Nvidia CEO Declares AGI Here, Hints at Massive Chip Deployment
Nvidia CEO Jensen Huang has sent shockwaves through the tech industry with a bold declaration that Artificial General Intelligence (AGI), defined as software capable of human-level cognitive tasks, has already arrived. Huang stated that OpenAI's latest ChatGPT model was trained on Nvidia's chips, pushing the AGI timeline forward significantly from optimistic predictions of 2029 to 2026. This advancement is supported by a massive engineering shift at Nvidia, moving from single-board processors to integrated 72-chip 'super-brains'.
Huang initially claimed OpenAI's model was trained on 300,000 of these massive chip units, but later deleted and revised the number to over 100,000. Nvidia has not provided an explanation for this discrepancy, fueling speculation. The dramatic statements coincide with Nvidia's immense financial success, reporting $89 billion in AI computer sales in three months, more than double the previous year. Huang emphasized the link between computing power and revenue, suggesting a strong incentive to promote the AGI narrative.
Despite Huang's enthusiasm, OpenAI has adopted a more cautious stance. Company president Greg Brockman has not confirmed AGI's achievement, leaving Huang as the sole prominent figure insisting on its arrival. Furthermore, a previously announced $100 billion deal between Nvidia and OpenAI in 2025 was reportedly never finalized, with reports indicating OpenAI has been reducing chip purchases. Huang promised an additional 400,000 chips coming online soon, but financial markets are expected to focus on tangible sales figures.
Nvidia's aggressive promotion of AI extends to Israel, where the company has appointed Israeli content creators Noa Zilberman and Lina Plat as official ambassadors. They will attend the GTC Berlin conference from October 20-22, 2026, to communicate Nvidia's technological advancements to the local audience.
