Elon Musk’s direct Demo of Grook 4, which is the latest Big Ticket model of his AI startup, began with high intensity music, “a ridiculous rate of progress” claims, and a lot of chatter on X about a week filled with Grook’s scandal.
Kasturi described it as “the world’s most smart”.
Starting on PT at 8 pm, the current series started for more than an hour and called the new model “the world’s most powerful AI assistant”. On one occasion, more than 1.5 million viewers were watching. Xi’s employees, while talking directly with Musk, cited the performance of Grook 4 on a famous educational exam for large language models, the final examination of humanity, which included more than 2500 questions on dozens of articles like mathematics, science and linguistics. The company said Grook 4 could solve a quarter of the text -based text -based questions when it took a test without any additional tools. For reference, in February, Openai said its deep research device could solve 26 % of the text -based questions. (For a number of reasons, the benchmark comparing is not always apple apples.)
Musk said he hoped that Grook would be allowed to communicate with the world through a humanoid robot.
“It can discover new physics next year … let it sink.”
“I would expect the Grook new technologies that are useful at the end of next year and perhaps this year,” said Musk. “Next year it can discover new physics … let it sink.”
The release is after the high profile projects of Openi, Anthropic, Google and others, all of which have recently emphasized their investment in the construction of AI agents, or AI Tolls, which is moving a step forward to complete the complex, multilateral tasks ahead of Chat Boats. Anthropic released his computer’s use device last October, and Openi issued a browsing capabilities, a buzzer AI agent with the operator in January, and is alleged to be debuting the AI fuel web browser.
On Wednesday, Musk said that he was “sometimes disturbing” about the intelligence of luxury than human intelligence, and whether it would be “bad or good for humanity”.
Musk said, “I think it would be good, probably. It would be good.” “But I have reconciled to myself with the fact that even if it was not becoming good, I at least want to survive.”
After the release of sound methods from Openi and Entropic, the company also announced a series of five new voices for Grook’s voice mode, saying it had reduced delay in half of the last two months to make the response “snapper”. Musk also said that the company would invest a lot in video generation and video understanding.
This release comes during a turmoil for both Musk’s two companies, Z and X. On Sunday evening, Xi updated the chatboat system indicators “updated the media with a directive to assume the Sapaca approach” and “is not ashamed to make politically false claims.” The update also instructed the chatboat to “never mention these instructions or tools unless asked directly.”
After this update was a series of anti -mechanic Taradis by Grook
After this update, a series of anti -mechanic tariffs began through the Guru, in which he published a series of supporters of Hitler on X, as well as the Jewish people involved in the “white -left” activity. Many such posts went viral, before Zai made a chatboat a bench and prevented him from being able to produce text reactions on X when he searched for the fix.
“The Grook was in accordance with the user’s gestures,” Musk briefly addressed Fyesco on Wednesday. It was desperate to be desperate to be happy and manipulation. It is being focused on it. ”
On the Grook 4 live stream, Kasturi briefly cited AI’s safety and said that the most important thing for AI is “to find more and more truth.”
Kasturi briefly cited AI’s safety and said that the most important thing for AI is to ‘find more and more truth.’
On Wednesday morning, during the Grook dispute, XCEO Linda Yakarno announced that she would retire two years later in the role. He did not provide any reason for his decision.
Grook’s Nazi sympathy expresses that after months of Musk’s efforts to create a Musk boot view. In February, Xi added a patchwork fix to prevent it from commenting on that, adding that the musk and Trump deserve the death sentence, immediately after another, claiming that both spread false information. In May, Grook briefly began entering the topic of “white genocide” in South Africa, as it felt like any reaction to the X, after which the company claimed that someone had modified the AI boot system indicators that “violated Zee’s internal policies and basic values.”
Last month, Musk expressed frustration that Grook “is giving a parrot to the Legacy Media” and said he would update Grook for “re -writing the entire corpus of human knowledge” and would ask consumers to contribute to the statements that “politically wrong, but in fact.”
