Algon comments on Deceptive AI ≠ Deceptively-aligned AI