Could AI really wipe out humanity? Six experts weigh in on the risks

Recent claims by AI industry leaders regarding existential risks, internet-wide botnet threats, and the potential for AI to surpass nuclear weaponry have sparked intense debate. Anthropic CEO Dario Amodei warned of potential internet takeovers, while other experts, including those from the AI safety community, remain skeptical, labeling such predictions as unscientific or alarmist. The discourse highlights a divide between those advocating for immediate, strict government regulation and critics who view these calls as a 'regulatory capture' strategy designed to stifle competition. Meanwhile, geopolitical tensions between the US and China continue to influence the narrative, with some arguing that slowing down AI development could result in a strategic disadvantage. As the industry grapples with the balance between innovation and safety, the debate remains polarized, with no consensus on the actual probability of catastrophic outcomes or the best path forward for global AI governance.