If the software is supposed to interact with people, but does not take human psychology seriously, then the problem is not in laziness, ignorance, collusion or greed. The problem is in the software. It is best studied in game development. When the game offers many ways to achieve goal and the most efficient is also boring, then players will overwhelmingly choose that path and complain about it. The same situation is with rewards on Hive and efficiency of bots.
There are two reward curves. First is how posts compete with each other for reward pool, second determines how curators compete for their share in post payout. First curve was changed to linear in HF19, then to convergent linear in HF21 and then in Hive back to linear in HF25. When it is not linear, it can be set to make popular posts get relatively more rewards (bots will vote for posts from authors historically most popular) or relatively less. Bots can adapt to the curve, because they can be made to observe all posts. Humans can't do that. Therefore nonlinear curve promotes use of bots over manual organic curation. Similar thing with curation reward curve - it was changed to square root in HF19, convergent square root in HF21 and linear in HF25. There is however something else that influences curations - voting windows. Currently it is fairly easy for normal people to achieve the same results as mindless bots (some people would prefer if the first window was wider though). More than that, people have higher chance to know if the topic will be controversial to attract downvotes - maybe AI bots will be able to do that in the future. Manual curators also have one more advantage. If they build considerable group of followers, by attracting votes through reblogs and promotional posts, they can pull late voters to the posts they voted on. Late voters are the source of extra rewards for early curators (they increase value of the post linearly, but their share in curation is slashed). I can't see bots doing that.
And "lazy, greedy" whales have very easy way to get consistent curation rewards - vote for hbd.funder posts (rewards going to support hbd-stabilizer). You don't need to be a whale to use that method and I also use it when I have too much voting mana before I have to leave for sleep or work. These posts receive hefty rewards day in day out. So, use of bots is not the most efficient in current environment - if you don't feel confident in your ability to curate, it makes more sense to delegate to team of manual curators than to use bot. It makes healthier environment than it was before.
This explains a lot.
It's a software that let's people interact with other people, though.
If your design is meant for multiplayer and large maps and the map isn't populated enough, then it won't work. Ever tried SQUAD with only a handful of players on the map?
My complaint was always that there were too few players (whales) so the game-theory regulating mechanics, that could have balanced it, never kicked in. Instead of letting others join the game, they were playing in an almost empty world and then acted like it was a design flaw.
It also always felt like a flaw (and I have talked about that extensively) that you need to vote to get better ROI. There should be a reward for not voting at all. Voting for
hbd.funderwould fix that, but I feel like there is a solution out there that's much more elegant and doesn't involve that hack.I'll have to think a bit more about rewards...
rewards over r_shares are strictly linear.
r_shares aren't, because the pool...
I have an answer, but I can't really form a cohesive text, yet.
I would be interested with what you might come up with.
Yes, an empty landscape is bad for a game, that's right.