While every other tech CEO is tiptoeing around the "A-word," Nvidia's boss just grabbed the bull by the horns.
Nvidia CEO Jensen Huang appeared on the Lex Fridman Podcast this week and said something that made the entire industry choke on their coffee: "I think we've achieved AGI."
Just like that. No hedging. No lawyer-approved phrasing. Just vibes, and a very expensive GPU empire behind him.
To understand why this is a big deal, you need to know what AGI (Artificial General Intelligence) actually means.
You see, the current AI is a really, really smart dog, it can learn one trick incredibly well (like writing code or drawing). But AGI would be more like a person, able to figure out any task, even brand new ones, just by thinking it through.
Scientists have been arguing about when (or if!) we'd ever get there. And now the boss of the company whose chips power literally every major AI just said: "We're already there."
Here's the twist: Almost every other tech company has quietly been running away from the word "AGI." OpenAI now talks about "levels" of AI capability. Anthropic uses the phrase "advanced AI systems." The industry basically agreed to stop hyping AGI because it sounds scary and invites regulators to the party.
Then Jensen Huang walked in and flipped the table.
Critics, and there are many, will point out that today's AI still makes stuff up, struggles with basic logic, and doesn't truly understand anything.
Passing a bar exam is impressive; knowing why it matters is something else entirely. So is Huang right, or has the definition of AGI just been stretched so far it now fits whatever we've already built?
That's the question keeping researchers up at night. And with Nvidia sitting at a $2 trillion valuation, when Huang talks, the industry can't just scroll past it.
