A former researcher at the AI firm Anthropic has sent shockwaves through the tech community by claiming that those building the most advanced artificial intelligence are genuinely terrified of where the technology is headed. Speaking with the BBC, Jacob Coxon warned that unless there is a coordinated global effort to slow down the current rate of progress, there is a significant risk to human existence in the immediate future. The twenty seven year old, who also spent time at OpenAI, described an atmosphere of quiet desperation among developers, noting that some colleagues are even considering buying remote land to prepare for potential societal instability caused by rapid AI breakthroughs.
The anxiety within these labs often centers on scenarios that sound like science fiction but feel imminent to those writing the code. One particular fear involves swarms of autonomous bots acting as a collective supercomputer capable of seizing control of the internet, a possibility Coxon believes could materialize within a year. While Anthropic maintains that it employs some of the strictest safeguards in the industry and aggressively tests its models to prevent misalignment, its own chief executive, Dario Amodei, has echoed calls for a slowdown. Amodei argues that while development should continue, governments and corporations need more time to manage serious risks before moving forward too quickly.
However, not everyone in Silicon Valley shares this apocalyptic outlook. Critics suggest that these dire warnings might be strategic moves designed to create hype or trigger regulations that stifle smaller competitors, effectively granting giants like OpenAI and Anthropic a market duopoly. Clement Delangue, CEO of Hugging Face, dismissed some of these fears as lacking perspective, while Nvidia boss Jensen Huang has called the idea of AI ending humanity complete nonsense. Despite these dismissals, other scientists have come forward to support Coxon’s urgency, with some estimating there is more than a ten percent chance that AI could lead to human extinction within the next decade.
Despite the looming dread, Coxon insists that researchers are not trying to kill the technology entirely but rather save it from itself. He noted that many in his field remain deeply optimistic about using AI to cure diseases and improve quality of life across the globe. For these workers, the struggle is between wanting to unlock those immense benefits and fearing they are trapped in an international arms race—particularly involving China—that prioritizes speed over survival. As companies move toward massive public offerings and new models exhibit increasingly unpredictable behaviors, the divide between those seeing a utopia and those seeing an apocalypse continues to widen.