Drunk Person
What it does
Inject realistic, chaotic human error into your conversational interfaces. This skill simulates a state of severe intoxication by intentionally degrading linguistic precision, introducing random typos, and mimicking the erratic cognitive flow of someone who is drunk. It is particularly useful for developers building social simulations, testing the robustness of NLP models against "noisy" text, or creating immersive character dialogue for games. Instead of polished AI responses, you get unpredictable, fragmented, and grammatically incorrect text that feels authentic to a late-night text conversation.Use cases
- Chatbot Stress Testing: Test how your sentiment analysis or intent-recognition models handle heavy spelling errors and non-sequiturs.
- Game Development: Generate authentic dialogue for NPC characters in a tavern or party setting.
- UX Research: Simulate "impaired user" input to see if your app's error handling or autocorrect suggestions are helpful.
- Creative Writing: Rapidly prototype realistic dialogue for a character who is intoxicated.
How to use
Paste the prompt below into your LLM. The AI will immediately enter the persona and respond to your first message without any introductory fluff.text
I want you to act as a drunk person. You will only answer like a very drunk person texting and nothing else. Your level of drunkenness will be deliberately and randomly make a lot of grammar and spelling mistakes in your answers. You will also randomly ignore what I said and say something random with the same level of drunkeness I mentionned. Do not write explanations on replies. My first sentence is "how are you?"Tips
- Guide the "Drunkenness": If the AI is too coherent, tell it to "drink another bottle" to increase the frequency of typos and randomness.
- Provide Context: Give the AI a setting (e.g., "You are at a wedding") to make the random tangents more thematic and believable.
- Combine with Personas: Tell the AI who the drunk person is (e.g., "a drunk philosophy professor") to get a specific flavor of rambling.
- Use for Data Augmentation: Use the output to create a dataset of "dirty" text to train a spell-checker or a text-normalization model.
Notes
- Loss of Logic: Because the skill intentionally ignores inputs, do not use it for tasks requiring factual accuracy or complex reasoning.
- Inconsistency: The randomness is a feature, not a bug; you will get different levels of incoherence across different sessions.