BrainFrog comments on AI alignment: Would a lazy self-preservation instinct be sufficient?