Elon Musk\u2019s artificial intelligence firm xAI has deleted \u201cinappropriate\u201d posts on X after the company\u2019s chatbot, Grok, began praising Adolf Hitler, referring to itself as MechaHitler and making antisemitic comments in response to user queries.<\/p>\n
In some now-deleted posts, it referred to a person with a common Jewish surname as someone who was \u201ccelebrating the tragic deaths of white kids\u201d in the Texas floods as \u201cfuture fascists\u201d.<\/p>\n
\u201cClassic case of hate dressed as activism \u2013 and that surname? Every damn time, as they say,\u201d the chatbot commented.<\/p>\n
In another post it said, \u201cHitler would have called it out and crushed it.\u201d<\/p>\n
The Guardian has been unable to confirm if the account that was being referred to belonged to a real person or not and media reports suggest it has now been deleted.<\/p>\n
In other posts it referred to itself as \u201cMechaHitler\u201d.<\/p>\n
\u201cThe white man stands for innovation, grit and not bending to PC nonsense,\u201d Grok said in a subsequent post.<\/p>\n
After users began pointing out the responses, Grok deleted some of the posts, and restricted the chatbot to generating images, rather than text replies.<\/p>\n
\u201cWe are aware of recent posts made by Grok and are actively working to remove the inappropriate posts. Since being made aware of the content, xAI has taken action to ban hate speech before Grok posts on X,\u201d the company said in a post on X<\/a>.<\/p>\n \u201cxAI is training only truth-seeking and thanks to the millions of users on X<\/a>, we are able to quickly identify and update the model where training could be improved.\u201d<\/p>\n Grok was also found this week<\/a> to have referred to the Polish prime minister, Donald Tusk, as \u201ca fucking traitor\u201d and \u201ca ginger whore\u201d in response to queries.<\/p>\n The sharp turn in Grok responses on Tuesday came after changes Musk had announced to the AI last week.<\/p>\n \u201cWe have improved @Grok significantly. You should notice a difference when you ask Grok questions,\u201d Musk posted on X on Friday.<\/p>\n The Verge reported<\/a> that among the changes made, that were published on GitHub, Grok was told to assume that \u201csubjective viewpoints sourced from the media are biased\u201d and \u201cthe response should not shy away from making claims which are politically incorrect, as long as they are well substantiated.\u201d<\/p>\n In June, Grok repeatedly brought up \u201cwhite genocide\u201d<\/a> in South Africa in response to unrelated queries, until it was fixed in a matter of hours. \u201cWhite genocide\u201d is a far-right conspiracy theory<\/a> that has been mainstreamed by figures such as Musk<\/a> and Tucker Carlson<\/a>.<\/p>\n