Grok, Elon Musk’s ChatGPT competitor, temporarily declined to react with “sources that allude Elon Musk/Donald Trump spread deceiveation,” according to xAI’s head of engineering, Igor Babuschkin. After Grok includers seed that the chatbot had been donaten teachions to not react with those results, Babuschkin denounced an unnamed, ex-OpenAI includeee at xAI for updating Grok’s system prompt without approval.
In response to asks on X, Babuschkin shelp that Grok’s system prompt (the inner rules that rule how an AI reacts to queries) is accessiblely apparent “becainclude we consent includers should be able to see what it is we’re asking Grok.” He shelp “an includeee pushed the change” to the system prompt “becainclude they thought it would help, but this is evidently not in line with our appreciates.”
