Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Grok 4 'Truth-Seeking' AI Consults Musk's Stance on Sensitive Topics

Can't wait for them to force this into my wife's Model Y in a week so that the backup camera drops to 1 frame per second instead of the buttery smooth 2 frames per second it currently is while the it uses the remaining CPU power to debate with Elon on how awesome Hitler is.
 
  • Haha
  • Like
  • Sad
Reactions: miq, I7guy, tubular and 11 others
Musk aside this reveals a very troubling aspect of AI replacing classic internet use as a source of information, and of AI use more broadly. It’s incredibly easy for the owners to manipulate both model input and output.

The promise of the internet was that it liberated our access to information - people could easily make a site and share anything they liked. If AI replaces the classic internet (through both AI search and AI slop drowning out human made sites) all access to information will pass through the prism of a very small number of billionaire owners.

Billionaires, if you hadn’t noticed, are mostly power hungry greedy madmen with god complexes. They’ve got too much control over our lives as it is.
 
  • Like
Reactions: heretiq, digimc, bbzt and 11 others
Grok works best on X content. Other than that it’s ChatGPT that is higher quality overall.

Elons input is just icing on the cake why Grok isn’t that serious outside X
 
stars_fan said:
I see no issue with it pointing out things that make some people uncomfortable.

The world is a pretty messed up place and some people live in a complete fantasy land.
Click to expand...

Pointing out things that are false, posited as truth, doesn't just make people uncomfortable, it makes them untrusting..because they are lies, and intentional.

Yes the world is a pretty messed up place. Elon putting his thumb on the scale doesn't make it less messed up.

I've chosen to just not use LLMs and chatbots, at all, from Grok, OpenAI, Apple, or otherwise, until this nonsense is over and the bubble has popped.
 
Last edited:
  • Like
Reactions: ateslik, heretiq, tubular and 12 others
Rogifan said:
Maybe it’s just me but why are people using AI chatbots to discuss sensitive political/cultural topics? Or asking it questions with the intent to draw out certain answers? I’ve never used Grok. Does anyone on Twitter use it for anything but trolling/s—tposting?
Click to expand...

People are lazy and ill informed and not willing to do any work to learn anything themselves.

The ramifications of turning more and more of society into useless blobs that just want a quick and easy "answer" for everything are going to be profound.

This feels appropriate as well:

1752242252896.png
 
  • Like
  • Love
Reactions: tubular, System603, MasterControlProgram and 8 others
Plerf said:
First Musk alienated half of the people (in the US) by going MAGA.

Then he alienated the other half of the people by feuding with their Orange Leader.

Now Musk trains an AI to be his friend since no one likes him.

Tell me I’m wrong.
Click to expand...
It became pro-Hitler for like a day. I am curious if that was based on instructions from Musk (I believe not long before he complained about the sources Grok was using, I guess he thought they were too woke) or some X employees sabotaging it because they didn’t like Musk’s request).
 
  • Like
Reactions: uacd
Rogifan said:
It became pro-Hitler for like a day. I am curious if that was based on instructions from Musk (I believe not long before he complained about the sources Grok was using, I guess he thought they were too woke) or some X employees sabotaging it because they didn’t like Musk’s request).
Click to expand...

Remember though .. it WAS NOT A NAZI SALUTE

He was just expressing his "love for everyone in the crowd"

/s
;)
 
  • Like
  • Haha
Reactions: tubular, Chidoro, System603 and 11 others
klasma said:
Are there still any left?
Click to expand...
Yeah, I still am.

I read the writeup that MacRumors referenced and I think that writeup was interesting - if you switch the phrasing from asking Grok who it supports to who one should support, it gives a very different answer. The author reasonably suggests that Grok lacks a sense of self. Who does Grok support? Grok is a program - you may as well ask Safari who it supports. Grok needs to assume whatever identity it thinks the asker wants it to have - it knows it's made by xAI and that Musk owns xAI, so it assumes the proper identity it should have is Musk's.
 
  • Disagree
  • Haha
  • Like
Reactions: ateslik, ryken229, Chidoro and 10 others
Rogifan said:
It became pro-Hitler for like a day. I am curious if that was based on instructions from Musk (I believe not long before he complained about the sources Grok was using, I guess he thought they were too woke) or some X employees sabotaging it because they didn’t like Musk’s request).
Click to expand...

It makes me wonder if Musk temporarily put Kanye West in charge…
 
  • Like
Reactions: steve333, maxoakland, StoneyG and 2 others
captain kaos said:
It also thinks its "Mecha-Hitler"......
Click to expand...

So, to settle it once and for all, why not turn to Musk's very own "maximum truth-seeking AI," Grok?

Asked to answer in one word the ideology it would most closely relate to Musk's onstage gesture, Grok didn't hesitate.

"Fascism," the AI chatbot replied.

futurism.com

Elon Musk's Grok AI Analyzes His Nazi Salute

Asked to answer in one word the ideology it would most closely relate to Musk's onstage gesture, Grok didn't hesitate.
futurism.com futurism.com
 
  • Like
Reactions: bbzt, StingRaySpeed, johnsawyercjs and 1 other person
turbineseaplane said:
People are lazy and ill informed and not willing to do any work to learn anything themselves.

The ramifications of turning more and more of society into useless blobs that just want a quick and easy "answer" for everything are going to be profound.

This feels appropriate as well:

View attachment 2527987
Click to expand...
I use ChatGPT frequently. I use Copilot at work frequently. I’ve never once thought to ask them inherently political questions. But there were some people on Twitter getting their jollies out of Grok correcting things Musk posted. And there are others who ask specific questions to try and show the AI has a bias.
 
  • Like
Reactions: sjjones, MasterControlProgram, gleepskip and 2 others
Ai is trained so of course it's going to take its responses from the primary source of its learning.

It just goes to show that Grok 4 training prioritized Musk views on some politics.
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back