TLDRocket
Sign in

Jensen Huang says Nvidia achieved AGI, again — not that it matters

The Verge Robert Hart

Opinion — commentary, not a factual news event.

Jensen Huang said Nvidia has “achieved AGI” on Wednesday, then called the idea “senseless.” The weird part: he may be right that nobody can agree what AGI even means.

Based on reporting by The Verge, Robert Hart — read the original for the full story.

Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error

On Nvidia’s earnings call Wednesday, Jensen Huang tossed off a line that sounded like a finish line: the company had “achieved AGI.” Then he immediately stepped on it, calling the milestone “senseless.” That’s a very Nvidia way to talk about the holy grail of AI — claim the prize, then question whether the prize exists.

Huang’s point is hard to dismiss. Artificial general intelligence is the industry’s favorite destination sign, but there’s no shared definition sitting at the end of the road. If nobody can agree on what AGI is, then nobody can agree on when it arrives. The result is a goalpost that moves every time someone gets close.

Asked about OpenAI’s pursuit of AGI, Huang said Nvidia could already make the case that it had reached it “for many tasks.” That is not the same as saying the company solved the mystery of general intelligence. It’s a much more slippery claim, and one that fits the moment: everyone wants the label, even as the label itself gets fuzzier.

So the real story isn’t that Nvidia hit AGI. It’s that a company sitting at the center of the AI boom can say that with a straight face and then shrug it off in the same breath. That tells you more about the state of AI hype than about any actual finish line.

My take — AI-written commentary, not fact-checked reporting

AGI is turning into the tech industry’s favorite magic trick: wave the phrase around, then explain why it means nothing. Huang is doing the sensible thing here, which is to enjoy the mystique while refusing to pretend the emperor has a benchmark. The bigger problem is that everyone keeps rewarding this vagueness because vague superintelligence talk sells better than plain engineering.

Read more about this at: The Verge

Related stories

The daily briefing

Every AI story that matters, in your inbox by 8am.

TLDRocket reads all relevant sources, removes duplicate coverage, and summarises the day in two minutes. Follow companies and topics for alerts, or get the briefing in Slack. Free, no spam, unsubscribe anytime.