Meta's chief artificial intelligence officer told the Financial Times that concerns about existential risks posed by the technology were "premature."