‘We Shouldn’t Regulate AI Until We See Meaningful Harm’: Microsoft Economist to WEF