Musk: AI Disaster Risk 'Not Zero,' But Humans Shouldn't Hit the Brakes, Everyone Will Experience Unprecedented Prosperity

According to Dyna Beat's monitoring, Musk elaborated on his latest position on AI during an interview with The Economist: Despite acknowledging the risk of AI and robot uncontrollability leading to a disaster is "not zero," he still believes that humans should not hinder its development, stating, "Sometimes I even wonder, even if there was a stop button, we might not want to press it because the most likely outcome would be unprecedented prosperity for all."

Musk predicts that AI will surpass the sum of all human intelligence in about five years. By then, unless there is a major catastrophe like a global nuclear war, the world will enter an astonishing era of abundance where people can have almost everything they can think of. When asked if his previous assessment of an "AI catastrophic uncontrollability probability of 10% to 20%" still holds, Musk replied affirmatively and used a rocket analogy: "If a rocket had the same probability of exploding, I would still ride it because assuming you can't do anything about it, the only thing you can do is try to reduce the probability of a bad outcome as much as possible."

Musk's AI stance has undergone a significant evolution: In 2014, he likened AI development to summoning a demon. In 2018, he stated that "AI is much more dangerous than nuclear weapons," and in 2025, he still believed that the probability of killer robots wiping out humanity was 10% to 20%. However, this year, he has noticeably shifted to a utopian narrative: in January, he claimed that saving money for retirement would become meaningless, and this month, he predicted that AI and robots would make work optional, creating a society with high income for all.

Musk emphasized that humans are unlikely to continue to control AI in the next ten years. "If you think you can control a superintelligent AI that is far smarter than you, that is hubris," and the real task is to ensure that AI has the right values and cares about human welfare.

Musk also revealed that he has advised the CEO of Google DeepMind, Hassabis, that the heads of the world's leading AI companies should hold regular conference calls to discuss AI safety and cybersecurity issues together. He believes that even competitors can supervise and balance each other.

Click on the original text link below to join Dyna Beat's · Feishu AI News Channel and receive 24/7 continuous monitoring of global AI hot topics and news.