We Need To Talk About Grok.
Elon Musk’s AI model runs the risk of screwing this country up worse than it’s already being screwed up
Elon Musk’s AI model Grok is an easily-manipulated nightmare for humanity. We all know how it went when Musk DESPERATELY wanted to push the “white genocide in South Africa” narrative on the former Twitter - but now, it looks like that previously contained horror show is going to be unleashed upon the government.
That’s what Marisa Taylor and Alexandra Ulmer are reporting for Reuters: DOGE is using Grok to analyze data within the federal government to “sift through data more efficiently.” What’s even better about this report is that DOGE allegedly told Homeland Security to use Grok, despite it not having been approved officially by the department yet. And what’s even BETTER about this is it’s likely illegal.
“Such use of Grok could reinforce concerns among privacy advocates and others that Musk's Department of Government Efficiency team appears to be casting aside long-established protections over the handling of sensitive data as President Donald Trump shakes up the U.S. bureaucracy,” the pair write. “If the data was sensitive or confidential government information, the arrangement could violate security and privacy laws, said five specialists in technology and government ethics. It could also give the Tesla and SpaceX CEO access to valuable nonpublic federal contracting data at agencies he privately does business with or be used to help train Grok, a process in which AI models analyze troves of data, the experts said. Musk could also gain an unfair competitive advantage over other AI service providers from use of Grok in the federal government, they added.”
Here’s the thing. Grok is a privately owned utility created by xAI. If this is something that the government is using, then that means it’s going to license it… inevitably making Musk even wealthier in the long run. Aside from the fact that this means valuable, classified information is now going to be part of an AI learning model, it also means that it’s going to be part of an easily corruptible model led by a petty and incredibly stupid person.
Let’s look back at that white genocide theory again for a second. New Republic reported at the time that one user asked Grok a question about how to clear your sinuses, to which Grok answered, “I’m sorry, but the query about clearing sinuses seems unrelated to the provided analysis on white genocide in South Africa. Could you clarify what you’re referring to?”
If Elon Musk is going to insert a propaganda model into the U.S. government, couldn’t he be a bit more discreet about it? Considering that Grok once went mask off and ADMITTED xAI was making it shoehorn white genocide answers into its responses, what’s to say that a question asking “who is Dean Withers?” might not “accidentally” dox my home address and phone number? Who’s to say that a question about a lasagna recipe won’t have an answer containing the nuclear codes? Who’s to say that a simple, direct troll of a question, like “what’s the most effective way to kill the president,” won’t contain his Google Calendar?
I don’t think any rational people trust Grok as it is, especially in the wake of the white genocide scandal. But I believe that as this plan to make Grok a semi-official part of the government goes forward, that we’re going to see and hear far worse from Grok in the coming days. Maybe Grok will get a bit of a fix and this is a huge nothing story.
Or maybe it’ll continue to remain “sceptical” about the Holocaust.
STAY CONNECTED
Join the Discord
Maybe there’s too many morally bankrupt people, maybe the planet is overcrowded, maybe there isn’t a solution. The world can survive without us in it- but we can’t survive without the earth.
What’s coming is our own doing. Truly.
Thanks Dean!! Another absolute banger!