During Nvidia’s quarterly earnings call on Wednesday, CEO Jensen Huang told analysts the company had "achieved AGI," a phrase that usually signals a breakthrough in artificial general intelligence. The proclamation was short‑lived; minutes later Huang called the milestone "senseless," adding that the industry’s obsession with defining AGI has become a distraction.
Huang did not offer a concrete definition of artificial general intelligence. Instead, he suggested that for many tasks Nvidia’s chips and software already operate at an AGI level, though he stopped short of naming specific benchmarks. He pointed to the next wave of AI—autonomous agents that can learn new skills, improve themselves recursively, and go beyond responding to simple prompts—as the practical frontier.
"What matters is AI doing productive and useful work," Huang said. He framed the conversation in terms of "profitable tokens" and the relationship between compute power and revenue, noting that more compute yields more tokens, which in turn drives profit. "This is the exact phase where we’re at, which is why everybody’s leaning in," he added.
The claim echoes a March appearance on the Lex Fridman podcast, where Huang asserted, "I think we’ve achieved AGI." That interview offered no clear definition either, prompting Fridman to propose his own: an AI system capable of running a tech company worth over $1 billion. Huang later qualified the statement, saying the odds of hundreds of AI agents building Nvidia from scratch are "zero percent."
Huang’s remarks arrive at a time when the tech community remains divided over what constitutes AGI. OpenAI, the organization that set out to build it, describes AGI as "highly autonomous systems that outperform humans at most economically valuable work," a definition its own leadership admits is hard to measure. Other industry figures, including Anthropic’s Dario Amodei and Google DeepMind’s Demis Hassabis, have called the term vague or a marketing buzzword, preferring descriptors like "powerful AI" or "personal superintelligence."
Despite the lack of consensus, executives continue to invoke AGI in earnings calls and interviews, often to underscore a company’s leadership in AI research. Nvidia’s latest assertion underscores the company’s confidence in its hardware and software stack, while simultaneously acknowledging that the industry’s language around AGI may be more hype than hard science.
Questo articolo è stato scritto con l'assistenza dell'IA.
News Factory APP - notizie agentiche per potenziare il tuo SEO e AEO.