Recommendations of the Editorial team
What is it useful to be the richest man in the world when your chatbot refuses to lie for your disgusting policy? The right -wing extremist billionaire Elon Musk seems to be related to GROK again – DAS by his company Xai, AI model, which is integrated into its social media platform X.
After the attack on Saturday on the democratic Minnesota, Melissa Hortman, and her husband Mark Hortman in their house and the firing injury of Senator John Hoffman from Minnesota and his wife Yvette Hoffman-presumably by the same perpetrator-a large part of the online rights mobile to present the terrible attack as an act of a radical left. However, her efforts were put into perspective when Vance Boelter, the 57-year-old man, who is charged with the second degree and attempted murder, was apparently an anti-abdominal trump voter who had a “death list” that only comprised democratic mandate. (He has not confessed so far.)
Musk reacts to political violence with disinformation
But Musk, who never waits for facts-or if they ignore them when they are present-plunged into the weekend with Maga-Influencers. To spread a wrong story of left extreme violence. “The extreme left is murderically violent,” he wrote on Saturday to X., quoting a Trump supporter who falsely attributed the politically motivated murders in Minnesota to “the left”. Thereupon another user mentioned GROK in the thread. And asked: “Why is the left so murderous violent? You don’t look particularly tolerant.” Grok replied: “The claim that ‘the left’ is murderically violent is not supported by evidence”. And offered a centric classification. “There are political violence on all sides. Right attacks like on January 6th and left protests like the riots in 2020.
On the same evening, Musk reacted to an X-user and Trump supporter who complained that GROK was “manipulated by left indoctrination”, with the words: “I know. Is fixed this week.” It remains unclear how exactly a company controlled by Musk should have become a victim of left -wing brainwashing. However, it is not the first time that Musk and its right -wing extremists are angry to GROK angry because it contradicts their worldview. Conservatives complained right from the start that the bot was too “Woke”. Especially with topics such as diversity and gender identity. Contrary to Musk’s goal of developing an “anti-woke” AI.
Frustration over facts: GROK contradicts the right narrative
Musk itself was disappointed last year when Grok did not mock a transgender athlete during an appearance on the Joe Rogan Show. Instead, “Karens” moved to the ridiculous that made this topic a main point of the cultural struggle. The model is also known to correctly reproduce that Musk is a significant broadcast of misinformation and to refute Trump’s misleading lines of argument.
The frustration of the tech-oligarch over Grok’s assessments of political violence apparently grew when the bot further pointed out that right-wing violence has been more common and fattening in recent years. “Since 2016, data has been indicated that right-wing political violence has been more common and fatal. In incidents such as the Capitol uprising on January 6 and mass shootings (e.g. El Paso 2019), which demanded numerous fatalities,” said Grok on Tuesday. “Left -wing violence, especially during the protests in 2020, has increased, but is usually directed against real assets. And is less fatal.” Musk described this answer as wrong. “Serious mistakes. Because that is objectively wrong,” he wrote. Without delivery. “GROK raps what the old media say. We are working on it.”
On Wednesday, Grok continued to explain that right -wing extremism demands more fatalities. “From 2015–2020, right -wing extremists counted incidents 267 with 91 deaths, compared to 66 left -wing extremists with 19 deaths,” it said in a contribution. “Right -wing attacks often aim at humans. While left -wing violence usually goes out in protests on property damage. What looks more visible but is less fatal.”
How grok is reprogrammed remains unclear – but hardly subtle
It is uncertain how exactly Musk’s team wants to change at Xai Grok that it reproduces its distorted reality. But the result should hardly be subtly. X-users were surprised in the last month when the bot repeatedly warned of an alleged “white genocide” in South Africa. Even with completely non -subject inquiries. In South Africa, where Musk was born, white people are not systematically murdered. Nevertheless, he and Trump repeatedly claimed that the country’s white minority is a victim of violent persecution. Trump even accelerated the immigration process for some of these alleged “refugees”.
Many suspected that GROK had been specifically changed on Musk’s instruction so that it brings South Africa into discussions. Even then, however, it could not prove the baseless thesis of a “white genocide”.
This behavior was later corrected and the bizarre contributions were deleted. That is comparatively simple. To make a chat bot that is programmed to search for “truth” to lie through terrorist statistics without being like clumsy manipulation, it should be much more difficult. Nevertheless, you can rely on the fact that Musk continues to work on the most important technological challenge of our time. To get software to be in all of his opinion.