Ex-Anthropic researcher, who said 'AI could kill us all', calls for 'kill switch'

Former Anthropic researcher Jacob Coxon, who sparked debate after saying AI "could kill us all by the end of the decade", has called for developing a "kill switch". "A kill switch probably would work on a lot of AIs for now," he said. However, he added that it might not work if AI agents launch an internet-wide hacking campaign.