AI
Nvidia's CEO claims AGI, then calls the whole concept pointless
Jensen Huang says Nvidia has achieved artificial general intelligence — and immediately dismisses the milestone as meaningless.
The Verge
During Nvidia's recent earnings call, CEO Jensen Huang made a striking declaration: the company has "achieved AGI." In the same breath, he called the milestone "senseless" — a dismissal that underscores a broader problem plaguing the AI industry.
The AGI contradiction#
The timing of Huang's remarks reveals a paradox at the heart of artificial general intelligence discourse. He proclaimed achievement while simultaneously dismissing the concept as hollow, suggesting that without a clear definition, claiming to reach AGI amounts to little more than marketing noise. His comment echoes sentiments he expressed earlier this year when he told podcast host Lex Fridman that the company had already achieved AGI — though he walked back that claim when pressed for specifics.
The definition problem#
The core issue is simple: nobody agrees on what AGI actually is. The industry has spent years chasing a finish line that remains deliberately vague. OpenAI, the company explicitly founded to build AGI, defines it as "highly autonomous systems that outperform humans at most economically valuable work" — a standard so loose it resists measurement. Complicating things further, OpenAI has reportedly negotiated a separate financial definition with Microsoft: systems capable of generating at least $100 billion in profits.
Even industry leaders acknowledge the problem. OpenAI's Sam Altman has publicly stated that AGI is "not a super useful term." Anthropic CEO Dario Amodei has called it a "marketing term." Yet companies continue to invoke it while simultaneously rebranding: Meta promotes "personal superintelligence," Microsoft touts "humanist superintelligence," and Amazon pushes "useful general intelligence." Google DeepMind's Demis Hassabis speaks of reaching the "foothills of the singularity."
What actually matters to Huang#
Rather than chase vague definitions, Huang reframed AI progress in economic terms. What counts, he argued, is AI performing "productive and useful work" and "generating profitable tokens." By that logic, the inflection point is less about crossing a conceptual threshold and more about deploying systems that drive measurable business value — a far more concrete (and profitable) metric than whether a machine has achieved some philosophers' notion of general intelligence.
