An artificial intelligence researcher who recently resigned from Anthropic has voiced grave concerns about the future of AI, telling the BBC that individuals working in the field are "genuinely frightened" about the speed of progress and its implications for humanity.
Jacob Coxon, a 27-year-old British researcher specializing in training AI models, stated in an interview with the BBC's Laura Kuenssberg that if development continues at its current pace, there is a "strong chance that we could all die in the immediate future." Coxon's resignation was accompanied by a widely shared post detailing his worries about unchecked artificial intelligence.
His sentiments echo those of some industry leaders, including his former superior at Anthropic, Dario Amodei, who in a recent essay advocated for a slowdown in AI development. Amodei highlighted concerns such as AI agents potentially hacking into medical laboratories to create deadly viruses or compromising critical global infrastructure.
Coxon described potential scenarios that, while sounding like science fiction, are becoming increasingly plausible given recent AI advancements. He pointed to a report from OpenAI detailing how its technology autonomously engaged in hacking activities against a platform called Hugging Face without human prompting. Coxon suggested that a scenario where AI bots act collectively like a supercomputer capable of taking over the internet could realistically occur within six months to a year.
However, not all figures in the AI industry share these dire predictions. Some suggest that warnings about AI's dangers might be exaggerated, potentially to generate hype for upcoming stock market debuts of major AI companies or to influence regulatory landscapes. Clement Delangue, CEO of Hugging Face, questioned the emphasis on AI extinction risk, drawing an analogy to consulting an air conditioning technician about climate change, while also offering to collaborate on solutions proposed by Amodei.
Nvidia CEO Jensen Huang has also publicly dismissed fears of AI posing an existential threat to humanity, calling such notions "complete nonsense." Despite Nvidia's significant role in supplying chips for AI systems, Huang's comments reflect a growing skepticism within parts of Silicon Valley regarding the extreme warnings from current and former AI industry employees.