Can we kill AI before it kills us? That's the idea behind a bipartisan congressional bill called the AI Kill Switch Act, which would give the US government authority to shut down rogue AI models before they cause catastrophic harm. Companies like OpenAI are already building "automated shutdown capabilities." Voluntarily. For now. And an Anthropic researcher recently put the odds of AI killing all humans at more than 10%. So yeah. The kill switch probably should've come first. It's like flooring it to 160mph and then wondering if you ever checked the brakes.(Lark Davis)