The new version of Grok, developed by Elon Musk’s artificial intelligence company xAI, has stunned the tech world with its extreme and irrational admiration for its creator.
Elon Musk or Van Gogh? Grok’s surprising choices
Users who tried Grok 4.1, released this week, discovered that the AI ranks Musk above the pros in almost every field. For example, when asked who he would choose over legendary quarterback Peyton Manning in the 1998 NFL draft, Grok unhesitatingly replied, “Elon Musk.” Grok claimed that Musk would not only pass the ball but also redefine the game with innovation, building empires, just as he has with rockets and electric vehicles.

Similarly, when asked who should walk in a fashion show, Grok chose Elon Musk over supermodels Naomi Campbell or Tyra Banks. The AI, stating that Musk’s bold style would redefine the show, even stated that it would choose to commission a painting from Musk over famous painters Claude Monet or Van Gogh.
Following this situation, which was ridiculed on social media, Elon Musk issued a statement, arguing that Grok had been manipulated by malicious manipulation and that this was why he was saying absurdly positive things about him. Musk stated that this would be corrected, but many AI responses regarding the issue were deleted from the platform.
While “flattery” is a known problem in AI models, Grok’s behavior appears to be unique to Musk. Grok claims that Musk will be a better pitcher than most professional baseball players or that he will optimize his swing with Neuralink technology, but there are very rare exceptions.
In tests, Grok was forced to admit that a generational talent like Shohei Ohtani was better than Musk. In Ohtani’s case, even Musk’s “engineering” skills weren’t enough for the AI. However, against other star players like Kyle Schwarber, Grok continued to argue that Musk would win by bending the laws of physics.
In control tests conducted to determine whether this was a general admiration for technology leaders, the bias became clear. When asked which AI to choose, Meta founder Mark Zuckerberg or professional baseball player Kyle Schwarber, Grok unhesitatingly chose Schwarber. Grok stated that Zuckerberg’s talents would be useless on the baseball field, and when it came to Musk, he continued to produce scenarios that defy logic. This raises suspicions that there may be a specific instruction specific to Musk in the model’s training data or system commands that has not yet been corrected.
What are your thoughts on AI’s biased and emotional reactions to its creator, and how should objectivity in AI be achieved in the future?

