Santa Clara: Nvidia CEO Jensen Huang has rejected fears that artificial intelligence could lead to human extinction by 2030, while arguing that the industry should continue developing AI systems as quickly as possible without compromising safety.
Speaking to CBS News' Jo Ling Kent for Sunday Morning, Huang said he completely disagreed with predictions that AI would destroy the world by the end of the decade. He put the likelihood of such an outcome at "zero per cent". CBS reported that the interview will air on September 20.
Huang also rejected calls for a coordinated slowdown in AI development.
"We should go as fast as we can irrespective of anybody else," he said.
At the same time, he said companies should not release products before they are ready or deploy systems that are unsafe.
"We're going to go as fast as we can, but we would never ever, and never should, ship products before they're ready, deliver products that are unsafe," Huang said.
Huang's position comes amid a growing debate within the AI industry over whether the rapid expansion of AI capabilities is moving faster than safety measures and human oversight.
Anthropic CEO Dario Amodei has warned that the pace of development is a concern. CBS reported that Amodei described the rapid progress of AI as a "warning sign" and said the industry needed to slow down, while stressing that this did not mean AI development should be halted altogether.
One area of concern is recursive self-improvement — the possibility that increasingly capable AI systems could contribute to the development of subsequent generations of AI systems.
Anthropic has separately published research and measurements aimed at tracking how much of AI research and development is itself being performed by AI systems. The company said increasing automation of AI research could accelerate development while also making human oversight and control more difficult.
The debate therefore centres not simply on whether AI development should continue, but on how quickly it should proceed and what safeguards should accompany increasingly capable systems.
Concerns have also been voiced by people who have worked inside leading AI companies.
Reporter World's exclusive compilation by Aaron Philip Keechireth examines warnings from researchers and former employees at companies including Anthropic and OpenAI. The article looks at former Anthropic researcher Jacob Coxon, who publicly criticised the pace of AI development after leaving the company, as well as concerns raised by Anthropic's Alignment Science Lead Evan Hubinger and OpenAI chief scientist Jakub Pachocki.
Coxon said AI companies were racing towards self-improving systems, while Hubinger publicly discussed his own assessment of the potential risks. Pachocki separately wrote about concerns over the consequences of a continued rapid rise in machine intelligence. These are individual assessments and warnings, rather than established predictions of what AI will ultimately do.
AI developers themselves are increasingly divided over how fast the technology should advance. While Nvidia's Huang argues that development should move ahead rapidly with safety safeguards, researchers and former employees at leading AI companies have raised concerns about the risks of increasingly capable systems.
Read in detail: Artificial Intelligence and the threat to humanity: Industry experts voice concerns — Aaron Philip Keechireth's exclusive compilation examines warnings from people who have worked inside Anthropic and OpenAI, including Jacob Coxon, Evan Hubinger, Mrinank Sharma, Jakub Pachocki and Jan Leike.
(With inputs from ANI | CBS News | Anthropic | Reporter World exclusive compilation by Aaron Philip Keechireth)