Warnings from current and former Anthropic researchers about artificial intelligence potentially threatening humanity are drawing skepticism from some Silicon Valley executives and investors. The debate intensified after researcher Jacob Coxon resigned from Anthropic and accused leading AI companies of “gambling with our lives” as they develop increasingly capable systems.
Coxon, 27, previously worked at OpenAI before joining Anthropic. He said people building advanced AI genuinely believe future systems could threaten humanity, including through capabilities powerful enough to hack computer systems.
Anthropic Alignment Science lead Evan Hubinger publicly supported Coxon’s concerns. He wrote that he personally estimates a greater than 10% chance that AI could kill all humans within the next decade.
Some Executives Question the Warnings
The comments became a topic at Goldman Sachs’ annual technology conference in San Francisco, where executives were already discussing growth and investment opportunities. Grindr CEO George Arison told the BBC that he viewed the warnings as evidence of what he called an “anti-civilisational worldview at Anthropic.”
Arison said the comments prompted him to tell some Grindr engineers to stop using Anthropic technology. He questioned whether the warnings reflected genuine beliefs or could also help AI companies justify extremely high valuations by presenting their technology as capable of transforming most industries and jobs.
Anthropic was valued at $965 billion in its most recent fundraising round earlier this year and is reportedly preparing for a possible public listing. OpenAI is also expected eventually to pursue an IPO, adding scrutiny to how the companies describe both the capabilities and risks of their systems.
Criticism has also focused on Anthropic CEO Dario Amodei, who has warned that AI could eliminate large numbers of entry-level white-collar jobs and create risks serious enough to require stronger safeguards.
Anthropic Continues to Call for Slower Development
Anthropic maintains that advanced AI presents both significant benefits and potentially severe risks. In a recent update on its alignment and security work, the company said it supports a lawful and verifiable mechanism for coordinating the pace of frontier AI development.
Amodei expanded on that position in his essay We Must Pace the Frontier, calling for slower development, independent safety evaluation and coordinated regulation. He argued that companies should continue developing AI while giving researchers and governments more time to address increasingly capable systems.
Coxon has explicitly rejected suggestions that his warnings are intended as marketing. His resignation adds to several high-profile departures from Anthropic and OpenAI in recent years involving concerns about AI safety and the pace of development.
Featured image credits: Magnific.com
For more stories like it, click the +Follow button at the top of this page to follow us.