United Kingdom • 🌿 Progressive

AI insiders warn of extinction while employers chase valuations

AI insiders warn of extinction while employers chase valuations

Jacob Coxon, 27, resigned from Anthropic warning that AI developers believe the technology could destroy humanity.

Jacob Coxon, a 27-year-old researcher who left OpenAI to join its rival Anthropic, resigned this week with a warning that rattled the Goldman Sachs tech conference in San Francisco: the people building artificial intelligence believe it could destroy humanity, and they are gambling with our lives. --- THE CONTEXT --- Goldman Sachs hosts an annual conference at San Francisco's Palace Hotel where Silicon Valley executives court investors. The BBC reported that Coxon's resignation landed during that event, forcing discussions about existential risk into a setting designed to project growth and financial confidence. --- THE FACTS --- Coxon published his resignation on Tuesday, describing forthcoming AI systems as superhuman systems that can hack anything. He stated explicitly that his warnings were not marketing. Evan Hubinger, a team lead at Anthropic, amplified the alarm on X, writing that he personally estimates more than a 10% probability that AI will kill all humans within the next decade. Anthropic CEO Dario Amodei had already drawn criticism for stating AI could wipe out half of entry-level white-collar jobs and would test who we are as a species. On Saturday, Amodei published an essay calling for a slowdown in AI model development and for global regulation, describing the associated risks as serious. Anthropic was valued at 965 billion dollars, roughly 713 billion pounds, in its most recent fundraising round. The BBC requested a response from Anthropic; this report does not document one. --- THE POSITIONS --- Grindr CEO George Arison told the BBC he viewed the statements as indicative of an anti-civilisational worldview at Anthropic, called them dangerous, and said he had instructed some of his engineers to stop using Anthropic's technology. Arison also framed continued reliance on Anthropic as irresponsible stewardship of shareholder money. Coxon and some current Anthropic employees echoed the same concerns in their own posts. --- WHAT REMAINS UNKNOWN --- This report does not establish whether Anthropic has internal safety procedures addressing the concerns raised by Coxon and Hubinger. The source does not specify the contractual conditions of Coxon's departure or whether other employees are considering similar steps. --- UNANSWERED QUESTIONS --- • What formal internal channels exist at Anthropic for researchers who identify serious safety risks before concluding that resignation is their only option? • How will the workers most exposed to AI-driven displacement — entry-level white-collar employees — have any voice in the global regulatory framework Amodei now endorses? • Does the resignation pattern at Anthropic and OpenAI reflect isolated dissent or a broader structural failure to incorporate safety concerns into product decisions? • What consumer protections would a global AI regulatory regime include, and who would enforce them? --- EPM ANALYSIS --- The human cost dimension is concrete: Amodei himself has named entry-level white-collar workers as the first group facing displacement, and no protective framework is in place. The irony is sharp — the same executives calling for global regulation are the ones accelerating development. Winners so far are the investors entering at a 965-billion-dollar valuation. The decisive variable is whether workers and consumers gain representation before regulation is captured by the industry it is meant to govern. 📌 📌 EPM Take: In EPM's view, the resignation of Coxon and the public statements by Hubinger expose a structural failure, not a communications problem. When a company's own team leads assign a greater than 10% probability to human extinction, and the company continues scaling toward a 965-billion-dollar valuation, the burden falls on regulators and democratic institutions to act — not on individual engineers to absorb that risk alone. The workers Amodei says will lose their jobs have no seat at the table where this technology's future is being decided.
📤 Share on Telegram

¿Te gustó este artículo? Recibe cobertura global en tu correo.

Suscríbete gratis / Subscribe free