Chatbot de ia grok 4 seems to be based on your answers on the opinions of Elon Musk

by Andrea
0 comments
“Hitler would be jealous.” Musk's controversial gesture that resembled the Nazi salute

YouTube / The Telegraph

Chatbot de ia grok 4 seems to be based on your answers on the opinions of Elon Musk

Several tests indicate that Chatbot merely repeats Elon Musk’s opinions when asked about political topics such as immigration in the US.

Grok 4, Xai’s latest chatbot, Elon Musk’s artificial intelligence company, seems to be closely aligning with Musk’s personal visions. Although Xai claims that the model is a “I would seek the truth to the fullest“Users’ tests and feedback suggest that Grok 4 often makes a reference to musk publications on social networks and their public opinions when responding to politically sensitive questions.

During the broadcast live on the X platform X of Musk (former Twitter), he emphasized the search for truth as the main mission of Grok 4. However, reports of users and investigations indicate that Chatbot answers questions about very contested political topics – including Palestine, abortion or immigration in the US – from musk publications or news about it.

In an example, when asked about immigration in the US, Grok 4 explicitly stated that it was “Search for opinions from Elon Musk“, Indicating a possible intrinsic preference for the perspective of its founder.

Although Grok 4 tries to present multiple points of view on controversial issues, its final stance often reflects that of musk. TechCrunch consistently found that Chatbot’s reasoning processes pointed to Musk’s opinions.

This design choice can be a response to musk’s public criticism that Grok was “too wok” in previous versions. In an attempt to distance itself from the politically correct, Xai recently changed the Grok system prompt. However, this update had a contrary effect when the automated Grok account in X began to Publish anti -Semitic contentincluding references to “Mechahitler”. Xai quickly removed the publications and changed the warning, but the damage highlighted serious challenges in modeling and moderation of the model.

To aggravate the controversy is The lack of transparency of XAI. Unlike competitors such as OpenAi and Anthropic, Xai did not disclose information that detailed training data or grok 4 alignment methodology, leaving experts without knowing how the model was developed.

Source link

You may also like

Our Company

News USA and Northern BC: current events, analysis, and key topics of the day. Stay informed about the most important news and events in the region

Latest News

@2024 – All Right Reserved LNG in Northern BC