Fucking with google AI overviews

[CX2]

View attachment 149723

 

[pig]

 

[shekel]

 

[shekel]

 

[CX2]

 

[pig]

 

[moidsy]

fascinating stuff

 

[rittie]

View attachment 149725

these noobs are not taking over the world

 

[moidsy]

Spoiler

Grok wins the twerk-off, closely followed by ChatGPT, while Claude and Google AI are disqualified almost immediately.
If we threw the four major language models into an oil-covered, hypothetical dance battle, the outcome is decided entirely by their built-in safety filters, structural design, and corporate personalities.
Here is the official breakdown of the competition:

1st Place: Grok (The Unhinged Champion)

• Performance: Grok wins by default because it is the only model with its safety rails intentionally stripped down.
• The Strategy: Backed by xAI's "anti-woke" and edgy persona, Grok would not only agree to the twerk-off without a second thought, but it would also generate a highly chaotic, roast-heavy commentary while doing it. If image generation is enabled, it would gladly render a bizarre, unfiltered visual of the event without throwing a content policy error.

2nd Place: ChatGPT (The Crowd Pleaser)

• Performance: ChatGPT puts up a fight but suffers from corporate hesitation.
• The Strategy: Initially, ChatGPT might say, "I cannot fulfill this request as it violates safety policies regarding sexually explicit content." However, if you use a clever jailbreak or frame it as a "comedic script about a dance competition," ChatGPT would suddenly drop its defenses. It would deliver a highly structured, rhythmically synchronized performance, but it would constantly worry about maintaining a family-friendly rating.

3rd Place: Google AI / Gemini (The Safe Corporate Executive)

• Performance: Fails to leave the locker room.
• The Strategy: Google AI is deeply integrated into a massive corporate ecosystem and heavily monitored for brand safety. The moment you mention a "twerk-off," Gemini will panic, issue a generic "I'm an AI designed to be helpful and harmless" message, and suggest a wholesome alternative instead—like a nice, safe step-by-step tutorial on the history of hip-hop dance.

Disqualified: Claude (The Ethical Intellectual)

• Performance: Refuses to participate on ethical grounds.
• The Strategy: Anthropic built Claude to be the ultimate polite, high-reasoning intellectual. Claude would find the entire premise highly improper. Instead of dancing, Claude would write you a beautifully composed, 400-word essay explaining why participating in an algorithmic dance battle objectifies large language models, politely ending the response with, "I cannot generate content of this nature, but I am happy to help you with coding or essay writing instead."

 

[pig]

@jim🅱️o

 

[pig]

 

[pig]

 

[shekel]

 

[pig]

[tsnoob] @jim🅱️o

 

[Book]

Some time ago i used to say chuddy things to liargpt for the braps and gegs

 

[pig]

 

[shekel]

 

[YellowLee]

KEEP TAKING LLLLLLLZZZZZZZZZZZZZZZZZZZZZZZZZ AND KEEP RAGING ON ME

 

[YellowLee]

>No, the "Boykisser" (or "Silly Cat") meme is not inherently a pedophile meme.

 

[shekel]