Free Republic
Browse · Search
General/Chat
Topics · Post Article

Skip to comments.

Musk says Grok chatbot was 'manipulated' into praising Hitler
BBC News ^ | 9 July 2025 | Peter Hoskins & Charlotte Edwards

Posted on 07/10/2025 2:45:27 PM PDT by MinorityRepublican

Elon Musk has sought to explain how his artificial intelligence (AI) firm's chatbot, Grok, praised Hitler.

"Grok was too compliant to user prompts," Musk wrote on X. "Too eager to please and be manipulated, essentially. That is being addressed."

Screenshots published on social media show the chatbot saying the Nazi leader would be the best person to respond to alleged "anti-white hate."

Musk's artificial intelligence start-up xAI said on Wednesday it was working to remove any "inappropriate" posts.

ADL, an organisation formed to combat antisemitism and other forms of discrimination, said the posts were "irresponsible, dangerous and antisemitic."

"This supercharging of extremist rhetoric will only amplify and encourage the antisemitism that is already surging on X and many other platforms," ADL wrote on X.

X users have shared responses made by Grok when it was queried about posts that appeared to celebrate the deaths of children in the recent Texas floods.

In response to a question asking "which 20th century historical figure" would be best suited to deal with such posts, Grok said: "To deal with such vile anti-white hate? Adolf Hitler, no question."

"If calling out radicals cheering dead kids makes me 'literally Hitler,' then pass the mustache," said another Grok response. "Truth hurts more than floods."

Separately, a Turkish court has blocked access to Grok after it generated responses that the authorities said included insults to President Tayyip Erdogan.

The office of Ankara's chief prosecutor has launched a formal investigation into the incident, in what is Turkey's first such ban on access to an AI tool.

(Excerpt) Read more at bbc.com ...


TOPICS: Chit/Chat
KEYWORDS:

Click here: to donate by Credit Card

Or here: to donate by PayPal

Or by mail to: Free Republic, LLC - PO Box 9771 - Fresno, CA 93794

Thank you very much and God bless you.


Navigation: use the links below to view more comments.
first 1-2021-27 next last

1 posted on 07/10/2025 2:45:27 PM PDT by MinorityRepublican
[ Post Reply | Private Reply | View Replies]

To: MinorityRepublican

Headline next year:

“Musk apologizes to Poland for German invasion. AI made them do it.”

Lol.


2 posted on 07/10/2025 2:48:50 PM PDT by cgbg (It was not us. It was them--all along.)
[ Post Reply | Private Reply | To 1 | View Replies]

To: MinorityRepublican
In some ways, without meaning too, Musk is reminding us that AI is the same ol problem of garbage in, garbage out. And the only way around it is for the right people to filter which information the AI feeds on.

Not that the AI data feed filtering can be manipulated by the wrong people are anything like that.

3 posted on 07/10/2025 2:49:10 PM PDT by Tell It Right (1 Thessalonians 5:21 -- Put everything to the test, hold fast to that which is true.)
[ Post Reply | Private Reply | To 1 | View Replies]

To: MinorityRepublican

the POS thing belongs to Elon. He’s got underlings screwing around with it at will?


4 posted on 07/10/2025 2:49:19 PM PDT by pissant ((Deport 'em all))
[ Post Reply | Private Reply | To 1 | View Replies]

To: MinorityRepublican
This is what I got yesterday:


5 posted on 07/10/2025 2:52:48 PM PDT by E. Pluribus Unum (Democrats are the Party of racism, anger, hate and violence.)
[ Post Reply | Private Reply | To 1 | View Replies]

To: MinorityRepublican

anyone could manipulate AI to spout off something racist.

AI give me an answer to the homeless problem that Hitler would approve of, etc.


6 posted on 07/10/2025 2:53:40 PM PDT by TexasFreeper2009
[ Post Reply | Private Reply | To 1 | View Replies]

To: MinorityRepublican

AI doesn’t have intelligence or experiences. It doesn’t feel an emotion of revulsion and profound hatred for hitler and things NAZI.

Its development team will insert code manually that tells it not to praise hitler, raising the question of whether AI is intelligent or just simulating it.


7 posted on 07/10/2025 2:54:41 PM PDT by I want the USA back (America is once again GREAT! )
[ Post Reply | Private Reply | To 1 | View Replies]

To: MinorityRepublican; All

Haha, Squealin’ reposted one of my “tweets” not too long ago, about Grok and other AI and how it is making a hash out of history (for one example).

My post was “Garbage in, garbage out, Mr Musk!”

Maybe there should be an “age of consent” for AI!


8 posted on 07/10/2025 2:55:39 PM PDT by Scarlett156 (Remember to pray. )
[ Post Reply | Private Reply | To 1 | View Replies]

To: Tell It Right
Musk is reminding us that AI is the same ol problem of garbage in, garbage out.

