I disagree that AI Ethics is a bullshit field-- but its one that belongs, has to belong, on the outside looking in. Its an academic field and a watchdog field. Its a subset of Ethics, not of AI. If Google were really concerned about AI ethics they would have set up an external entity with a consortium of other industry partners, seeded them with an endowment large enough to have some measure of intellectual freedom, put them under the auspices of an external parent organization with a long and positive reputation and severed all ties with them. They wanted the optics of ethics, the new car smell-- but they didnt wan't to breech the membrane of plausible deniability when someone internal pointed out that their golden geese were causing some problems.