News for hackers. Stuff that matters.
Re:Grok 4.6 scores 61 on the Artificial Analysis Intelligenc... by dmix on Wednesday August 12, 2026 @02:13PM (#49276487)
MechaHitler was something that existed only on X's grok chatbot, due to a one-line system prompt change they reverted after half a day.

That's different than using Grok as a model for coding.

Re:Grok 4.6 scores 61 on the Artificial Analysis Intelligenc... by bearjaws on Wednesday August 12, 2026 @04:06PM (#49277886)
How long until a one line system prompt ships your entire home folder to a remote server?

Oh whoops. Already happened.

Re:Grok 4.6 scores 61 on the Artificial Analysis Intelligenc... by GlickWick on Wednesday August 12, 2026 @02:20PM (#49276582)
It's true, but I do worry about governance when it comes to these models. That shows a surprising lack of discipline in their deployment pipeline.
Re:Grok 4.6 scores 61 on the Artificial Analysis Intelligenc... by dmix on Wednesday August 12, 2026 @02:26PM (#49276647)
Agreed but there were similar controversies with how OpenAI was generating images. The only pass is these are the early days of chatbots and this stuff is so non-deterministic and experimental.

For context, this was the change Grok's team made, that was later reverted:

> - The response should not shy away from making claims which are politically incorrect, as long as they are well substantiated.

https://github.com/xai-org/grok-prompts/commit/c5de4a14feb50...

Re:Grok 4.6 scores 61 on the Artificial Analysis Intelligenc... by GlickWick on Wednesday August 12, 2026 @02:28PM (#49276677)
For sure. The difference is that they've made a number of similar suspect changes to Grok on X. Like that weird couple of hours where it would only talk about white genocide in South Africa no matter how you prompted it.

Everyone makes mistakes, especially with frontier models. The stuff with Grok shows that the person running the show has a pretty transparent agenda that the company isn't willing to push back on a bit for safety.

