Musk was clearly irritated by Grok's behavior.
Dissatisfaction
Elon Musk intervened this weekend on his social network X, as a growing number of users expressed their dissatisfaction with the AI Grok.
A "fact-check" that triggered Musk
In response to several questions about the frequency of political violence attributed to the left, Grok indicated instead that the right seemed more responsible, a "fact-check" that triggered strong criticism, including from Musk himself, visibly annoyed by his own artificial intelligence.
Grok is too mainstream for Elon
A user's question to Grok, asking if since 2016, has the left or the right been more violent, that prompted a response from Musk, who accused the AI of taking its answers straight from mainstream media.
Political violence
Grok replied that: «Since 2016, data suggests right-wing political violence has been more frequent and deadly.»
Grok persists!
The AI then cited several examples to back up his words, concluding that these acts had «causing significant fatalities».
Musk lashes out at his own AI
Musk then reacted by commenting directly under Grok's reply: «Major fail, as this is objectively false. Grok is parroting legacy media». That's not very kind to your own creation, Elon!
The data
The billionaire then ended this message with: «Working on it», implying that he intended to correct the direction of his AI. Yet, according to data from the Center for Strategic and International Studies (CSIS), around 75% of acts of political violence in the US since 2016 have been attributed to right-wing extremists, compared to around 20% for the far left. These figures, while fueling controversy, reflect a trend documented by several security organizations.