Microsoft’s Mustafa Suleyman said in an essay that trying to develop AI consciousness would be “catastrophic” for humanity.