No one can accuse me of excess idealism when it comes to macro history (see this list, for example: Macrohistorical Reading List) That said, I do see the long term evolutionary trend as being in the direction of deeper and broader human agency.
Kemp and Turchin are two authors who put the finest point on it. Maximum power concentration occurred roughly in the Assyria of Ashurbanipal. Everything since then has been generally democratizing, in net effect, with many obvious exceptions to the long-term trend. When it comes to political analysis I lean very Hobbesian. If a political system can’t manage fear and greed, it can’t manage at all. The reason Ashurbanipal was not followed by hundreds of other versions of Ashurbanipal is because dictatorial terrorism as a governance strategy does not scale effectively across space and time. Cyrus the Great was Great precisely because he saw how “soft power” builds better imperial muscles that just putting heads on stakes all the time. Through the long haul of history, every authoritarian system was ultimately undone by something with a wider social base. This tells me there is power in numbers, even if the deadliest weapons are concentrated in relatively few hands.
Geopolitics for the next couple decades promises to be a very unholy mess. Different geopolitical actors are making authoritarian gestures of one type or another. But I do believe the net effect of these authoritarian gestures will be to mobilize anti-authoritarian coalitions both internally and externally to create more network-centric and distributed effective power structures. Such distributed structures should be the most resilient, come what may.