The conclusion? Living systems are computational systems

And this should put into greater context the concept of Steemit as a Social Computation Game rather than just a "Social Media" platform. Steemit can actually compute using it's swarm intelligence in the form of collaborative filtering upvoting. In the experimental sense of the word, this ability to compute is what makes Steemit most intriguing.

Steemit needs to implement different forms of human computation

In some of my other posts I discussed with @craig-grant and @ned the topic of reward capping. I still do not support reward capping but to move on from that as now I expect a trial, I will offer my deeper and philosophical suggestion, that Steemit if it is seen as a Social Computation Game rather than a mere Social Media Platform, then the incentives in my opinion should be moved so that the playing styles and roles can be more diverse.

Yes, bloggers have to blog, curators have to curate, but there could be roles for people who want a more stead stream of rewards and I think human computation could fill that void long term. Subjective Proof of Work is only one example of a kind of human computation within a Social Computation Game, and many other examples are possible.

The only issue I have is when rewards are cut in a way which removes entire opportunity paths. In essence it becomes a road block and with nothing to replace it, the game becomes less fun for certain people who feel held back. The better solution in my opinion is to continue to open new opportunity paths when incentives are moved. So if they are going to do this reward cap trial, why not also at the same time activate the referral system, or perhaps something else for people to do who have the time and brain power?

What if Steemit can't currently open more doors or opportunity paths?

This would be a problem but not in the short term. In the short term, Steemit has limited or perhaps no competition. In the long term there is Synereo which is using a composable system design which means they will be able to implement and reward all forms of human computation from launch. The point is, for certain kinds of tasks you can have a fixed rate per task, for certain kinds it's subjective, for certain kinds it could be based on time, and a variety of these microtasks is the only way to satisfy everyone.

If we look at games we see there isn't just one way to play. There are usually different character types and playing styles. Depending on the character type you pick, it opens up different opportunity paths. The point is it doesn't get boring, and in an RPG some players might never hunt monsters and simply do blacksmithing or crafting, and still have fun. The variety of activities is what keeps it from being a situation where developers have to choose to take away fun from one group to give it to another group.

References
http://livingcomputation.com/
http://www.cs.unm.edu/~ackley/papers/ackley-aaai16-smpt.pdf