Dr Roman Yampolskiy, AI safety expert covers the dangers and implications of AI/AGI between now and 2030