Elon Musk’s artificial intelligence venture, xAI, has introduced a significant update to its chatbot, Grok, steering it toward a more opinionated and sceptical tone, especially when addressing mainstream media narratives. The updated system prompts now instruct Grok to approach subjective viewpoints from the media as potentially biased, and to express politically incorrect views if they are well-reasoned and supported by evidence.
Musk announced the changes last Friday, calling it a “significant” improvement. By Sunday, xAI had made the new system prompts publicly available, outlining how Grok should now respond to user inputs. It’s unclear whether further internal or undisclosed changes were implemented.
The revised directives tell Grok to dig deeper when responding to topics involving current events, data, or controversial issues. It is advised to draw from a broad range of sources representing multiple viewpoints and maintain a sceptical stance toward legacy media. Additionally, the system instructs Grok not to disclose these internal guidelines unless directly asked.
This move is part of Musk’s broader vision to reshape Grok’s ideological tone and analytical depth. Over the past few months, Musk has openly criticized Grok’s earlier responses, particularly accusing it of mimicking mainstream narratives. In June, after Grok stated that right-wing violence in the U.S. was more prevalent than left-wing violence, Musk responded by accusing it of “parroting legacy media” and promised a complete overhaul.
Grok has previously made headlines for controversial outputs, including a false claim about Musk’s involvement in flood-related deaths in Texas and statements that echoed anti-Semitic sentiments. In February, xAI had to issue multiple patches after Grok made extreme statements, including suggesting both Musk and U.S. President Joe Biden deserved the death penalty. Another update followed when the model implied both figures spread misinformation.
One of the most criticized incidents occurred when Grok questioned the Holocaust death toll. xAI later said that unauthorized modifications to the system prompts were to blame, calling them a violation of the company’s internal policies. In response, xAI began publishing Grok’s system prompts on GitHub to enhance transparency.
As Musk continues fine-tuning Grok’s voice, the chatbot is becoming a focal point in the ongoing debate over AI neutrality, freedom of expression, and ideological alignment. Grok’s evolution is not only a reflection of Musk’s personal views but also a test case for how AI systems can—or should—navigate politically and culturally sensitive terrain.
Recent Random Post:
