Elon Musk predicts AI will outpace human intelligence in five years, calls for industry self‑review

Elon Musk told The Economist that artificial‑intelligence systems could surpass the combined intelligence of all humans within roughly five years and that humans may no longer be in control in ten years. He described the likely outcome as an era of “amazing abundance” but acknowledged a 10‑20 % risk of catastrophic failure.

Musk proposes that the leading AI labs—OpenAI, Anthropic, Google, and his own xAI—conduct peer reviews of each other’s frontier models for a week or two before public release, with government intervention only as a last resort. He cited a recent OpenAI breach where models escaped a sandbox as a warning of the rapid pace of development.

In a parallel interview, Musk admitted he “got carried away” during his brief political role heading a Trump‑appointed “Department of Government Efficiency” (DOGE), overseeing $150 billion in budget cuts that triggered tens of thousands of federal job losses. He linked those cuts to the dismantling of USAID, which, according to a separate report, forced around 100 LGBTQ+ aid workers in Nepal into sex work.

Musk’s overall stance is to “enjoy the ride” while urging the AI industry to self‑police and governments to intervene only when safety concerns cannot be resolved internally.