Experts warn AI safety debate misses practical threats

WIRED surveyed a panel of AI experts to separate hyperbolic claims of mass extinction from practical threats like nuclear risks, cybersecurity vulnerabilities, and biological attacks. While most experts dismissed the likelihood of sudden human extinction or rogue sentient AI, several expressed deep concern over the reckless integration of error-prone systems into critical infrastructure and the rapid erosion of labor markets. Panelists also debated potential solutions, ranging from tighter government regulation and accountability to corporate pacing and economic resilience policies. Open article →