AI researchers say industry is 'gambling with our lives'

Senior AI safety researcher Paul Christiano, who co-authored foundational alignment work with Anthropic's Dario Amodei, now warns of "meaningful risk" of catastrophic and irreversible loss of human control over AI "in the very near term," stating "most people could die." Geoffrey Hinton, one of AI's three founding "Godfathers," separately told the BBC he estimates a 10% probability of human extinction from AI, a figure he called "not unreasonable." Christiano announced he is joining OpenAI's nonprofit safety team in response to these concerns. The warnings reflect a notable shift among researchers who previously expected AI development to proceed gradually enough to allow for adequate safety measures.