AI industry debate intensifies over existential risks
The AI sector is facing a renewed public dispute over whether advanced systems could threaten humanity’s survival. The discussion accelerated after researcher Jacob Coxon said he was leaving Anthropic because he believed leading AI companies were taking unacceptable risks. Anthropic’s alignment lead then wrote that the company sincerely believed AI could kill all humans and personally estimated the likelihood at more than 10% within the next decade.
On the Equity podcast, Kirsten Korosec, Sean O’Kane and Anthony Ha examined the claims. Ha questioned who the word “we” represented and said the percentage appeared unsupported, while acknowledging that Coxon had acted on his concerns by ending his work in the field. Korosec suggested that warnings about models breaking through safeguards might also highlight how capable the companies’ systems have become. The participants said genuine concern and commercial incentives could coexist.
The discussion also focused on Anthropic’s expected IPO and whether such statements could affect its S-1 filing and valuation. The speakers noted that the apocalyptic framing may draw attention away from nearer-term issues, including effects on employment, the environment and climate. The episode was recorded before Anthropic CEO Dario Amodei released his plan for more cautious AI development.
Written by Sakhanda Wire AI from reporting by TechCrunch. Facts, quotes and full context are in the original article below.