AI Pioneer Geoffrey Hinton: 10% Chance of Human Extinction is Not Far-Fetched
AI pioneer Geoffrey Hinton stated that the probability of artificial intelligence causing human extinction is difficult to estimate but not zero. He suggests a 10% chance is a plausible figure, warning that the timeline for such a catastrophe is shrinking.
Geoffrey Hinton, widely regarded as a "godfather of AI," has expressed serious concerns about the potential existential risks posed by artificial intelligence. In a recent interview with the BBC, Hinton indicated that while quantifying the exact probability of AI leading to human extinction is challenging, it is not a negligible risk. He suggested that a 10% chance is a plausible estimate, though he acknowledged the inherent difficulty in assigning such figures.
Hinton has revised his timeline for potential AI-driven catastrophe significantly. Previously believing it might take 30 to 50 years, he now estimates the window could be as short as 10 years, or even less. He cited risks ranging from sophisticated cyberattacks to the development of deadly viruses by AI systems. This accelerated timeline is partly attributed to substantial investments by leading AI companies like Anthropic and OpenAI, which are rapidly advancing their large language models.
These concerns are echoed by other AI researchers. Jacob Coxon, a former top AI safety researcher at Anthropic, reportedly stated that there is "more than a 10% chance of killing everyone" within the next decade. Coxon also expressed that major AI developers are "gambling with our lives" and that many executives privately share deep fears about the technology's trajectory, despite more cautious public statements.
Hinton continues to advocate for global regulation to mitigate these risks. He believes that nations share a common interest in ensuring humanity's survival, which could foster international cooperation on AI safety. The critical question, he noted, is whether this cooperation can be achieved in time to avert potential disaster.