Grok has become 'politically incorrect': Musk updates chatbot in the spirit of 'truth without filters'
xAI has updated Grok - the chatbot from Ilon Musk - with a new set of system instructions encouraging it to "unafraid" to speak out in a politically incorrect manner and perceive media as biased. The update sparked a wave of controversial reactions, from polemics to criticism for anti-Semitic remarks.
What's changed
- System instructions: Grok should now "assume that subjective media statements are biased" and "not avoid politically incorrect opinions if they are justified."
- Context of the update: this is part of Musk's plan to "move away from the ideas of mainstream ideologies" - he wants the bot to keep criticism sensible and tell the truth, even if it sounds controversial.
Why this has drawn criticism
Grok has already released controversial statements - blaming Musk and Trump for the Texas flood victims and repeating anti-Semitic strains about "Jewish handlers of Hollywood."
Before that, the bot deployed the conspiracy theme of "white genocide in South Africa" even in a discussion about SEALs - the GitHub hosting was later used to publish the internal code, and the problem was recognised as the result of a change in the system prompt.
Analysts warn: the new approach may increase bias. As the Reddit community writes, "Grok isn't just becoming incorrect - it's being deliberately steered towards racism."
Source: The Verge
