Jensen Huang on AI safety: "We all need to hope it's an engineering problem. If it's not an engineering problem, it's not solvable." very obvious atp that the AI alignment/safety needs more engineering mindset and less throwing doom percentages around.