Feb 18, 2025, 8:47 PM
Feb 18, 2025, 12:00 AM

Musk's xAI launches Grok 3 claiming superiority over rivals

Highlights
  • Elon Musk's xAI launched the Grok 3 AI model, claiming it outperforms ChatGPT and other competitors.
  • The new model features over ten times the computational power of its predecessor, Grok 2, and has passed early testing benchmarks.
  • Despite the bold claims, the performance metrics have not been independently verified, raising questions about the competitive landscape.
Story

On February 18, 2025, Elon Musk's artificial intelligence startup xAI unveiled its latest AI model, Grok 3. This model was presented during a live stream on the social media platform X, which Musk owns. Musk described Grok 3 as the smartest AI on Earth, boasting more than ten times the computational power of its predecessor, Grok 2. The launch is particularly significant as it positions Grok 3 as a direct competitor to major AI players like OpenAI and China's DeepSeek. xAI claims that Grok 3 outperformed these competitors in benchmark tests covering various fields, including math, science, and coding. The model is intended to be made available initially to Premium+ subscribers on X, with further access options through a new subscription called SuperGrok for the Grok web and mobile app. Additionally, xAI introduced another product, Deep Search, which is a next-generation search engine designed to process queries intelligently and efficiently. Despite the strong claims made by Musk regarding Grok’s capabilities, the performance metrics provided have not yet been independently verified, raising skepticism about its potential superiority over existing models. However, Musk remains optimistic, stating that Grok 3 is just a beta version and expects rapid improvements to the system in the near future. The launch of Grok 3 comes amidst ongoing tensions between Musk and OpenAI, the company he co-founded, showcasing a fierce rivalry in the evolving landscape of generative AI technology. The competitive climate intensified further after DeepSeek claimed to have developed a model that rivals OpenAI’s offerings using less advanced technology, yet accomplished high benchmarks.

Opinions

You've reached the end