OpenAI develops bel model with over 10t parameters as AI race intensifies

A report from the X platform claims that OpenAI has completed the pre-training of its next-generation model, internally known as Bel, which is positioned as the direct successor to the Doug model. The Bel model is said to have more than 10 trillion parameters, positioning it as a potential foundation for future large language models and possibly advancing toward artificial general intelligence (AGI). AGI refers to a hypothetical intelligent system capable of learning, understanding, and performing any intellectual task that a human can accomplish. Unlike narrow AI, which is designed for specific tasks, AGI would possess cross-domain generalization, reasoning, and self-learning capabilities in unfamiliar environments. The development of the Bel model is seen as a significant move by OpenAI, aiming to establish it as the base model following GPT-6. The report suggests that OpenAI believes the Bel model could push the boundaries of what is currently achievable in AI, potentially reaching the threshold for AGI. The announcement is expected to put pressure on Anthropic, as OpenAI claims that the company is unprepared to respond to the upcoming release of Astra. According to the report, Anthropic has already accepted that the current landscape will favor OpenAI for the remainder of the year, but the company remains confident it can regain its competitive edge by early next year. The report highlights the growing competition in the AI field, with major players like OpenAI and Anthropic vying for dominance. The development of models with such vast parameter counts underscores the increasing scale and ambition of AI research, as companies strive to achieve breakthroughs in artificial intelligence.