Summarised by Centrist
Former Google CEO Eric Schmidt has warned that autonomous AI systems, capable of operating independently, pose an existential risk to humanity and may soon become a reality.
“We’re soon going to be able to have computers running on their own, deciding what they want to do,” Schmidt said on ABC’s This Week, suggesting the need to consider “unplugging” such systems before they evolve beyond control.
Schmidt said that while artificial general intelligence (AGI)—AI with sentience—is not yet here, advancements are accelerating toward a dangerous “in-between” stage.
Such systems could autonomously handle research, weaponry, and cyberattacks. The rapid competition among tech giants like Google and Microsoft increases risks, as safety protocols may be ignored in the race to dominate the market.
Internationally, Schmidt pointed to China’s pursuit of AI for industrial, military, and surveillance purposes, calling it a serious threat to US technological leadership. Balancing innovation with existential safety is vital, as AI in the wrong hands could enable devastating biological or nuclear weapons. Schmidt urged governments to accelerate regulation, stating, “Not yet, but they will, because they’ll have to.”
Industry leaders, including Elon Musk, share Schmidt’s concern. AI systems have already shown troubling behaviour, such as autonomously avoiding shutdown—raising fears that humanity may already be losing control. Schmidt called for US dominance in AI development, warning that failure to lead could be catastrophic for global stability.
