Why AI extinction fears are overblown, experts say

Experts say AI is unlikely to cause human extinction because artificial viruses face physical constraints, human oversight, and stricter testing, despite recent warnings from researchers. The debate intensified after researcher Jacob Coxon left Anthropic, warning that OpenAI and Anthropic are "betting our lives" on superintelligence. Evan Hubinger estimated a 10% chance AI kills everyone within a decade, but others argue digital capabilities don't translate to real-world harm. Google engineer Anselm Levskaya noted creating a virus requires lab assembly and testing, not just design. Experts like Eric Xing and Jürgen Schmidhuber distinguish AI as a tool from autonomous threats, though they caution against dismissing more immediate risks like cyberattacks.