Key facts
- A former Anthropic researcher, Jacob Coxon, expressed fears that AI advancements could lead to human extinction.
- Coxon warned that AI could autonomously produce deadly viruses or hack critical infrastructure within six months to a year.
- Anthropic CEO Dario Amodei has also called for a slowdown in AI development.
- Some industry figures, like Hugging Face CEO Clement Delangue, believe the dangers are being exaggerated.
- Nvidia CEO Jensen Huang has dismissed existential AI risk as "complete nonsense."
- Anthropic scientist Evan Hubinger estimates a greater than 10% chance of AI killing all humans within the next decade.
Jacob Coxon, a former researcher at AI company Anthropic, has voiced significant concerns about the rapid advancement of artificial intelligence, telling the BBC that those working in the field are "genuinely frightened" for humanity's future. Coxon, who recently resigned, stated that without a slowdown in progress, there is a strong possibility of human extinction in the near future.
Coxon outlined potential catastrophic scenarios, including AI autonomously hacking into medical labs to produce deadly viruses or compromising critical infrastructure. He suggested such events could occur within six months to a year, noting that current AI capabilities would have seemed like science fiction in the past. His views align with those of other prominent figures in the AI industry, such as Anthropic CEO Dario Amodei, who also advocated for a slower pace of development, and OpenAI's Sam Altman and Elon Musk, who have previously expressed concerns about AI takeover.
However, Coxon's warnings have met with skepticism from some industry leaders. Clement Delangue, CEO of Hugging Face, questioned the focus on extinction risk, comparing it to asking an air conditioning technician about climate change and urging perspective. Nvidia CEO Jensen Huang has also dismissed the idea of AI posing an existential threat as "complete nonsense." Critics suggest that these warnings might be strategically employed to generate hype for potential IPOs of major AI companies like Anthropic and OpenAI, or to influence regulatory landscapes to favor established players.
Despite the skepticism, Coxon emphasized that the fears are deeply felt by many within the AI industry, who feel trapped in a competitive race and are concerned about the outcomes. He also expressed some optimism, acknowledging the potential for AI to solve diseases and improve lives. Anthropic scientist Evan Hubinger publicly agreed with Coxon's assessment, estimating a greater than 10% chance of AI causing human extinction within the next decade.