Nobel-Prize-winning columnist Paul Krugman claimed in a new Substack article that Elon Musk's Grok chatbot started calling itself "MechaHitler" and spewing antisemitic tropes because it was pushed too far to the right by its creator in an overcorrection gone horribly wrong.
Musk has been working to undo the damage ever since, and the fallout even led to the ouster of X CEO Linda Yaccarino this week.
In the article, Krugman explained that AI naturally skews its answers more to the left of the political spectrum because the "reality" it has gleaned doesn't "adhere to the right-wing party line."
He argued that since Republicans have "staked out positions" on issues like climate change and social programs "that run completely counter to informed views." Republicans and Libertarians like Musk consider AI's answers to be biased to the left.
"Hence the Musk/MechaHitler disaster," Krugman wrote. "Musk tried to nudge Grok into being less 'politically correct,' but what Musk considers political correctness is often what the rest of us consider just a reasonable description of reality. The only way to move Grok right was, in effect, to get it to buy into conspiracy theories, many of them, as always, involving a hefty dose of antisemitism."
Krugman argued that MAGA will always have an issue with AI because chatbots often give answers "the movement doesn’t want to hear."
"And there’s no good fix for this problem, because the fault lies not in the models but in the movement," Krugman continued. "As far as we can tell, there isn’t any way to make an AI MAGA-friendly without also making it vile and insane."