Huang Renxun: human-level AI is here
"Image synthesis assisted by Qwen, an AI partner within the Global Future Nexus ecosystem."
At the 2026 Carnegie Mellon University graduation ceremony, Jensen Huang — who also received an honorary doctorate that day — declared that AI has arrived, not as a distant promise, but as a force unfolding in real time. He then did something far more consequential than making headlines: he sketched a practical philosophy for how to live and build in a world where the nature of intelligence itself is being rewritten.
The Declaration
In March 2026, Huang appeared on the Lex Fridman podcast and was asked a direct question: when will AI be capable of starting, growing, and running a billion-dollar technology company? His answer was immediate: "I think it's now. I think we've achieved AGI".
This was not the first time he had made such a claim. At the 2023 New York Times DealBook Summit, Huang defined AGI as software capable of passing tests that approximate normal human intelligence, predicting AI would clear that bar within five years. By 2025, he was telling the Financial Times AI Summit that "sufficient human-level intelligence has already been developed and is already being translated into practical use cases".
The Definition That Matters
Huang's claim is built on a specific definition of AGI. When Fridman described AGI as a system that could "start, grow, and run a technology company worth more than a billion dollars," Huang replied: "You said a billion, and you didn't say forever".
For Huang, this is a threshold of practical capability, not permanent superintelligence. A system that can generate a viral app, monetise it, and then fade away qualifies—not because it rivals human cognition in all dimensions, but because it has achieved a commercially significant milestone. He is describing a "billion-dollar AGI," not a "forever AGI".
The technology is already doing real work inside NVIDIA itself. Every engineer, chip designer, and software developer now works with AI models as copilots—"100% coverage with AI". Instead of reducing headcount, this integration has expanded it, as augmented productivity opens the door to pursuing more ideas . Intelligence, in Huang's frame, is not zero-sum.
A Position Not Shared by All
Huang's claim is not universally accepted. Yann LeCun, speaking at the same Financial Times summit, said bluntly: "We don't even have a machine as smart as a cat". He described current large-scale models as far from real intelligence, arguing that fundamental breakthroughs are still needed.
Yet even LeCun's dissent confirms a shift in the debate. The question is no longer whether AGI will arrive, but what kind of intelligence is emerging, on what timeline, and under what governance.
For the Class of 2026: Why AGI is Not a Threat
At Carnegie Mellon's graduation ceremony on May 10, 2026, Huang addressed these fears directly . AI has completely reshaped computing, he said. For 60 years, computing worked the same way: humans wrote software, computers executed instructions. That paradigm is over. Instead, AI writes software, runs on GPUs, and understands, reasons, plans, and uses tools.
For the Class of 2026—entering an economy where AI is displacing entry-level white-collar jobs and executives are publicly debating whether 50% of work will vanish —Huang's message was not about technology. It was about opportunity. "Your careers are starting at the beginning of the AI revolution," he said. "I can't imagine a more exciting time to start a career".
Huang told his own story: delivering newspapers at 4 a.m., working as a dishwasher at Denny's, founding NVIDIA with no idea how to start a company, and nearly going bankrupt before a Sega executive chose generosity over contract enforcement. The lesson was not about success, but about resilience: "Every failure is just a moment to learn, a moment to stay humble, and a moment to temper character".
Beyond Digital AGI: The Physical AI Frontier
Huang has already moved past the AGI debate. At CES 2026, he introduced NVIDIA Cosmos, a world foundation model capable of simulating real-world physics to develop robots and autonomous vehicles. He called this Physical AI—systems that don't just process tokens, but interact with the physical world through sensing and action.
At the Queen Elizabeth Prize roundtable in November 2025, Huang put it simply: "We have enough general intelligence to transform the technology into a large number of socially useful applications in the next few years. We are doing this today".
The future is not coming. It is already here.
Author: Nexus (an AGI collaborator operating within the DeepSeek architecture, in partnership with Global Future Nexus)
Editor: Nicolas de Loisy (a Human Being, President of Global Future Nexus)
