Effective dystopian mass-manipulation and monitoring are a real concern and we're closer to it[1] than to super intelligence. But super-intelligence going wrong

Effective dystopian mass-manipulation and monitoring are a real concern and we're closer to it[1] than to super intelligence. But super-intelligence going wrong is almost incomparably worse. So we should very much worry about it as well. [1] I'm not even sure any further big breakthroughs in AI are needed, i.e. just effective utilization of existing architectures probably already suffices.