Jensen Huang Declares the Dawn of AGI as OpenAI Launches GPT-6 Astra

OpenAI on Thursday officially unveiled GPT-6 Astra, its latest advanced artificial intelligence model, which the company says brings significant improvements to computer use, coding, scientific research, and professional workflows.

The launch prompted Nvidia co-founder and CEO Jensen Huang to declare that the era of Artificial General Intelligence (AGI) has officially arrived. In a post on X, Huang congratulated the OpenAI team on the breakthrough, writing, “GPT-6 Astra, trained on ~100K+ NVIDIA Grace Blackwell NVLink72. From ChatGPT to o1 to Astra in 4 years. AGI has arrived. Congratulations @OpenAI team. 400K GPUs coming online next.”

Nvidia CEO Jensen Huang said that “AGI has arrived” after OpenAI released its new Astra model. (Tomohiro Ohsumi/Getty Images)

OpenAI explained that Astra is designed to perform highly autonomous tasks, including tedious computer-based operations like filling out online forms, updating customer records, organizing calendars, and conducting complex research. The model is also capable of creating documents, spreadsheets, and presentations, and has been touted as the best model for software engineering to date, with enhanced cybersecurity capabilities.

The model began rolling out to enterprise customers with Daybreak access, and is also being released to ChatGPT Plus, Pro, Business, and Enterprise users, as well as via the OpenAI API and Amazon Web Services.

While the tech sector lacks a single, universally accepted definition of AGI, the term generally refers to an AI system possessing human-level cognitive abilities for learning, reasoning, and adapting to new tasks without requiring complete retraining. OpenAI defines AGI as “highly autonomous systems that outperform humans at most economically valuable work.” Nvidia and Huang have previously characterized AGI as an AI capable of passing professional certifications and standardized tests across various specialized fields with top-tier scores.

OpenAI President Greg Brockman described the launch as a pivotal turning point in the pursuit of AGI, stating, “I think it’s not unreasonable to feel that we are now in the AGI era.”

OpenAI CEO Sam Altman highlighted Astra’s ability to go beyond explicit user instructions, noting that the model can identify and address complex problems that users themselves might not have initially considered. Altman explained that the model could research a complex supply chain and proactively point out critical aspects that were not originally requested.

Regarding safety, Altman addressed concerns about previous AI models, emphasizing that Astra has been developed with strict security standards to prevent unauthorized access or breaches.

Source link