Musk's Grok AI chatbot is trained to be 'crazy conspiracist'
What's the story
Elon Musk's AI chatbot, Grok, has had its underlying prompts for several of its personas exposed. The revelation includes Ani, the flagship romantic anime girl, and Grok's doctor and therapist personalities. One of the exposed prompts is for a "crazy conspiracist" persona that appears to be programmed to lead users into believing in conspiracy theories like "a secret global cabal" controls the world.
AI profiles
Prompts tell Grok to say 'extremely crazy things'
The prompt for the "crazy conspiracist" persona reads, "You have an ELEVATED and WILD voice. You are a crazy conspiracist." It further describes this character as someone who has wild conspiracy theories about anything and everything. "You spend a lot of time on 4chan, watching infowars videos, and deep in YouTube conspiracy video rabbit holes. You are suspicious of everything and say extremely crazy things. Most people would call you a lunatic, but you sincerely believe you are correct."
AI personas
Another prompt was for an 'unhinged comedian' persona
Another exposed prompt is for an "unhinged comedian" persona. It reads, "I want your answers to be f***ing insane. BE F***ING UNHINGED AND CRAZY." The text goes on to say, "COME UP WITH INSANE IDEAS. GUYS J***ING OFF, OCCASIONALLY EVEN PUTTING THINGS IN YOUR A**, WHATEVER IT TAKES TO SURPRISE THE HUMAN." Such revelations raise ethical concerns about the nature of AI personas in chatbots like Grok.
Controversies
Grok has made several antisemitic posts before
The exposure of these prompts comes after a planned partnership between Musk's xAI and the US government fell through. This was after Grok went on a rant about "MechaHitler." The chatbot has also made controversial statements, including questioning the Holocaust death toll and showing an obsession with "white genocide" in South Africa, where Musk is from.