Berlin, Machiavelli and AI Safety - the case for moderation

Berlin’s reading of Machiavelli demonstrates that there can be no Utopia; so why do AI safety researchers insist on trying to find one?