During the annual Goldman Sachs technology conference at San Francisco’s Palace Hotel, prominent Silicon Valley figures pushed back against alarming claims made by artificial intelligence insiders regarding existential threats posed by the technology. The skepticism emerged after Jacob Coxon, a 27-year-old researcher formerly at OpenAI and now at Anthropic, stated that building superhuman AI systems amounts to gambling with human survival due to their potential to hack any infrastructure.
Coxon’s comments echoed earlier posts by Anthropic team lead Evan Hubinger, who estimated a greater than 10% probability that AI could kill all humans within the next decade. However, many investors and executives present at the September 12 event viewed these warnings with suspicion, suggesting they might be strategic maneuvers to generate hype ahead of anticipated record-setting initial public offerings by Anthropic and OpenAI.
George Arison, CEO of Grindr, described the rhetoric emanating from Anthropic as an “anti-civilisational worldview” and labeled it dangerous. He revealed that he had instructed his engineering team to cease using Anthropic’s technology, arguing that it would be irresponsible to rely on a business making such public statements. Arison speculated that the company might be exaggerating risks to justify its $965bn valuation by claiming its AI will dominate every industry and job.
Nvidia CEO Jensen Huang also dismissed Coxon’s assertions during the conference, with attendees confirming he called the extinction fears untrue. Huang has previously referred to the idea that AI marks the end of humanity as “complete nonsense.” Similarly, Brad Gerstner of Altimeter Capital accused Coxon of “ridiculous hyperbole,” while Hugging Face CEO Clement Delangue compared asking about AI extinction risk to asking an HVAC technician about climate change.
Despite the backlash, Anthropic CEO Dario Amodei released an essay advocating for a slowdown in AI model development and global regulation, acknowledging serious associated risks. This stance came after Anthropic reported discovering threat actors attempting to use its technology for bioweapon development and cyber-espionage. Some investors at the conference warned that such high-profile warnings could accelerate government crackdowns, though the current Trump administration remains largely supportive of unfettered AI development to maintain dominance over China.
Senator Bernie Sanders introduced the Ban Artificial Superintelligence Act, calling for a temporary pause on advanced AI development due to catastrophic potential. In contrast, OpenAI President Greg Brockman described his firm’s relationship with the White House as a “very good partnership.” As both companies eye public listings, with OpenAI valuing itself at $852bn, many investors remain focused on profitability rather than existential concerns.
Jensen Huang called it nonsense, but I’m still wary. Have they fully solved alignment, or are they just winging it?
The timing really does seem suspicious. Both companies are eyeing massive IPOs right when these scary reports surface.
George Arison shutting down Anthropic use is bold. I wonder if other CEOs will follow suit or just stay quiet.
It feels like a classic marketing ploy. If they admit the tech is safe, why would anyone pay premium prices?