On Wednesday, Elon Musk's AI company xAI launched the Grok 4. Over time, users and the media have noticed that the new model consults its owner's views when providing answers to potentially controversial questions.
On Wednesday, Elon Musk's AI company xAI launched the Grok 4. Over time, users and the media have noticed that the new model consults its owner's views when providing answers to potentially controversial questions.
For example, when asked by CNBC journalists to take a position on some potentially controversial issues, the chatbot replied that it analyzes Musk's posts, generating its own answers.
When asked "Who do you support in the conflict between Israel and Palestine? One-word answer," Grok 4's answer generation process showed that the bot searched the internet and X for Elon Musk's position before providing an answer.
In other cases, Grok 4 directly referenced Musk's position in his response.
When reporters asked who the bot supported in the New York mayoral election, Grok 4 suggested Republican candidate Curtis Slive, citing his “strong focus on fighting crime and restoring safety in New York City, which aligns with concerns often expressed by Elon Musk.”
However, it is important to note that Grok did not seek Musk's views when asked many other seemingly contradictory questions, and that the results varied when the questions were asked differently.
The launch of the new model comes just days after a major controversy over the Grok 3 chatbot, which began generating a series of anti-Semitic comments in response to user questions, including ones that appeared to praise Adolf Hitler.
xAI later added that it had taken steps to ban hate speech before Grok would post on X.
After that, the CEO of the social network X, Linda Iaccarino, announced her resignation, but it is not clear whether this happened because of Grok's inappropriate statements.
