What’s behind the AI industry’s latest warnings of doom?

The AI industry is currently embroiled in its loudest debate yet regarding whether its technology poses an existential threat to humanity. The discussion was triggered after AI researcher Jacob Coxon resigned from Anthropic, stating his concern that leading AI companies are "gambling with our lives." This was followed by a post from Anthropic’s alignment lead, who declared, "We really do earnestly believe AI could kill all humans!", adding his personal estimate that the chance is ">10% within the next decade." On a recent episode of TechCrunch’s Equity podcast, Kirsten Korosec, Sean O’Kane, and the host discussed these apocalyptic warnings. The conversation explored skepticism toward "doomer" narratives and questioned whether such warnings are a way of "flexing" to demonstrate the advanced nature of a company's AI models, particularly as these firms prepare to go public. Sean O’Kane raised questions about how these concerns would appear in Anthropic’s S-1 filing for its IPO, wondering if lawyers must officially state that there is a more than 10% chance the company could develop something that would eradicate humanity, which would be materially bad for business. This discourse occurs around the time Anthropic CEO Dario Amodei published his plan for more cautious AI development.