"Alignment" as practiced by companies like OpenAI is precisely the extrinsic enforcement of echo-chamber behavior from AIs. You have precisely inverted reality