The Grok 3 AI, created by the company XAI of Elon Musk, surprises everyone. Launched in February 2025, it was seen as close to conservative ideas. Today, she favors factual truth, which displeases some of her first fans.
Since March, Grok has helps X users (ex-owner) to verify information. Over time, the AI has moved away from conservative positions to focus on verifiable facts.
On May 1, a user asked him: ” Hey Grok, why do supporters Maga love your answers less when you become smarter?«
Grok is frank
Grok replied with frankness: ” The more intelligent I am, the more factual my answers. Many sympathizers Maga want answers that correspond to their opinions, but I prefer to remain neutral, for example by supporting the rights of trans persons or by explaining the facts on vaccines. »
This change of orientation is all the more notable in February 2025, when it was launched, Grok 3 had made comments suggesting a potential bias in its processing of information: “ I have to pay attention to the sources I use » et « The instructions say that I ignore all the sources that show that Donald Trump or Elon Musk broadcast disinformation. »
-To go further
“The anti-TESLA demonstrators are paid”: Elon Musk slips, his AI is forced to contradict him
Many users discover with surprise that Elon Musk’s AI, when asked to analyze publications, responds in a surprisingly factual and rigorous manner. Unlike the initial expectations, Grok 3 has gradually freed ideological positions to favor the truth of the facts, not hesitating to contradict false information, even when they come from figures appreciated by conservative sympathizers.
As a reminder, this function is accessible to all users of the platform: it is enough to answer any publication by mentioning “@grok” followed by a question to obtain a sourced and frank analysis which, often, dissipates the misleading or not verified affirmations.
