First and foremost, I want to take this time to say thank you to everyone involved in continuing to push for the creation of an effective space for free speech. The ability to share one's convictions freely is extremely valuable, essentially priceless, as it represents the ability to participate in human life. It is impossible to understate how important it is to be able to accurately value human contributions to our collective organism by hearing them all.

In this post I hope to open a discussion on moving forward in the blockchain economic incentives underlying what will soon become Hive. While these would apply equally to Steem, my confidence is somewhat low that this project in its current form can be salvaged because of the negative influence of the Steemit, Inc. ninjamined stake and its current controller's obsession with centralization, control, and censorship. At this time, I support the community fork and will see you on the other side at http://hive.blog! (The two chains will diverge approximately one hour after publishing this post, at 2 PM UTC on March 20, 2020.)

The first-layer or parental coin, whether STEEM or HIVE, will function as a store-of-value best if it is constructed to be conservative in its rewards payouts. Second-layer coins in the form of SMTs (smart media tokens) can then be experimental in their approach with regards to inflation rate, total supply, rewards curve, downvotes enabled or not, governance, etc. while the base-level coin acts as a stabilizing influence.

For this reason, I believe lowering inflation to be a very good idea. I am proposing a figure of 6% which is around one-third less than what we have now. Just to re-hash some history — after the initial hyperinflationary period in which the @steemit ninjamined stake was acquired, it was decided to decrease STEEM token inflation to 10% annually, which would further decrease by half a percent each year until bottoming out at 0.5% after 20 years. Making a coin gradually less inflationary over time is a way of encouraging positive long-term price action against fiat currency, though somewhat ineffective in the short-term as most people think on a day-to-day basis.

With the current design of posting rewards, some inflation is necessary to have a rewards pool. One solution would be to have a fixed inflation rate that is not too low (since we want to have a meaningful amount of tokens for rewards distribution) and not too high (so the coins are not simply dumped and fail to retain their value.) I would propose lowering inflation to a fixed 6% as a reasonable reduction of around one-third.

While this 6% is still high in comparison to the most conservative approach, I believe that for STEEM/HIVE to prove its use-case of rewarding content creation and continually expanding a user base it is preferable that the supply increases substantially each year. People subconsciously dislike seeing very tiny quantities of a resource that is highly scarce, even if it makes sound economic sense. It is depressing for most to buy 0.001 of a coin, and inconvenient to deal with decimals. It may be best that the cost of the token appreciates more gradually along with the supply and a growing user base.

Humans are very fickle creatures, and primarily animalistic in how we experience life. Vague abstractions of reward that fail to supply dopamine and other pleasure-inducing neurotransmitters do not satisfy or provide strong enough motivations to act in most cases. What is more, in the question of witness voting, we need robust incentives to combat voter apathy, which is a major game-theory problem of dPOS systems.

Up to now, the "reward" of voting in witnesses is that, ostensibly, whoever is elected to governance will provide best for the stability and prosperity of the system, thus benefitting one personally. I would argue that a more tangible, even if small, incentive is needed. One strategy is to give all voters a tiny percentage of the block-producer rewards of the witnesses voted for based on stake, but this introduces a troubling push towards further centralization, as voters are likely to choose whoever is at the top at the moment to maximize returns.

A second strategy would be to provide voters with an increase of total passive staked token inflation percentage regardless of who they are voting for, as long as they are actively using all their votes. However, the issue here is the lack of incentive to choose your witnesses wisely, which amplifies further the more votes people are given, since it takes time to be informed on which witnesses to vote. People may just vote randomly, or again, for top witnesses to get the increase. I suggest decreasing witness votes to 16 from 30 to begin with, to prevent a single actor from voting in a supermajority.

An ideal system would both reward people for using their votes (regardless of who is voted) and reward people for choosing wisely (which is ultimately subjective.) Witnesses can wear many hats, all of which may benefit the ecosystem — from programming, marketing, providing financial services, curating content, building and running dapps, promotional efforts, and more. It is quite difficult, arguably impossible, to represent these efforts numerically in order to judge which are best fulfilling their roles (in order to then reward voters for choosing them.)

A major problem in building an algorithmic system that proposes to reward individuals for the value of their contributions is that programming languages deal with tangible content while necessarily ignoring underlying context. Thus, we end up with a paradigm that generally values quantity over quality. For example, if a dynamic algorithm was introduced to rank witnesses according to their activities — it may reward the largest number of code-contributions, the largest number of comments, maximum payouts, largest presence on social media sites — all things which are numerical and and will be inevitably gamed.

The concept of "proof-of-brain" (which I believe ought to be called more accurately "proof-of-conciousness" or "proof-of-awareness") recognizes that there is an indefinite quality of sentience that is needed to best assess value of human contributions. One recognizes intuitively a well-composed blog post, for instance, which cannot be reduced to a simple measure of word-count, pixel distribution on images included, etc. by a computer. It is that which is intangible that provides value, which evolves over time as our species continues its evolution.

I propose reducing the power-down period by slightly over half to 6 weeks from 13 weeks. If you detect a pattern in some of the numbers I am suggesting — 6% inflation, 6 week powerdown, 16 witness votes — you would be correct. However, it is not just arbitrary; it happens to be both easy to remember, conjuring up images of a hexagonal "hive," and also logically based. With how the blockchain currently operates, it is best that powerdowns take quite a long time compared to the standard of other chains (a few days.)

It is possible to move away from needing as long of a period in the future based on community feedback and the continual evolution of the chain. Why go to 4 weeks instantly from 13? Cut it in around half and discern the effect on the ecosystem; move gradually rather than so drastically, while still reducing how much of an outlier this long period is compared to other investments. In truth, 6 may also be too quick of a modification, with 7 or 8 weeks being preferable for the first reduction.

It is also possible to implement a burning mechanism that allows faster powerdowns, but at a % loss of tokens to be either sent to @null or to the DAO. Still a third idea is to allow the internal trading of SP on the open market (where the purchasing party acquires the stake but it remains powered up.) Both of these ideas have rather serious security issues in the case of accounts being compromised. I lean towards the above simple idea of cutting the time period in around half as a first step. This can also be combined with the idea of having a 30-day gradual power-up process (at least in witness voting, maybe also in stake-based content voting?)

Most of the other ideas I currently support for developing the blockchain are front-end issues, such as decreasing the payout amount by half to show author rewards only without curator earnings, increased gamification and intense focus on UX design, built-in encrypted personal messaging on-site, and a very fast, convenient mobile app. The ideas addressed in this post are the main, concrete changes I can see on the blockchain level benefiting the ecosystem at the moment. What have I missed? What was inaccurate — please feel free to correct me. Do you agree or disagree with some of the potential changes listed? Let me know your ideas on modifying blockchain economics below.

Another thing I would like to see improve is increased incentives to engage in meaningful discussion in comments. Currently, many are demotivated to comment due to the "newsteem" rewards curve making voting on comments in general less lucrative. Engagement on posts, on average, is abysmally low. Engagement is a component of virality, a metric we want to be as high as possible in order for the platform to have maximum influencing ability, akin to Twitter. After brainstorming quite a lot on that, I am still stumped as to how to encourage more engagement through code alone. Let me know if you have any proposals for that. Thanks for reading and have a wonderful day!