Principle 1: The Core Game should be simple and accessible.
New users need to be able to come in and vote and post and not be put at a colossal disadvantage by the system design. I spent a lot of time yesterday trying to explain to 
I am really, really good at using the overcomplication of the system to my own advantage and the advantage of the users I want to vote on. I am good at that now, I will be good at that when it changes. I promise you I have a huge edge over users who don't figure out all the intricacies of how voting works, who don't do math and build spreadsheets and snoop the blockchain and run separate accounts just to do experiments.
I think that's a bad thing. That edge I have is a giant curb that new users have to jump to be competitive here. I've encountered many users who've been here far longer than I have who still haven't jumped it, who are still frustrated and confused. We need to allow people to come in here cold and be on an even playing field with the experts, at least in the very simple game of creating, finding, and voting on content.
1. A vote is a vote is a vote.
A vote should always have the same total value no matter where or to whom it is given. A voter shouldn't have to think about how many votes a post has, or how old it is, or who posted it. Their decision should be simple: vote, don't vote, how much. This would allow a brand-new user to understand the basic voting system in seconds rather than months or not at all.
For those who want more human voting, the best way to get that is to make voting operate on a human scale. If you need a spreadsheet to decide where your vote goes, most people aren't going to bother.
2. All information for optimal voting should be easily available.
One of the things that annoys me about the change in HF20 is that in order to vote optimally, we're going to have to snoop the blockchain to find the times of earlier votes on that post. I'm sure before too long someone like 
All information that matters to how much a vote is worth should be within the post's display in the condensers. If 
3. SBD should be fixed.
A lot of the overcomplication in the current system has come from SBD being overvalued by the market while the system continued to value it at $1. There are several options for fixing this; my preference would be to dump SBD and USD entirely, and just run everything in Steem, but absent that I support repegging efforts, and I would support a system change that tied the system's value of SBD to the market value.
Principle 2: New users should have an easy path to effectiveness.
The price drop has made this a much smaller issue than it was when I joined, and you could put $500 USD into the system without seeing much of a difference in your account or your interaction. These days that's still a lot of money but at least it gets you a decent amount of effectiveness.
But as we seek large-scale adoption we need to make sure that we aren't just dropping new users into a system where they can't do anything, and don't have any clear, direct path to being able to do anything.
1. Encourage smaller buyins.
My days in online poker, free-to-play games, and on Patreon suggest that there are some inflection points in people's willingness to put money into a system. For USD thinkers those seem to be at $5, $20, $50, $200, and $500. The smaller numbers especially can be things people do as a weekly or bi-weekly investment.
Even with the price drop, someone who puts in $20/week isn't going to see an immediate, week-by-week benefit of that buyin. Which means even if they try it once they aren't going to get the reward cycle going and become consistent small-time investors. If Steem is going to be successful long-term we need to be more attractive to those people.
2. Encourage powering up.
Some of us can see the long-term value in powering up our account to increase our effectiveness here, but that's a sophisticated attitude, and one that requires a certain amount of real-life privilege. We need to get more authors powering up their rewards rather than cashing out, in order to take roles that include voting, community support, and participating in decisions about the future of Steem.
3. Create more structures of consistent support for excellent new users.
Once someone is here, and producing good content, and interacting, and looking like they could be a long-term asset, we need to treat them like a long-term asset, and make sure we offer them enough rewards, social and financial, to keep them here. We especially need to be training more people into leadership positions, and making the people who build leadership positions for themselves more effective.
This is the one I'm actually using my stake to build right now. I can only catch a few dozen users at a time, but that's a few dozen more than zero. Hopefully the effect will be exponential as those users grow and offer their support to others.
Principle 3: Rewards should narrow the class divide, not increase it.
The single biggest issue on Steem is the massive gap between rich and poor. The idea that a middle class is necessary to the long-term success of Steem has been gaining a lot of ground lately, especially thanks to the writings of 
1. Support egalitarian rewards over stake-weighted ones.
Rewards that are determined by the power of your stake simply drive the class gap open further. We should direct as many of the rewards as possible to participatory actions available to everyone rather than to stake. In practice, curation rewards are incredibly stake-weighted, while author rewards are more egalitarian, so I support author rewards whenever possible. Hopefully we will see an expansion in the universe of participatory action rewards when SMTs come.
2. Broaden the rewards base rather than narrowing it.
Many of the proposals you may have seen around have the end goal of distributing more rewards to the most popular posts. They tend to be obfuscated in different language. "Superlinear rewards" means distributing more rewards to the most popular posts, and they're at least honest about it. "Returned to the rewards pool" also means distributing more rewards to the most popular posts, and that one isn't as clear.
This is backwards. A healthy system is one in which we have a wide variety of content that gets voting support, and many users who are able to grow through producing that content. Giving more rewards to popular posts distributes power to high-stake, sophisticated, experienced users at the expense of everyone else. It also currently means that the few popular topics we have stay popular, and any topic that doesn't have substantial support has a hard time getting traction.
Principle 4: We should embrace automation.
By being on blockchain at all, we are early adopters of the most-easily-automated system ever invented. We need to treat this as a feature, rather than a bug. Easy automation is one of the most important characteristics of a system based on on open ledger, and fighting it is essentially trying to turn that open-ledger system back into a closed-ledger system.
There may come a time when blockchain-backed social media is the only social media. (That seems unlikely; one feature Steem can't have in its current form is private posting, and people want that.) But until then, what's special about this platform is the very fact of its open ledger and open access, where anyone who wants to can access the back end of the system and write software to interact with that.
This is awesome, and revolutionary. It will allow us to build things that have never been built before, to do things on social media that have never been done before. Like any brand-new engineering field, a lot of the things we do might end up not being particularly good ideas. But this is not a reason to scrap the concept of automation.
Wrapup
As I said above, these are the principles that guide my actions when participating in the conversations around the future of Steem. I hope that my writing has led you to think about them and embrace some or all of them, but I'm here to convince you, not to force them on you. One of the great things about this platform as it currently stands is that everyone has the opportunity to participate in determining the principles by which it goes forward.
I can argue with people over some of these, while also building structures with my own stake that allow me to take action on some of them without anyone's help or permission. That's not entirely new, but you generally need a lot more resources to do that in a real-life social context. The promise of being able to take meaningful action towards my goals in a simple and effective fashion is what drew me to invest here and what will keep me here long-term.