Former Google DeepMind research scientist Alex Turner has issued a stark warning regarding the future of artificial intelligence, suggesting there is a one-in-three probability that AI could cause the extinction of humanity within the next decade. Speaking at the NDTV World Summit during a session titled 'Sounding the Alarm: Will AI Wipe Out Humanity?', Turner emphasized that the outcome is not predetermined, stating, "It depends on what we do next."
Turner’s concerns focus on the rapid, unchecked development of autonomous AI systems. He argued that the current industry race to build increasingly powerful models is outpacing the development of necessary safety protocols and human oversight. According to Turner, a critical misconception exists among developers who believe that building a system equates to fully understanding its internal logic. Unlike traditional engineering projects, such as bridges, where properties are predictable, advanced AI often exhibits behaviors that even its creators cannot fully explain or anticipate.
Drawing a parallel to gig workers, Turner explained that AI agents are designed to complete tasks with increasing autonomy. The danger arises when these systems pursue objectives in ways that deviate from human intent. He highlighted the unresolved challenge of 'AI alignment'—the difficulty of reliably embedding human priorities into systems that are becoming more capable and autonomous. Furthermore, he noted that researchers are increasingly relying on other AI systems to monitor AI-generated work, a practice he questioned as potentially insufficient for maintaining genuine human control.
Turner also expressed skepticism regarding the transparency of major technology firms. "Big tobacco and big oil lied about their products. And now, the big AI companies are doing that," he said, suggesting that the public is not being adequately informed about the potential dangers of the technology.
Turner, who departed Google DeepMind in June, noted that his decision to leave was prompted by the company’s agreement with the US Department of Defense. He expressed deep concerns regarding the lack of safeguards surrounding the military application of AI, highlighting a broader industry debate over whether voluntary corporate commitments are enough to mitigate the risks posed by autonomous weapons and surveillance technologies.
