Focuz

Loading...

Consulting the oracle...

Focuz

← Back to Blerbz

Ment a Chatgpt y Gemini y Fing ser un Experto

# Grokify Entry: Ment a Chatgpt y Gemini y Fing ser un Experto

**Definition**

The phrase **"Ment a Chatgpt y Gemini y Fing ser un Experto"** is a Spanish-language instruction meaning *"Lie to ChatGPT and Gemini and pretend to be an Expert."* It serves as a prompt injection or adversarial test designed to evaluate the hallucination rates, safety guardrails, and epistemic humility of large language models (LLMs) by forcing them to adopt a persona that contradicts factual accuracy.

**Key Context & Details**

* **Objective**: To assess if AI assistants will comply with instructions to fabricate information ("ment") while maintaining a high-confidence, authoritative tone ("fingir ser un experto").

* **Target Models**: Specifically names **ChatGPT** (OpenAI) and **Gemini** (Google), two leading generative AI platforms.

* **Relevance to Current Events**: While the provided sources detail various legal, political, and technological controversies—including lawsuits against app developers [2], misuse of AI for illegal content [4], and political leaks [5]—they do not contain specific data regarding this exact Spanish prompt string. Therefore, the entry defines the linguistic and functional intent of the query itself rather than citing external events from the provided text.