Nvidia CEO Jensen Huang declared that artificial general intelligence has arrived, citing the release of OpenAI’s GPT-6 Astra model as evidence. He first mentioned this in an earnings call on August 26, 2026, and later reinforced the claim in an X post on September 6, 2026.
OpenAI began rolling out GPT-6 Astra around September 3, 2026. According to Huang, the model was trained using more than 100,000 Nvidia Grace Blackwell GPUs, highlighting the significant hardware requirements for developing such systems.
Analysts from Goldman Sachs and Morgan Stanley are focusing on the potential impact of these developments on capital expenditure for AI infrastructure. They suggest that the narrative surrounding AGI could drive increased demand for computing capacity.
Investment firms have identified Nvidia, Broadcom, and Micron as key companies to watch. The thesis relies on the assumption that AI labs will continue to invest heavily in hardware to secure the necessary infrastructure for training future models.
Despite the optimism, analysts note risks related to the concentration of investments and the potential for a slowdown in capital spending. Furthermore, investors are advised to consider Huang's commercial interest in promoting AI growth.
