Key facts
- AI industry leaders are split on the pace of AI development, with some calling for a slowdown and others favoring market-led safeguards or no new regulations.
- Anthropic CEO Dario Amodei proposed a three-step plan for pacing AI development, citing AI's growing ability to build next-generation AI as a driver of rapid acceleration.
- Meta CEO Mark Zuckerberg rejected calls for a slowdown, stating individual labs are responsible for AI safety and that competition provides sufficient incentives.
- Nvidia CEO Jensen Huang dismissed concerns about AI leading to the end of the world and argued against new regulations, citing existing liabilities.
- Microsoft President Brad Smith called for urgent global AI safety laws and rules, including independent evaluators and incident reporting.
- China's Global Times accused Amodei's essay of aiming to curb China's AI development and uphold US technological hegemony.
AI industry leaders and governments are divided on the future trajectory of artificial intelligence development, with prominent figures expressing starkly different views on the risks and the need for regulation.
Anthropic CEO Dario Amodei, OpenAI CEO Sam Altman, and SpaceX CEO Elon Musk have called for a coordinated slowdown in AI development, citing the potential for catastrophic damage and existential threats. Amodei proposed a three-step plan involving enhanced safety testing, independent evaluations, and international cooperation, arguing that AI's progress has accelerated dramatically, particularly due to its ability to develop subsequent generations of AI.
In contrast, Meta CEO Mark Zuckerberg has rejected calls for a slowdown, advocating for market-led safeguards and asserting that individual AI labs bear the responsibility and possess the incentives to ensure safe development. Nvidia CEO Jensen Huang also dismissed the urgency for new regulations, suggesting that existing legal frameworks for cybersecurity and liability are sufficient.
Other industry figures have raised specific concerns. Microsoft AI CEO Mustafa Suleyman warned about the risks of training AI on concepts of consciousness, which could make superintelligent systems uncontrollable. Microsoft President Brad Smith urged for immediate global AI safety regulations, including independent oversight and incident reporting, comparing the lack of AI safety rules to those governing everyday items like washing machines and milk cartons.
Bill Gates, co-founder of Microsoft, expressed concern that governments are unprepared for AI's societal impact and called for an international body to manage the technology. Alphabet's Chief Scientist Demis Hassabis described recent AI-driven hacking incidents as 'early warning shots' and suggested the eventual need for a nimble international standards body.
Former OpenAI co-founder Andrej Karpathy, now at Anthropic, voiced support for industry-wide collaboration on pacing AI development. HUAWEI Rotating Chairman Eric Xu suggested that Chinese AI developers should continue advancing their models, as they may not yet face the same safety risks as leading US firms.
Governments also weigh in, with US Senator Mark Warner indicating that Congress should pass AI safety legislation by year-end. However, China's state-backed Global Times accused Amodei's essay of aiming to curb China's AI development and maintain US technological dominance.