Nvidia CEO Jensen Huang has declared that artificial general intelligence (AGI) has arrived following the launch of OpenAI’s latest AI model, GPT-6 Astra. Huang made the statement in a post on X, congratulating OpenAI and pointing to the rapid progression from ChatGPT to newer generations of AI systems. However, the claim remains contested, with AI researchers questioning whether Astra meets the standards generally associated with AGI.
Huang wrote that the progression from ChatGPT to OpenAI’s o1 and then Astra had taken four years before stating that “AGI has arrived.” He also highlighted the computing infrastructure behind the model, saying Astra was trained on approximately 100,000-plus Nvidia Grace Blackwell NVLink72 systems. Huang further said that another 400,000 GPUs were coming online.
OpenAI describes GPT-6 Astra as its latest generation of AI intelligence, combining advances in pre-training, reinforcement learning and alignment. The company says Astra is state-of-the-art in areas including computer use, web browsing, software engineering, cybersecurity, science and professional work. OpenAI’s published evaluation results include a 98% score on FrontierMath Tier 4, 99.9% on ARC-AGI-3 and 100% on ExploitBench. These figures are OpenAI’s reported benchmark results and should not by themselves be interpreted as proof that the system has achieved AGI.
The launch has also highlighted the close relationship between OpenAI’s AI development and Nvidia’s computing infrastructure. India Today reported that most of OpenAI’s training fleet and inference infrastructure continues to run on Nvidia GPUs. Huang’s announcement of another 400,000 GPUs coming online underscores the scale of computing resources being deployed as AI models become increasingly capable.
OpenAI President Greg Brockman also responded positively to Huang’s assessment. He said the industry was moving into the AGI era, while acknowledging the importance of close partnerships in reaching this stage.
The definition of AGI, however, remains an important part of the debate. OpenAI’s stated mission refers to AGI as AI systems that are generally smarter than humans. The term is commonly used to describe systems capable of performing a broad range of cognitive tasks at or beyond human levels, rather than being limited to a narrow set of functions.
AI researcher and critic Gary Marcus challenged Huang’s conclusion shortly after the Nvidia CEO’s post. Marcus argued that Huang had provided neither evidence nor a clear definition for the declaration and said that, under conventional definitions, Astra still falls short of AGI. His comments illustrate the lack of consensus surrounding the point at which increasingly capable AI systems should be classified as artificial general intelligence.
As a result, Huang’s statement is best understood as an industry leader’s assessment of Astra’s capabilities rather than a universally accepted declaration that AGI has been achieved. What is clear from OpenAI’s own release is that Astra represents a substantial expansion in AI capabilities across computer use, software engineering, cybersecurity, science and professional tasks. The continuing debate over whether those capabilities constitute AGI is likely to remain central as increasingly powerful AI systems move into real-world use.
 has arrived following the launch of OpenAI’s latest AI model..){kind=link}