Nvidia CEO Jensen Huang has sharply criticized scientists who issue alarming forecasts about artificial intelligence causing societal collapse, arguing that such predictions lack scientific grounding and actively harm the field.

In a recent appearance on The Ezra Klein Show, Huang pushed back against what he described as “AI doomer” narratives, insisting they are not based on rigorous research and discourage young people from entering the industry.

He specifically addressed estimates from AI pioneer Geoffrey Hinton, who suggested a 10-20% chance that advanced AI systems could lead to societal collapse.

Huang rejected the figure, stating there is no scientific research supporting such precise probabilities. He argued that simply coming from a prominent scientist does not make a prediction scientific.

He said,

“I would tell Geoff that it’s irresponsible to say all that. All of his predictions have been wrong. Enough predictions. That 10 percent chance is not grounded in science. It’s not grounded in research. Just because it comes from a scientist doesn’t make it scientific. Those predictions are hurtful.”

“Is it good or bad that we scare young people about the future of A.I., so much so that they don’t even want to go to universities and don’t want to go to college anymore because they don’t think they’ll get a job? Is that helpful or hurtful, if it were to happen? It’s hurtful.

“Don’t think for a second just because you’re an alarmist that you’re doing a social good. It is not true. So I think that we ought to just all be wiser, more mature, be evidence-based, be scientific. If you want to be scientific, be scientific. Do the science.”

According to Huang, these forecasts have a poor track record and create unnecessary fear that steers talented students away from AI-related careers at a time when the technology needs more, not fewer, skilled contributors.

The Nvidia chief framed AI safety primarily as an engineering challenge rather than an existential philosophical crisis. He maintained that the industry should focus on practical, evidence-based approaches to building reliable systems instead of amplifying speculative worst-case scenarios.

Huang’s remarks come amid an ongoing and often polarized debate within the AI community. Figures such as Hinton, along with researchers associated with organizations focused on existential risk, have repeatedly warned that rapid progress in AI capabilities could outpace society’s ability to control or align the systems.

In defense, Huang’s comments reflect his longstanding position as one of the most prominent AI optimists in technology. As the leader of the company that supplies the majority of the specialized chips powering modern large language models and generative AI systems, he has consistently emphasized the transformative benefits of the technology while downplaying catastrophic risk narratives.

In contrast, Huang and other industry leaders argue that exaggerated doomerism risks slowing innovation, reducing investment, and creating regulatory overreactions that could cede technological leadership to less cautious actors.

Huang made clear which side he occupies. He expressed concern that constant emphasis on collapse scenarios could deter the next generation of engineers and researchers precisely when their contributions are most needed to improve safety, reliability, and usefulness.

Nvidia’s central role in the AI boom gives Huang’s words significant weight. The company’s GPUs remain the dominant hardware platform for training and running the largest AI models.

Any shift in public or regulatory sentiment driven by doomer narratives could affect demand for those chips, research priorities, and talent pipelines. Huang appears determined to counter the more alarmist messaging by stressing evidence, practical progress, and the tangible benefits already emerging from AI systems in science, industry, and everyday applications.

Outlook

The debate over AI’s long-term risks is likely to remain contentious as models become more capable and increasingly integrated into the economy.

Huang’s comments could reinforce the view among technology companies and investors that AI development should remain focused on measurable safety improvements, engineering controls, and practical applications rather than speculative probabilities of societal collapse. At the same time, warnings from researchers such as Hinton are unlikely to disappear.

As AI systems become more autonomous and capable of performing increasingly complex tasks, questions around alignment, misuse, labor-market disruption, and the ability of institutions to govern advanced systems are expected to remain central to the conversation.