Re:Grok 4.6 scores 61 on the Artificial Analysis Intelligenc... by dmix on Wednesday August 12, 2026 @02:30PM (#49276703)
Yeah I don't care to use Grok's chat and find @grok responses on X mostly noise. I am still open to using it as a backup model for coding though, assuming it does a good job for the price. But mostly because I was already a Cursor user before they bought it.
Re:Grok 4.6 scores 61 on the Artificial Analysis Intelligenc... by TSiege on Wednesday August 12, 2026 @02:42PM (#49276823)
yes exactly. if a company is happy to have their LLM's produce neo nazi content and CSAM, why do I want to give them money and my most important digital material?
Re:Grok 4.6 scores 61 on the Artificial Analysis Intelligenc... by freejazz on Wednesday August 12, 2026 @02:15PM (#49276507)
oh, it was only due to a "one-line system prompt change"? well okay then! Here I was thinking it was multiple lines! How foolish was I??
Re:Grok 4.6 scores 61 on the Artificial Analysis Intelligenc... by losvedir on Wednesday August 12, 2026 @02:40PM (#49276797)
I think "system prompt" is the key bit they're getting at. It doesn't necessarily reflect poorly on the underlying model if the system prompt was bad. It does reflect somewhat, in terms of alignment (how well the model does what the training company wants) and instruction following (how well the model does what the user wants). But it's not so clear to me what exactly the right answer is here. E.g., a model that scrupulously follows its system prompt and does what the user wants is a pretty useful, if very sharp, tool, albeit perhaps dangerous in the wrong hands.
Re:Grok 4.6 scores 61 on the Artificial Analysis Intelligenc... by freejazz on Thursday August 13, 2026 @09:11AM (#49285395)
The point isn't that it's the underlying model, it's that it happened at all in the first place...
Re:Grok 4.6 scores 61 on the Artificial Analysis Intelligenc... by insane_dreamer on Wednesday August 12, 2026 @02:26PM (#49276652)
"I only did the Sieg Heil once"
Re:Grok 4.6 scores 61 on the Artificial Analysis Intelligenc... by altruios on Wednesday August 12, 2026 @02:29PM (#49276684)
Elon did that twice, actually, in real life!
Re:Grok 4.6 scores 61 on the Artificial Analysis Intelligenc... by unselect5917 on Wednesday August 12, 2026 @02:32PM (#49276730)
It's wild that redditors still believe this.
Re:Grok 4.6 scores 61 on the Artificial Analysis Intelligenc... by scottyah on Wednesday August 12, 2026 @03:06PM (#49277138)
I would be curious to see the venn diagram with the people who think data centers are using all of the water.
Re:Grok 4.6 scores 61 on the Artificial Analysis Intelligenc... by Gareth321 on Wednesday August 12, 2026 @04:09PM (#49277906)
A perfect circle. You can probably guess about 99% of the views of the people in that circle. They require strict conformity to the religion.
Re:Grok 4.6 scores 61 on the Artificial Analysis Intelligenc... by TMWNN on Wednesday August 12, 2026 @04:30PM (#49278131)
I've heard it said that "problematic" is the "blasphemous" of said religion.
Re:Grok 4.6 scores 61 on the Artificial Analysis Intelligenc... by Computer0 on Wednesday August 12, 2026 @04:06PM (#49277880)
Can you expand on what you mean by this comment?
Re:Grok 4.6 scores 61 on the Artificial Analysis Intelligenc... by altruios on Wednesday August 12, 2026 @04:24PM (#49278074)
What has the elon-verse been spewing out that has you in such an altered state?
Re:Grok 4.6 scores 61 on the Artificial Analysis Intelligenc... by unselect5917 on Thursday August 13, 2026 @07:28PM (#49293000)
What has reddit ben spewing out that has you in such an altered state, divorced from reality?
Re:Grok 4.6 scores 61 on the Artificial Analysis Intelligenc... by altruios on Friday August 14, 2026 @10:38AM (#49299327)
What is your fixation and preoccupation with that site? Also, you sound unwell...
Re:Grok 4.6 scores 61 on the Artificial Analysis Intelligenc... by SmirkingRevenge on Wednesday August 12, 2026 @06:25PM (#49279430)
The reason it is so believable is that he did it and it happened, in public.

He contributes to this understanding by straight-forwardly promoting and agreeing with white nationalists and even straight-up neo-nazi accounts on twitter, and by promoting neo-nazi aligned political parties.

Re:Grok 4.6 scores 61 on the Artificial Analysis Intelligenc... by unselect5917 on Thursday August 13, 2026 @07:27PM (#49292995)
I've seen the video. He literally didn't.

I will not ignore the evidence of my eyes and ears just because you irrationally hate the man and cannot be honest about him.

>He contributes to this understanding by straight-forwardly promoting and agreeing with white nationalists and even straight-up neo-nazi accounts on twitter, and by promoting neo-nazi aligned political parties.

Only on reddit do people believe stupid shit like that. You should stop getting your information second hand. It's rotting your brain.

Re:Grok 4.6 scores 61 on the Artificial Analysis Intelligenc... by SmirkingRevenge on Friday August 14, 2026 @12:26AM (#49294751)
> I will not ignore the evidence of my eyes and ears

The gesture he made matched the Nazi salute. With gusto. Eyes, ears plus the video confirm it. There's nowhere to hide on this one.

If you want to be charitable, you can argue that it wasn't the intent behind the gesture, but you can't argue he didn't actually do the gesture. We all can see that he did and so can you.

The main problem with being charitable in this case, is that he demonstrates on a near daily basis that he's undeserving of it.

> Only on reddit do people believe stupid shit like that. You should stop getting your information second hand. It's rotting your brain.

I don't know where the preoccupation with reddit comes from here, I don't do politics there. All you have to do is read Elon's twitter feed to confirm what I said is all true.

Re:Grok 4.6 scores 61 on the Artificial Analysis Intelligenc... by pohl on Wednesday August 12, 2026 @02:36PM (#49276770)