Elon Musk's AI startup, xAI, has launched its new chatbot, Grok-2, claiming its performance rivals that of OpenAI, Google, and Anthropic products.
During the launch event on Wednesday, xAI showcased Grok-2, a large language model that has performed remarkably well in independent AI benchmarks. According to testing, it ranks among the top five AI chatbots, just behind Google’s Gemini and OpenAI’s ChatGPT. AI expert Ethan Mollick from the Wharton School mentioned on platform X, “Currently, there are five AI models equivalent to GPT-4: GPT-4o, Claude 3.5, Gemini 1.5, Llama 3.1, and Grok-2.” He emphasized that while all labs agree that there is still room for improvement, none have surpassed GPT-4 yet.
xAI describes Grok-2 as a significant advancement in AI development, capable of providing a more intuitive, controllable, and versatile experience across a range of tasks, including answering questions, collaborative writing, and coding challenges. According to rankings from AI assessment organization LMSYS, Grok-2 outshines top models from Meta and Anthropic.
In its internal evaluations, xAI focused on ensuring that Grok-2 adheres to user instructions and delivers accurate and truthful information. Notably, its predecessor faced criticism for "AI hallucination," a phenomenon where false information was incorrectly presented as fact—an issue seen as a significant barrier to the adoption of AI systems in businesses.
Initially, Grok-2 will be available to paid users on Musk's social media platform, X. Additionally, xAI plans to release the model to developers this month, enabling them to create applications for businesses.
Founded in March of last year, xAI has rapidly enhanced its technological capabilities with substantial investments. This year, the company secured $6 billion in funding, achieving a valuation of $18 billion. Recently, Musk has sought Tesla’s board approval for a $5 billion investment in xAI, bringing the company's total investment closer to OpenAI's $13 billion and surpassing Anthropic's $9 billion.