Jensen Huang Dismisses AI Doomsday Warnings and Defends Nvidia’s Financing Strategy

Nvidia CEO Jensen Huang offered a sweeping defense of the AI buildout at Goldman Sachs’s Communacopia + Technology Conference. He rejected both the existential-risk case made by some AI researchers and investor concerns that Nvidia’s infrastructure investments are helping fund demand for its own chips.51

Huang’s answer to AI extinction warnings

Huang described fears that advanced AI could lead to human extinction as “complete nonsense.” His comments followed public warnings from former Anthropic researcher Jacob Coxon, who argued that leading labs were racing toward self-improving AI without acting responsibly.53

The dispute is not simply about whether current AI systems are dangerous. Coxon’s warning concerns the trajectory toward more capable systems and whether developers can reliably control them. Reporting on the debate noted that Anthropic alignment lead Evan Hubinger publicly supported the broader concern and said he personally assigned a greater than 10% chance to an AI-driven extinction outcome.8

Huang’s response was a rejection of that premise, not a technical resolution of it. He presented AI progress as a field that can be engineered and deployed responsibly, while the researchers pressing for caution argue that safety work and governance may not be keeping pace with capability development.36

His criticism of cybersecurity alarmism

Huang also suggested that cybersecurity fears can become commercially useful: he said the industry’s intense discussion of cyber threats coincides with cybersecurity products coming to market, asking what better way there is to create demand than to create a problem.7

That remark frames some AI-risk rhetoric as market-driven rather than as evidence that systems are inherently uncontrollable. But it does not settle the underlying security question. The relevant issue is whether AI systems materially expand cyber capabilities and, if so, whether safeguards can manage that risk—not merely whether security vendors stand to benefit from heightened concern.67

Why Huang rejects the “circular financing” label

Huang laughed off accusations that Nvidia is financing its own sales. Asked about financial backstops for data-center projects and AI infrastructure, he replied: “I put in one, and a hundred comes back. Is that circular?” He characterized the arrangements as a strategy to build distribution channels for Nvidia’s architecture.1

In Huang’s view, investments or commitments involving AI infrastructure customers are not equivalent to fabricated demand. The intended outcome is a larger ecosystem of AI data centers, cloud capacity, and applications that creates durable demand for Nvidia systems.

That distinction matters. An investment in a customer is not automatically sham revenue. Yet critics of circular financing are focused on economic dependency: when chip suppliers, cloud providers, infrastructure companies, and AI labs have overlapping investments and purchasing commitments, a slowdown in outside funding or end-user demand could affect several linked businesses at once.

The credibility test for Nvidia’s AI thesis

Huang’s argument is fundamentally techno-optimist: AI demand is real, infrastructure is still being built, and Nvidia’s capital can help unlock a much bigger computing market. He reiterated a forecast of $3 trillion to $4 trillion in global AI-infrastructure spending by 2030, while Nvidia continued to target roughly 70% year-over-year revenue growth.218

The opposing view does not require assuming that Nvidia’s reported sales are illegitimate. It asks whether the scale and durability of AI infrastructure spending will ultimately be supported by independent customer revenues, rather than by a financing-intensive expansion cycle.

For readers and investors, the two debates should be kept separate:

• AI safety: Huang rejects extinction-risk warnings, while some researchers argue that more capable AI warrants slower development and stronger safeguards.56
• AI economics: Huang says Nvidia’s investments create productive infrastructure and distribution; skeptics focus on how much future demand depends on interconnected capital commitments.1

Huang’s conference remarks made clear that Nvidia sees neither concern as a reason to slow the buildout. Whether that confidence proves justified will depend on the real-world returns from AI deployment—and on whether safety and financing risks are managed as the industry scales.