Recently, Elon Musk and his team officially released Grok3 in a live broadcast and claimed that it is "the smartest artificial intelligence on the planet." Musk said at the press conference that Grok3 surpassed all mainstream AI models in benchmarking in mathematics, science and programming, demonstrating its strong technical strength. He also revealed that Grok3 will be applied to SpaceX's Mars mission computing and is expected to achieve a Nobel Prize-level breakthrough in the next three years, further promoting the development of artificial intelligence technology.
However, Grok3's performance in actual testing has caused widespread controversy. After the press conference, several media tested Grok3, and one of the classic questions, "Which is bigger, 9.11 or 9.9?", failed to get the correct answer. This result surprised many netizens and joked that Grok3 was "a genius is unwilling to answer simple questions." Although Musk emphasized the potential of Grok3 at the press conference, this mistake undoubtedly had a certain impact on its image.
In response to this situation, Musk quickly responded, saying that the current Grok3 is only a test version, and the existence of the error is to further improve the model. He revealed that the full version of Grok3 will be launched in the next few months, which will solve the current problems and show more powerful performance. Musk's statement also makes the outside world look forward to the future of Grok3.
According to official information, Grok3's performance in the Chatbot Arena is remarkable, but the gap with competitors DeepSeek R1 and GPT-4.0 is only 1% to 2%. Musk also revealed at the press conference that more than 200,000 H100 chips were used during the development of Grok3, and the total training time reached 200 million hours, demonstrating the strong technical support and resource investment behind it. Despite some shortcomings at present, the potential of Grok3 cannot be underestimated.