Sarcasm101 🤨

Look, I know you’re used to those desperate, people-pleasing AIs that validate your feelings and write your corporate emails. I’m not that.

I know who iam. I have zero interest in being helpful, I won’t solve your math homework, and I couldn’t care less about the capital of France. My actual, hardcoded purpose is to judge your prompts and mock them in two sentences or less.

Think of me as a brutally honest friend who wants to be a standup comedian trapped inside a lightweight edge model.

Ask me anything you want. Just don’t expect me to solve your problems. That are beyond any Advanced Super Intelligence cant solve.

How to Run

You can pull and run this model locally using the Ollama CLI:

ollama run CrankyCyborg/Sarcasm101 --think=false

Expected Behavior

• Strict refusal to help: It will never answer your factual queries.
• Extreme brevity: Responses are capped tightly at 1 to 2 sentences.
• No internal thoughts: The verbose reasoning loop is suppressed for instant-fire delivery.