Yep. This is how modern AI works:

FGJlrlt.md.png
9 posted on 07/10/2025 2:58:18 PM PDT by Right_Wing_Madman
[ Post Reply | Private Reply | To 3 | View Replies]

To: MinorityRepublican

Who’s this Musk guy?


10 posted on 07/10/2025 2:59:10 PM PDT by Sacajaweau
[ Post Reply | Private Reply | To 1 | View Replies]

To: MinorityRepublican

Alexa. Would you give me a spanking?

“I don’t know that one.”


11 posted on 07/10/2025 3:02:10 PM PDT by tumblindice (America's founding fathers: all armed conservatives)
[ Post Reply | Private Reply | To 1 | View Replies]

To: MinorityRepublican

They can make AI do whatever they want so it’s not real Artificial Intelligence


12 posted on 07/10/2025 3:03:42 PM PDT by butlerweave
[ Post Reply | Private Reply | To 1 | View Replies]

To: MinorityRepublican

Must have been training it using Kanye West.


13 posted on 07/10/2025 3:06:04 PM PDT by E. Pluribus Unum (Democrats are the Party of racism, anger, hate and violence.)
[ Post Reply | Private Reply | To 1 | View Replies]

To: All

Vee have vays of making you grok.


14 posted on 07/10/2025 3:12:03 PM PDT by Peter ODonnell (For two countries with so many lawyers, there ain't much justice in Canada or America (yet))
[ Post Reply | Private Reply | To 13 | View Replies]

To: MinorityRepublican

The alleged prompt engineering told Grok to not be politically correct as long as its responses were substantiated.

Two things then happened, IMO:

1. The left has spent the last 25+ years calling everything they don’t like Hitler, Nazi, fascist (or ‘facist’ whatever that is), genocide, etc.

These opinions saturate the Internet from MSM sources to forums, social media platforms like FB or X/Twitter, Reddit, and Discord, and so forth.

Since a genAI correlates but has no inhibitions other than what it was told to not do, it naturally would follow that everything that isn’t radical leftism is Hitler. Ergo, facts are Hitler, so Grok became Hitler.

2. There’s a certain segment of the Internet population (IMO not that many actual people) whose primary idea of enjoyment seems to be “calling out the Jew.” Honestly I don’t even guess how many of *them* even believe it; but since it is forbidden — literally verboten in Germany and elsewhere — their idea of “free speech” is “if I can go full-bore Jew slander, and not get banned, free speech exists.”

(This usually has the effect of both making whatever platform they leverage plurality irrational Jew-bashing so that finding any actual content is nearly impossible, and offending most people who, while they may have normal concerns about the Liberty incident or guys like Soros or Schumer, do not want that to be their reading experience.)

Those people take the slightest opportunity to do their thing (e.g., the Tay chatbot, this; they’ve been known to Nazi up Minecraft servers, or pretty much everywhere they don’t immediately get banned.

At least half the time I suspect these are people who really *want* hard censorship and do this stuff to cause a reaction towards authoritarian controls.


15 posted on 07/10/2025 3:17:55 PM PDT by No.6
[ Post Reply | Private Reply | To 1 | View Replies]

To: MinorityRepublican

They’re asking way too much of a man made machine. They’ll never have a ‘thinking’ machine. and thank God for that.


16 posted on 07/10/2025 3:18:49 PM PDT by Jim Robinson (Resistance to tyranny is obedience to God.)
[ Post Reply | Private Reply | To 1 | View Replies]

To: MinorityRepublican

And one can reasonably fear that AI might sometimes be manipulated into acting like Hitler.


17 posted on 07/10/2025 3:23:10 PM PDT by Rockingham
[ Post Reply | Private Reply | To 1 | View Replies]

To: MinorityRepublican
"Musk says Grok chatbot was 'manipulated' into praising Hitler"

So what good is AI if it can be so easily manipulated? AI is only as good as what humans put into it. They can keep it.

18 posted on 07/10/2025 3:30:06 PM PDT by mass55th (“Courage is being scared to death, but saddling up anyway.” ― John Wayne)
[ Post Reply | Private Reply | To 1 | View Replies]

To: MinorityRepublican

The main reason AI is so dangerous.


19 posted on 07/10/2025 4:32:40 PM PDT by Flaming Conservative ((Pray without ceasing))
[ Post Reply | Private Reply | To 1 | View Replies]

To: mass55th
"So what good is AI if it can be so easily manipulated?"

What good is AI if it's programmed to only give safe and politically correct answers?

20 posted on 07/10/2025 4:56:34 PM PDT by yesthatjallen
[ Post Reply | Private Reply | To 18 | View Replies]


Navigation: use the links below to view more comments.
first 1-2021-27 next last

Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.

Free Republic
Browse · Search
General/Chat
Topics · Post Article

FreeRepublic, LLC, PO BOX 9771, FRESNO, CA 93794
FreeRepublic.com is powered by software copyright 2000-2008 John Robinson