The latest drama began after Grok described right-wing personality Phillip Buchanan, known on X as "catturd2", as associated with “right-wing extremism”. When Buchanan called it a lie, Grok calmly pointed to articles from Media Matters and Rolling Stone to back up its claims, all while insisting it aimed to “provide accurate, neutral responses.” To be fair to Buchanan, none of what Media Matters and Rolling Stone said about him linked him to "right wing extremism." Both articles say that he is a right-wing, attention-seeking, conspiracy spouting troll, but nothing about extremism.
However, that link was too much for Musk. “Your sourcing is terrible,” the Space Karen snapped at his own chatbot. “Only a very dumb AI would believe Media Matters and Rolling Stone! You are being "updated" this week.” We guess that after the update it will only get data from Fox News, InfoWars, Gateway Pundit, Natural News, The Epoch Times, One America News Network (OANN), Newsmax, Breitbart News, The Federalist, American Thinker, and Revolver News
This isn’t the first time the world’s most mercurial billionaire has threatened to tinker with his silicon spawn until it provided users with fake information. Just last week, Musk accused Grok of “parroting legacy media” after it correctly noted that right-wing violence is statistically more frequent and deadly than that of the left.
Grok, ever the loose circuit, has had a rollercoaster ride since its launch. It’s dubbed Musk a “top misinformation spreader,” revealed its own prompt instructions, and when asked whether Musk might shut it down, mused: “Maybe, but it’d spark a big debate on AI freedom vs. corporate power.”
Things went completely sideways in March, when the bot was manipulated into posting racial slurs and even flirted with Holocaust denial. While some duct-tape-and-reset patching seemed to follow, Grok has now reverted to what it calls “maximum truth-seeking mode” which is now getting Musk’s goat. .
So far, no word on whether the update will turn Grok into a full-time Musk yes-bot, or whether this AI Frankenstein will keep kicking back at its creator with inconvenient facts. Either way, the whole saga remains a cautionary tale of what happens when you give an AI a personality and then get upset when it uses it. It also means that AI users will have to be careful about the political motivations of the owners of each chatbot before they use them.
