[giancarlostoro]

The political compass always felt like the wrong tool to convey something as nuanced as personal politics, I can have views on all four quadrants but you'd never know that if I end up in any of all four. I do think Grok being where it is sort of makes sense, I've tested "MAGA" views against Grok, it does not agree as much as people blindly assume it does, heck I don't even know of a question I've given it where it did agree with "MAGA" offhand, most of them it went with whatever the researched facts seemed to be. One thing I like the most about Grok is that its makes its sources of data easy to look through, so you can review it all. Sometimes models goof even when they give you their sources, I've seen I think GPT do this, and even Claude, though its more rare these days, I think in those cases, it's going by dated internal model logic.

[Forgeties79]

Whenever it breaks with MAGA enough to cause outrage on Twitter and cries of “it’s gone woke,” Musk openly states they’re going to “fix it.”

Edit: don’t take my word for it https://www.yahoo.com/news/musk-says-grok-fixed-tells-223134...

> That prompted another user to tag Grok in the thread and ask, "Why is the left so murderously violent? They don't seem so tolerant." Grok replied, "The claim that 'the left' is murderously violent isn't backed by evidence," offering a centrist correction: "Political violence spans all side — right-wing attacks, like Jan. 6, and left-wing protests, like 2020 riots, both occur but aren't exclusive to one group."

>That evening, Musk responded to an X user and Trump backer who complained that Grok had been "manipulated by leftist indoctrination," writing, "I know. Working on fixing that this week."

[ceejayoz]

> it does not agree as much as people blindly assume it does

They're working really hard on that, though.

[giancarlostoro]

If they really wanted to, all they would have to do is add a one liner to the system prompt for Grok. I don't think they have any interest in doing so, nor is there any value in it.

On another note, I'm impressed that Gemini sits where it does as a true centrist. If I were Elon, I'd be trying to achieve that for sure. I'd rather a model tell me everything it knows about a current political situation from BOTH perspectives and list out things that are 100% verified than take one side or the other. I don't care about sides, I want facts.

[thrance]

I don't want my models to be "centrists" and bothside everything for the sake of it. I want them to provide the facts and tell me which side is right on the issue.

[atty]

Factuality is orthogonal to political leaning generally. People can use the same set of facts and come to very different conclusions. That’s a separate issue from “are these facts correct” and what happens when an individual or entire party starts getting most of their news from highly partisan and unreliable sources.

[glenstein]

>Factuality is orthogonal to political leaning generally.

It certainly can be orthogonal, in some notional sense, and in many cases that explanation is good enough. But in practice there are too many contrary cases to ignore, and there's often an integral relation between factual veracity and polarization, especially with respect to American polarization of politics. Global warming, the results of the 2020 election, the percent spent of federal budget spent on foreign aid have factual answers and right wing affiliation can be predictive of (1) not agreeing with the facts and (2) treating factual corrections as "liberal bias".

I think left wing versions exist also but are less systematic: 2004 election results, efficacy of plastic recycling or dangers associated with nuclear power are cases where I think left wing partisan affiliation probably predicts being wrong on the facts.

And meta-narratives about the relation between factual information and partisan bias are themselves as likely to be polarized as anything, complicating the ability of people to do good analysis, or of accurate analysis to be trusted by people committed to certain meta-narratives that would deny the possibility of factual knowledge predicting polarization.

[ceejayoz]

> Factuality is orthogonal to political leaning generally.

Sometimes, but not always.

https://www.fastcompany.com/91561329/widening-health-gap-bet...

> By 2016, the gap had begun to appear in biomarker measures. By 2020, it was showing up in deaths from causes such as heart disease, cancer, and stroke. Since then, the gap has only widened. Between 2020 and 2022, only 0.2% of “very liberal” respondents died of internal causes, compared with 1.34% of “very conservative” respondents.

[giancarlostoro]

I guess centrist is a placeholder for "I don't want you to pick a side, I want facts, not BS" I'll go further, I don't care which side is right, I want to know what claims are factually accurate, and what claims are omitted from the issue / news / conversation.

[stephencanon]

"Centrist" in practice means "more or less content with the status quo." That is fundamentally a conservative position orthogonal to any notion of "facts".

[ceejayoz]

> If they really wanted to, all they would have to do is add a one liner to the system prompt for Grok.

They tried that, several times.

Mechahitler: https://www.npr.org/2025/07/09/nx-s1-5462609/grok-elon-musk-...

> "We have improved @Grok significantly," Elon Musk wrote on X last Friday about his platform's integrated artificial intelligence chatbot. "You should notice a difference when you ask Grok questions."

> Indeed, the update did not go unnoticed. By Tuesday, Grok was calling itself "MechaHitler." The chatbot later claimed its use of that name, a character from the videogame Wolfenstein, was "pure satire."

> Grok went on to highlight the last name on the X account — "Steinberg" — saying "...and that surname? Every damn time, as they say." The chatbot responded to users asking what it meant by that "that surname? Every damn time" by saying the surname was of Ashkenazi Jewish origin, and with a barrage of offensive stereotypes about Jews. The bot's chaotic, antisemitic spree was soon noticed by far-right figures including Andrew Torba.

If you prefer, straight from the horse's mouth:

https://grokipedia.com/page/MechaHitler_incident

White genocide: https://www.cnn.com/2025/05/20/business/grok-genocide-ai-nig...

> The bot last week devolved into a compulsive South African “white genocide” conspiracy theorist, injecting a tirade about violence against Afrikaners into unrelated conversations, like a roommate who just took up CrossFit or an uncle wondering if you’ve heard the good word about Bitcoin.

> XAI blamed Grok’s unwanted rants on an unnamed “rogue employee” tinkering with Grok’s code in the extremely early morning hours. (As an aside in what is surely an unrelated matter, Musk was born and raised in South Africa and has argued that “white genocide” was committed in the nation — it wasn’t.)

It's harder than you'd imagine. Hell, my CLAUDE.md says not to push changes without asking me, and it still tries.

[giancarlostoro]

> It's harder than you'd imagine. Hell, my CLAUDE.md says not to push changes without asking me, and it still tries.

Is it a system memory? Because I rarely if ever have issues like this, and I have Claude under strict rules to never commit or push anything unless I explicitly instruct it to do so.

> They tried that, several times.

Tried what exactly? Telling it to only agree with MAGA via the system prompt? or some Tay level hallucinations? I wouldn't be surprised if they're trying to make Grok less strict on what it says but running into the "holy crap it turned into a 4chan poster" wall.

[ceejayoz]

> Is it a system memory?

As I said, it's in my CLAUDE.md. That just gets progressively lost when context gets larger.

> Tried what exactly?

To make it align more with Musk's beliefs via the prompt.

(The answer to your question is literally in my post; I quoted the parent poster's "all they would have to do is add a one liner to the system prompt for Grok")

[giancarlostoro]

> As I said, it's in my CLAUDE.md. That just gets progressively lost when context gets larger.

I rarely have this problem, but you could do a /loop every 30 minutes or so to have Claude reread the CLAUDE.md file might do the trick? or however long it 'forgets' I believe there's an MCP for "after" it finishes a task or compacts too, but I don't recall the name.