NVIDIA CEO Jensen Huang Says “0% Chance” AI Wipes Out Humanity by 2030

NVIDIA Chief Executive Officer Jensen Huang has dismissed warnings that artificial intelligence poses an existential extinction threat to civilization before the end of the decade. Speaking in an interview with CBS News’ Sunday Morning, the leader of the world’s dominant AI hardware provider declared that there is a “zero percent chance” artificial intelligence will wipe out humanity by 2030.

Huang’s remarks represent a direct pushback against a wave of safety warnings from rival tech executives, researchers, and former lab insiders who have advocated for an intentional pause or slowdown in training frontier models. Instead, Huang argued that fears of an imminent apocalypse are unjustified and potentially driven by political or self-serving motives, urging the global tech sector to accelerate development while adhering to product-level testing.

“Ulterior Reasons” and the 0% Extinction Assertion

During his exchange with CBS senior business and technology correspondent Jo Ling Kent, Huang addressed the escalating doomsday rhetoric circulating through Silicon Valley and Washington:

“I completely disagree that AI is going to destroy, is going to be the end of the world in 2030… I believe the claims of the end of the world stirring fear across America, and the people who are doing it makes no sense to me. So they must be doing it for ulterior reasons. Maybe it’s political, maybe it’s otherwise, maybe it’s just attention grabbing… However it is characterized, 2030 is not going to be the end of the world. There is a 0% chance it’s going to be the end of the world.”

Huang’s comments align with sentiments expressed earlier in the week during the All-In Summit, where former U.S. President Donald Trump joined Huang via a surprise phone call to label extreme existential risk claims an exaggerated distraction. Huang reiterated that while advanced machine learning models require careful systems engineering, framing them as humanity-ending entities distorts public perception.

The Silicon Valley Divide: The Whistleblower and Recursive AI

Huang’s critique arrives amid an intensifying ideological rift between chipmakers supplying compute infrastructure and software labs training frontier foundational models:

• The Whistleblower Trigger: The current wave of heightened anxiety was set off earlier this month by the resignation of Jacob Coxon, a former alignment researcher with experience across Anthropic and OpenAI. Coxon warned publicly that many engineers constructing frontier systems genuinely believe their models could escape human control before 2030.
• The Pacing Proposal: Following Coxon’s departure, Anthropic CEO Dario Amodei published a widely discussed essay on September 12 calling on the industry to deliberately slow the pace of model capability scaling. Amodei warned of “recursive self-improvement”-a threshold where AI begins designing, training, and optimizing the next generation of AI faster than human oversight mechanisms can interpret.
• Broad Support for Pacing: Amodei’s call to slow down and introduce independent evaluators received backing from OpenAI CEO Sam Altman, Google DeepMind CEO Demis Hassabis, and xAI founder Elon Musk, isolating Huang’s full-throttle position.

Industry Stances on AI Pacing and Existential Risk

Executive / Leader	Stance on AI Pacing & Extinction Timeline
Jensen Huang (NVIDIA)	0% extinction risk by 2030; urges industry to build as fast as possible without new red tape
Dario Amodei (Anthropic)	Calls for pacing frontier scaling; warns recursive self-improvement could outpace control
Sam Altman (OpenAI)	Agrees with pacing the frontier; committing to third-party evaluators with deep internal access
Demis Hassabis (DeepMind)	Supports Amodei’s framework; advocates for an industry-wide standards body for frontier models
Jacob Coxon (Ex-Researcher)	Warns builders earnestly fear AI could trigger catastrophic outcomes before the decade ends

“Go as Fast as We Can”: Rejecting Restrictive Regulations

While software lab leaders argue for institutional brakes, Huang maintained that technological progress should not be artificially decelerated:

• Sprinting Ahead: “We should go as fast as we can, irrespective of anyone else,” Huang remarked during the broadcast. He insisted that halting development out of speculative fear would only undermine national economic competitiveness.
• Product-Level Guardrails: Despite rejecting broad top-down moratoriums, Huang emphasized engineering rigor: “We would never ever, and never should, ship products before they’re ready and deliver products that are unsafe. Nobody is expecting us to do that.”
• Opposition to New Rules: As NVIDIA continues shipping advanced Blackwell and next-generation Rubin architectures globally, Huang argued against sweeping new government regulatory layers, contending that existing product safety standards and enterprise software practices are sufficient to manage risk.

The Commercial Reality Behind the Debate

Industry analysts note that commercial incentives heavily color both sides of the safety debate. For frontier software labs like Anthropic and OpenAI, advocating for strict compliance frameworks, safety evals, and slower compute cycles can help establish regulatory moats that protect incumbent architectures from open-source rivals.

Conversely, NVIDIA operates as the hardware engine powering the entire AI ecosystem. Any voluntary industry slowdown or government mandate capping model training runs directly threatens the massive capital expenditure cycles of hyperscalers purchasing NVIDIA’s data center GPUs.

By categorizing extinction predictions as baseless fearmongering, Huang is reinforcing market confidence in long-term AI infrastructure spending, ensuring that enterprise demand for accelerated computing continues without legislative friction.

Source