Muhammad Selim Korkutata | Anatolia | Getty Images
Elon Musk’s XAI unveiled its latest artificial intelligence model, Grok 3, on Tuesday, claiming it can be indexed based on the enablement of early tests and in-depth indexes in China, which include standardized tests on mathematics, science and coding.
“We are excited to showcase the Grok 3, which is a more capable order than the Grok 2 in a very short time,” Musk said in his demo of Grok 3 which was broadcast on social media. Platform X.
The team also said it is launching a new product called Deep Search, which will act as a “next-generation search engine.”
The XAI team said that later that day, Grok 3 will be launched for premium X subscribers and will also be subscribed separately through the web and application versions of the model.
Speech at the World Government Summit in Dubai last week Called the model “Horror Intelligence” with strong reasoning capabilities claims it outperforms all other existing models in XAI’s internal testing.
“This may be the last time AI is better than Grok,” Musk said at the time, adding that it is a training for “many synthetic data” and is able to reflect on its errors to achieve logical consistency.
XAI team claims that early iterations of Grok 3 are better than existing competitors’ ratings Chatbot ArenaIt is a crowdsourcing website that is in opposition to AI models in blind testing.
Musk said the company will continue to improve the model at the end of the product demonstration.
“We should emphasize that this is a beta, which means you should expect some imperfections at the beginning, but we will improve it quickly almost every day,” he said.
Fierce competition
Musk, Who speaks out Regarding the potential dangers of artificial intelligence, XAI began entering the generative AI market including Openai Chatgpt in 2023.
Last September, Openai launched its most advanced model, O1, and with it Reasoning ability and able to solve relatively complex scientific, coding and mathematical tasks.
Musk, along with Sam Altman, helped create a nonprofit in 2015.
But in recent years, Musk and Openai’s leadership Always arguing. Musk recently led an investor group that submitted proposal OpenAI rejected the nonprofit parent who bought an AI startup for $97.4 billion.
Last month, Chinese startup DeepSeek shocked the AI market by publishing a technical paper, which claims one of its open source models can rival the performance of OpenAI’s O1 model, despite using cheaper, less energy-intensive type process.
Facing the U.S.’s Feat of Leading AI Chip Manufacturers nvidia From using its cutting-edge GPU (for training AI models) to China.
Xai has one”Giant supercomputer,“For training AI, last year said the training was a cluster that used 100,000 advanced NVIDIA GPUs for AI training. On Tuesday, the company revealed that its GPU cluster doubled to train Grok 3.
Although many AI and technologists have Tell CNBC DeepSeek has exacerbated AI competition, showing what less advanced technologies can do, and others are more skeptical of its impact.
