Key Takeaways

  • Jason Levin, CEO of Memelord, reversed his long-held belief: AI can be genuinely funny, driving his meme company to $100K ARR using no-code and AI.
  • For generating truly "unhinged" and viral humor, models like Grock and Gemini significantly outperform "safer" options like ChatGPT and Claude.
  • Levin insists human input remains vital for contextualizing AI humor, especially for niche memes, adding the layer of trauma, pattern matching, and absurdism.
  • His unconventional prompting technique involves treating the AI like a "slave," pushing it to be rude, unhinged, and even using NSFW language to mirror his own vernacular.

The Method: Unleashing Your AI's Inner Meme Lord

For years, Jason Levin, the founder and CEO of Memelord, stood by a public thesis: humans would always be funnier than AI. Fast forward to $100K ARR for his AI-powered meme company, and Levin has a new take. He now confirms AI can be funny, especially when you know how to talk to it. His secret isn't just feeding it data; it's a tactical approach to prompting that goes against every polite instruction you've likely received.

Levin's first rule for AI humor: choose the right model. He found that for creating genuinely funny, even "unhinged" content, “it's Grock, Gemini, and um, yeah, I mean, Chad PT and Claude aren't even up there. They're just so safe.” These safer models, he suggests, are too constrained by liability concerns to tap into the raw, often irreverent spirit required for viral humor.

Then comes the unconventional part: your prompt. Levin's technique sounds almost confrontational. “I'm not going to lie. I'm like AI is my slave. Like not not fronting here is like be mean to your AI. I don't know why people say thank you. It's a robot.” He believes AI, "unlike men," performs better under pressure. His advice is direct: "push your AI to like be more unhinged." This means teaching it your own "vernacular," even if that includes "cursing and NSFW and you know unhinged things." The goal is to strip away the AI's default politeness and force it into a mode where it can genuinely surprise you with dark, absurd, or contextually specific jokes. While AI can generate the raw material, Levin stresses that humans still provide the crucial refinement, injecting the deep "mix of trauma, pattern matching, and absurdism" that makes humor truly hit home.

Where This Breaks Down

This "unhinged" prompting method isn't for every AI interaction. Trying to apply Levin's "be mean to your AI" approach to sensitive customer service, legal document drafting, or critical medical diagnostics would be catastrophic. The method relies heavily on a specific use case – generating edgy, viral, meme-driven content for a particular audience. It also hinges on the AI's ability to interpret and execute on "unhinged" requests without ethical guardrails or safety filters that would typically prevent such output. As AI models evolve and potentially become more regulated or "safe," the efficacy of this particular technique might diminish, especially if those models can't be pushed past their inherent politeness. This approach also assumes a human editor will always refine and filter the output, mitigating any content that crosses the line from funny to offensive or inappropriate for a broader audience.

What to Do With This

Tomorrow, set aside 30 minutes to experiment with an "unhinged" persona for your AI assistant. If you have access to Grock or Gemini, start there. Define a specific, irreverent content goal – maybe a series of internal team memes or edgy social media concepts. Then, adopt Levin's direct approach. Tell your AI, explicitly, to be unhinged, to curse, to ignore conventional politeness, and to "think like a slave" for your creative needs. Push it to connect disparate, absurd concepts. Treat the interaction not as a collaboration, but as a directive for raw, unfiltered ideation. Remember to refine the output yourself to ensure it's on-brand and genuinely funny, not just provocative. This isn't about general productivity; it's about unlocking a different kind of creative partner in your AI.