rki.news | Sources Anadolu
WASHINGTON, Sept. 14 — Tech entrepreneur Elon Musk said Monday that his concerns about the potential risks of artificial intelligence date back more than a decade, highlighting his early involvement in discussions about advanced AI safety.
Musk said on X that he had been thinking about AI safety long before 2014. He cited his contribution to philosopher Nick Bostrom’s 2014 book Superintelligence: Paths, Dangers, Strategies, which examines the possible consequences of AI surpassing human intelligence.
Musk said he was thanked by name in the book’s foreword. He also recalled recommending the book in an August 2014 post, warning that advanced AI could be “potentially more dangerous than nukes.”
His comments come as technology leaders debate safeguards for increasingly advanced AI systems.
Anthropic Chief Executive Dario Amodei recently proposed allowing independent safety evaluators to access AI companies and verify their safety practices. The proposal has received support from OpenAI Chief Executive Sam Altman and Musk.
Amodei said there are real dangers in AI development, while stressing that building systems safely could keep the probability of serious harm low. He warned that developing AI incorrectly could make the risk significantly higher.
Musk Says AI Safety Concerns Began More Than Decade Ago
