A new LessWrong essay argues for pausing AI development immediately, regardless of whether a long-term recovery plan exists. The author uses a house-fire analogy to prioritize stopping immediate risks over planning future resumes. This perspective pushes AI policy toward urgent interventionism, suggesting that survival must precede the strategic design of a post-pause era.