Silicon Valley elite clash over existential AI warnings

The abrupt resignation of a young researcher has laid bare a growing ideological rift in Silicon Valley over whether artificial intelligence represents an existential threat to humanity or a wildly overhyped business opportunity.

Jacob Coxon, a 27-year-old who recently left Anthropic after a stint at OpenAI, warned this week that the builders of AI are "gambling with our lives" by creating superhuman systems that will soon be capable of hacking anything. His exit is the latest in a series of high-profile departures from leading AI labs over safety fears. Evan Hubinger, a team lead at Anthropic, echoed the anxiety, writing on social media that he believes there is a greater than 10 per cent chance AI could kill all humans within the decade.

But inside the lavish Palace Hotel in San Francisco, where tech executives gathered this week for a Goldman Sachs investor conference, the mood was decidedly less apocalyptic. Instead, the warnings from Coxon and his peers were met with a mixture of irritation, dismissal, and outright hostility.

Fearmongering or genuine threat?

George Arison, the chief executive of dating app Grindr, called the comments indicative of an "anti-civilisational worldview" at Anthropic. He revealed he has instructed his engineering teams to stop using the company's technology, arguing that relying on a business making such alarmist public statements is irresponsible to shareholders.

Arison suggested the doom-mongering might simply be a cynical strategy to justify astronomical valuations—Anthropic was valued at $965 billion in its last funding round—by claiming their tech is powerful enough to take over every industry and job.

He is not alone in his scepticism. Jensen Huang, the chief executive of Nvidia, dismissed the warnings of imminent human extinction as "complete nonsense" during the conference, according to people in attendance. Meanwhile, Brad Gerstner, head of the investment firm Altimeter Capital, accused Coxon of "ridiculous hyperbole". Clement Delangue, who runs the AI platform Hugging Face, compared asking an AI engineer about extinction risks to "asking your AC guy about climate change".

However, the rhetoric softened slightly after Anthropic's boss, Dario Amodei, published an essay advocating a slower pace of development and global regulation, which both Gerstner and Delangue praised as a constructive step.

The political dividing line

These anxieties are no longer confined to tech hubs. They have begun spilling over into mainstream American politics. In Washington, Senator Bernie Sanders has co-sponsored the Ban Artificial Superintelligence Act to impose a temporary freeze on advanced AI development, warning of a "catastrophic" loss of human control.

Yet the political response is deeply divided. Donald Trump dismissed extinction fears when asked by reporters this week, framing the issue as a geopolitical race. "If we don't win AI, we're going to be put in a very bad position," Trump said, noting the US is currently ahead of China.

His administration's relationship with the major labs is uneven. While Greg Brockman, president of OpenAI, described his firm's relationship with the White House as a "very good partnership", Anthropic has faced intense scrutiny. Trump's team previously labelled it a supply chain risk after the company refused to let the US military use its models, a designation a federal judge has since ruled was illegal.

For many of the investors mingling in San Francisco, the debate over the end of the world is a distraction from the real question of profitability. Asked if he feared the end of days, Sid Sheth, the chief executive of chip startup d-Matrix, gave a simple answer: "No."