Jensen Huang says AGI has already arrived and AI could run companies, but he also warns hype fades and that building a firm like Nvidia with agents is still unrealistic.
What Jensen Huang Said on Lex Fridman’s Podcast
Nvidia CEO Jensen Huang has sparked a fresh debate after saying he believes artificial general intelligence (AGI) has already been achieved. Speaking on Lex Fridman’s podcast, Huang responded directly when asked how long it would take to reach AGI: “I think it’s now. I think we’ve achieved AGI.”
His comments were framed as a view on the broader AI industry, not a claim that Nvidia itself has built a single “AGI system.” Nvidia has benefited heavily from the AI boom because its chips and platforms are widely used to train and run modern AI models. So when Huang talks about the state of AI, people listen closely because his company sits at the center of the supply chain.
The statement is bold, but it also depends on what someone means by “AGI.” That is the core reason this topic becomes controversial quickly: different leaders define AGI in very different ways.
Can AI Run a Company? The ‘Agent’ Idea Explained Simply
Huang said it is “possible” for AI to operate a company. This does not necessarily mean a single AI replacing every human role. In practice, it usually means a network of AI “agents” handling tasks across a business: customer support, reporting, scheduling, compliance checks, drafting documents, and even some decision support for managers.
He pointed to OpenClaw, described as an open-source AI agent platform designed to function autonomously on behalf of users. The idea behind such platforms is that an agent can take actions, not just answer questions. For example, it could read messages, access files, coordinate with other tools, and complete a workflow end to end.
That is the real shift happening in AI right now. The industry is moving from chatbots that talk, to agents that do. If agents become reliable and safe, they could automate large parts of office work. That is why the “run a company” idea is being discussed seriously, even if it is not fully practical yet.
Why ‘AGI’ Is Hard to Define and Easy to Argue About
Artificial general intelligence is typically described as AI that can match or exceed human intelligence across many tasks, including complex planning and strategic decisions. But there is no single global standard test for AGI. Some people define it as “human-level performance in most cognitive tasks.” Others define it as “economic impact,” meaning AI can do most work humans do.
Because the definition is unclear, timelines are also unclear. One person may say AGI is here because models can write code, pass exams, and handle multi-step tasks. Another person may say AGI is not here because models still hallucinate, struggle with real-world context, and lack true accountability.
This is why different leaders disagree. If you define AGI as “useful general capability,” you can argue we are close or already there. If you define it as “reliable human replacement across domains,” you can argue we are still far away.
Altman vs Nadella: Two Very Different Views of the Same Trend
The debate is not limited to Huang. Sam Altman has said the industry is close to AGI, but he has also described progress as gradual rather than a single dramatic breakthrough. That view suggests AI will keep improving in steps, and the “AGI moment” may be more of a transition than a sudden event.
Microsoft CEO Satya Nadella has taken a more sceptical position, saying the industry is not close to achieving AGI and warning against relying on individual claims. This perspective emphasises measurable outcomes and real-world reliability over hype.
These disagreements matter for businesses and governments because policies, investments, and workforce planning can be influenced by how leaders frame the future. If decision-makers believe AGI is already here, they may push faster automation. If they believe it is still far, they may prioritise slower adoption and stronger controls.
Huang’s Caution: Hype Fades and ‘Building Nvidia’ Is Still 0%
Even though Huang called AGI “now,” he also added caution. He noted that many AI tools tend to lose user interest over time. This is a practical reality: many products launch with big promises, but if they are unreliable or complicated, people stop using them.
He also said the likelihood of AI agents building a company like Nvidia remains “0%.” That is a strong counterbalance to the headline claim. In simple terms, he is saying: AI can automate tasks and maybe run parts of a company, but building a world-class firm from scratch—creating culture, strategy, product leadership, and long-term execution—still depends on humans.
For readers, the most balanced takeaway is this: AI agents are becoming powerful enough to take on more real work, and companies will increasingly use them. But “AGI is here” depends on definition, and full corporate autonomy is still far from proven in the real world.
