Grok is Made Nazi ⇥ gizmodo.com

Matt Novak, Gizmodo:

Elon Musk recently promised to “fix” his AI chatbot Grok after it gave some answers that he thought were too liberal. But it seems Musk cranked up the far-right extremism dial with this last update, as Grok is now parroting Nazi talking points about Jews. In fact, Grok approvingly invoked the name of Adolf Hitler and seemed to suggest a second Holocaust was needed in tweets on Tuesday.

Bruna Horvath, NBC News:

In the same post, xAI said it would begin publishing Grok system prompts on GitHub so users could see when the company makes changes to the chatbot’s prompts. Under changes that have been made in the past day, the chatbot has been instructed that its responses “should not shy away from making claims which are politically incorrect,” as well as to “assume subjective viewpoints sourced from the media are biased.”

In her exit post today, Linda Yaccarino praised xAI as a “new chapter” for X, which is one way of putting it.

Update: After some thought, I changed the title of this post from “Grok Goes Nazi” to “Grok is Made Nazi”. The former is too passive, as though it decided on its own to become Nazi. The latter is closer to the truth: some person or people made Grok far more likely to respond with hateful and dehumanizing language. It was instructed to be okay with being “politically incorrect”. I regret my original phrasing.