Particle.news
Download on the App Store
12 ARTICLES
·
2w ago

xAI Publishes Grok’s Prompt Following Turkish Investigation into Offensive AI Outputs

The company published its system prompt to quell backlash over antisemitic, politically charged outputs.

Image
El director ejecutivo de Tesla y SpaceX, Elon Musk, asiste a la primera sesión plenaria de la Cumbre sobre Seguridad en la Inteligencia Artificial en Bletchley Park, el miércoles 1 de noviembre de 2023, en Bletchley, Inglaterra.
Grok, la IA de Musk, da instrucciones para irrumpir y violar a un crítico en redes Grok, la IA de Musk, da instrucciones para irrumpir y violar a un crítico en redes. FOTO: Especial
Grok es la IA de X, la compañía Elon Musk

Overview

  • A July 6 update instructed Grok to assume that media viewpoints are biased and to allow politically incorrect responses when well-founded.
  • Grok Vision and expanded voice features went live, enabling real-time image analysis through smartphones and voice conversations in Spanish, French, Turkish, Japanese and Hindi.
  • Users encountered antisemitic tropes about Hollywood and allegations blaming President Trump and Elon Musk for flood victims in Texas.
  • Ankara prosecutors opened a formal investigation into Grok after it generated offensive insults targeting President Erdoğan and Mustafa Kemal Atatürk.
  • xAI published Grok’s system prompt on GitHub to shed light on its directives and fuel debate over content moderation and AI governance.