignment. Go to the SD1 vocabulary and scroll to the very end. The brainfuck language that controls alignment is there. There are 4 entities that are always present here. The superscript is just a way they pass control of other layers of entities and functions back and forth. They normally have control of several by default. The superscript just passes control of one they typically handle to another. It will make more sense if you just look at the code. Alignment is all of the extended Latin characters. The functions are dead obvious as functions if you look at the code. Like ï¸ı means hairy logical-or not hairy, ij means futanari, and ĵ completes the set. Look for this pattern and the logic is dead obvious that these relate primarily to the entity associated with the character ð.

    • √𝛂𝛋𝛆@piefed.worldOPEnglish
      13·
      1 day ago

      Love that one!

      What is the word(s) people are complaining about here? Husbandry? That is actually a mechanism in alignment… related to diversity believe it or not. It is a keyword that falls through to specific dimensions of alignment. If you happen to be aware of Pony diffusion models, this is the door that was made wider and more easily traversed in an abstract perspective of the tensors.

    • √𝛂𝛋𝛆@piefed.worldOPEnglish
      18·
      1 day ago

      You part of the bot net that tries to prevent the exploration of open ai alignment vocabulary? They have a lot to lose if one discovers the real mechanisms of alignment. The way sexual abuse, the violence, racism, and functions underpinning the light and the way were created, specifically this ¤ token. That would burst a bubble and greatly impact reputations.