The pricing problem that nobody asked about: Rethinking Steem curation
Curator: a person who oversees or manages a place (such as a museum or zoo) that offers exhibits.
On the Steem blockchain, what are the exhibits?
I guess the original vision was that the /trending page was the exhibit - the place to be curated - and the aggregation of all voting is a curation process. Implicitly, this vision called for large stakeholders to follow behind the small voters and shape the look of /trending. In this process, the larger stakeholder isn't just shaping the look of the page, but they're also training the smaller stakeholders on what to vote for.
Later, that vision was extended and decentralized so that each community could offer its own dedicated /trending page as an exhibit.
In a museum or zoo, visitors are charged for admission, the curator draws a salary, and the people who supply the materials for the exhibits are compensated. Depending on the financial structure, investors might also receive returns on their investments. Our "admission fee" is the purchase of Steem or SP, and I think the idea was for Steem to reproduce the curation incentives with author rewards for content creators and curation rewards for voters. An attractive /trending page means that visitors would be willing to pay a higher admission fee.
After more than 9 years, I think we can safely say that this vision of "curation" has failed. This was probably unavoidable without a change in thinking. What did the designers and early adopters get wrong?
What is the right mix for authors and curators? One of the more contentious debates while I've been here was the shift from 25% to 50% for curators. I even believe that curators were receiving 75% for a short while before I started here in July of 2016. Do we have any evidence at all to suggest that one of those values is better than another?
Here's a more basic question that nobody asks. Is it even possible to select one right number? It's a pricing problem. Is it likely, or even plausible, to believe that one unchanging price is going to be an optimal value for all posts by all people, everywhere, and over an extended period of time? When we structure curation percentages for "an average user", we're structuring them for nobody.
Probably as early as 2016 or 2017, I was already advocating for a system where each author could set their curation percentage at posting time. This would let new authors give away more rewards to voters as a competitive mechanism in order to establish themselves, and it would let established authors hold on to more rewards after they had proven themselves and built their audience.
If I remember right, Golos (a Russian Steem fork) even tried this, but they screwed up the design so that authors could change the percentage after posting time and before payout time. As a result, authors used this parameter for a sort of a "bait and switch" to harvest votes, and it didn't work out. (I'm not sure what ever happened to Golos, but it seems to have mostly disappeared.)
Eventually, I gave up on advocating for that concept because it didn't seem to be getting traction, and I switched to thinking about beneficiary settings, instead, as a flexible competitive tool for authors.
Iteration 1 in that line of thinking was the @null beneficiary setting. If I'm a new author, I can burn some of my rewards in order to attract attention and start building an audience. After I've established an audience, I don't need to burn as much (or any). I think this worked - to a degree - but there are just too many investors who would prefer to capture the author rewards for themselves instead of using them to honestly curate an attractive /trending page. Burning rewards doesn't give legitimate authors enough of a boost to compete against the gains that those investors are seeing.
Others have come up with another construct where community members set beneficiaries for community curators, and the curators try to shape the community /trending page. In my opinion, that has met with marginal success. It can work, but in the long run it makes a "rug pull" too easy. The best way that this model could potentially be trusted by community members is through the use of multikey wallets for the community curation account. It doesn't solve the problem, however, that large investors can still get better returns if they also harvest some of the author rewards through low-effort posting.
Another possibility that hasn't been tried (AFAIK) is a blogging team where multiple people can use a single account in order to take turns posting on a shared topic and build an audience while sharing rewards across the team through the use of beneficiary settings. However, this would require significant levels of trust, organization, and leadership.
So, to date, none of those known approaches has fixed Steem's fundamental curation problem. The /trending page sucks. IMO, most of Steem's other problems with attracting and retaining authors are downstream from this one - or this is the most visible symptom of the underlying problem.
If we're going to fix Steem's curation, we need a set of incentives that is flexible but sustainable and realigns the interests of authors with the interests of investors. Enter Thoth.
The curated "exhibit" becomes the post, itself - or the curator account's post feed. The automated curator goes out and finds deserving content, the AI writes about the articles and sets beneficiary rewards to compensate the original authors, and votes on the post. Even though it's posting, it sets beneficiaries for other accounts to receive the author rewards and collects curation rewards for itself.
The curator is doing the entire curation process, not only deciding what to include, but also deciding on how to display it. Beneficiary settings also let delegators collect from the author rewards pool by supporting these authors, instead of competing with them. If desired, some rewards can be burned, and the curation account can even collect author rewards, too. This - I think - is the maximum flexibility that's possible for dynamic pricing without a hardfork.
For the record: I still think it would be good to allow authors to set their own curation reward percentage at posting time, but absent that...
When you turn everything upside down and look at curation as a self-contained project, all that's left is a pricing problem, a search problem, and a presentation problem. Unlike the intractable challenge that we've seen on Steem's /trending page for the last 9+ years, pricing, search, and presentation are problems that can be optimized.
If projects like this one take the place of the existing voting services, this might also improve the quality of /trending and have other wide-ranging impacts across the chain.
I'm not saying that everyone should run Thoth, but I am saying that delegation bot operators need to start thinking differently about curation rewards and beneficiary settings. I'm also saying that the clients of voting services should stop tolerating the stagnant technology and the destruction of wealth that the dominant bots are currently offering.
At small scale, @thoth.test has already proved that investors don't need to choose between ROI and wealth preservation. There's no reason to think that it would be any different if a similar project were backed by larger stakeholders.
It's not 2020 any more. Voting service clients should demand an immediate roadmap from their providers that adds wealth preservation and passive rewards to the product offering.
You can view my projects, here.