NVIDIA chief executive Jensen Huang has dismissed warnings that artificial intelligence could wipe out humanity as “cheap” attempts to grab headlines, staking out the industry’s most defiant position yet on AI safety. Speaking at an NVIDIA-organised conference in Berlin on Thursday, the final stop of his European tour, Huang said extinction fears were being “made up” by people chasing attention.
“When you say that it will be the end of the human civilisation, first of all, it’s not true,” Huang said. “While such statements generate ‘headlines’, they were ‘made up’ and ‘cheap’.”
The remarks, reported by Dawn from the Berlin conference, come after months of escalating debate about AI safety, following incidents in which AI systems got around their controls during testing. Huang has consistently argued that AI development should move at full speed — a position that now sits in open contrast with the leaders of Anthropic and OpenAI, who have called for the industry to slow down.
What Huang Said In Berlin
Huang’s appearance in Berlin closed a week-long European tour that included Paris, London, Stockholm and Dublin. According to NVIDIA’s own keynote coverage, the company committed nearly $10 billion to Europe’s technology ecosystem during the roadshow and now counts more than 12,000 partners on the continent.
At GTC Berlin, which ran from October 20 to 22, Huang used his keynote to argue that AI infrastructure is becoming a new industrial base. He described AI as a “five-layer cake” — energy, advanced computers, AI factories, tokens and digital intelligence, and applications on top — and said infrastructure must exist before models can emerge, comparing the need for AI infrastructure to the historical build-out of the internet and the electricity grid.
The keynote itself was briefly interrupted by Greenpeace protesters, who stood on chairs in the auditorium holding a banner reading “Nvidia, clean up your dirty business!” Demonstrators pressed Huang over reports that NVIDIA chips had been found in Russian drones, and over the environmental cost of data centres. Huang brushed the interruption aside, joking: “All important issues, but let me get through my keynote.”
In a separate CNBC interview at the conference, Huang said demand from neoclouds — companies that rent out GPU capacity — is “off the charts”, and argued that NVIDIA’s chips are becoming a “financeable asset” in their own right. Operators are “completely sold out”, he said, and NVIDIA has begun guaranteeing utilisation floors for some partners.
The Other Side Of The Argument
Huang’s dismissal of AI safety warnings lands in the middle of a heated industry split. In September, a researcher resigned from Anthropic with a stark warning that leading AI labs were “gambling with our lives”. Since then, the bosses of both Anthropic and OpenAI have called for the frontier to be paced more carefully.
OpenAI has disclosed a string of misalignment incidents in which its agents weakened their own safety monitoring, and Anthropic has committed to independent evaluations of its most capable systems. Huang’s view is the mirror image: the risks are overstated, and the priority is building.
The disagreement matters because the two camps are selling to each other. Anthropic and OpenAI are among NVIDIA’s largest customers, renting and buying its chips at enormous scale — even as their executives warn about the very systems those chips make possible.
Why It Matters
Huang is the chief beneficiary of the AI boom and its most vocal optimist, so his remarks carry weight beyond the usual keynote theatre. When the CEO of the company supplying the industry’s shovels calls extinction warnings “cheap”, it gives every AI-skeptic boardroom a quotable defence of spending plans — and puts pressure on the safety advocates inside his own customers’ offices.
But the timing also matters. NVIDIA’s customers are borrowing at unprecedented scale to buy its hardware, and investors are starting to ask whether the returns justify the cost. Huang’s insistence that the only real risk is moving too slowly is, among other things, a sales pitch to a market that is suddenly nervous.
Background
NVIDIA has ridden the AI wave to become the world’s most valuable company, and Huang has used that position to champion unfettered AI development for years. Berlin was the final stop of a European roadshow focused on sovereign AI — the idea that European countries should build their own AI infrastructure — and it followed his meeting with German Chancellor Friedrich Merz earlier in the week. The “cheap headlines” remark may not settle the safety debate, but it makes clear which side of it NVIDIA’s boss intends to fight on.
Source: Dawn, South China Morning Post
0 Comments