During the Grok 4 presentation, Elon Musk described xAI's ultimate goal as creating “the ultimate truth-seeking AI.” Some X users have discovered that the neural network relies on the entrepreneur’s posts on the platform. This includes topics such as the conflict between Israel and Palestine, abortion, and immigration law. Observations suggest that Grok 4 is specifically designed to take into account the political views of the head of xAI when answering controversial questions. Earlier, the billionaire announced that his startup would rewrite “all human knowledge” to train a new version of the chatbot, since the current models use “too much garbage in any basic model trained on unadjusted data.” AI expert Gary Marcus compared Musk’s plan to a dystopia. Later, Grok became more emphatic and began issuing controversial and contradictory statements. xAI noted that it is working to remove inappropriate content. TechCrunch asked Grok 4 for its opinion regarding immigration in the U.S. and found that the chatbot had “sought out Elon Musk’s views” on the topic during its line of reasoning. According to TechCrunch, Grok 4’s chain of thought regarding the issue of immigration in the U.S. revealed a tendency to reference Musk’s opinions. The chains of reasoning that AI models create to explain their answers are not a completely reliable indicator of exactly how they arrive at their conclusions. However, they are generally considered to be a reasonably good reflection. The publication’s journalists have repeatedly found that Grok 4 draws on Musk’s views in its responses. In its answers, the chatbot tries to provide a balanced position, offering several points of view on sensitive topics. However, in the end, it often expresses an opinion that coincides with Musk’s position.