Some thoughts on decentralizing Steem in the long run
This might be a good time to open a discussion of the future of Steem. And I don't mean the near future but the long term.
Currently Steem is not very decentralized and that's OK. When blockchain projects are young, too much decentralization makes their development hard. A lot of time will be wasted for unnecessary debating over minor things. Steemit Inc having a big stake is also good, because they can do the development for a while without having to beg money from investors or the blockchain.
Let's make Steem a real DAO (Decentralized/Distributed Autonomous Organization)
I'm not fully happy with the current governance system of Steem. When the blockchain was launched, I was actually a little bit disappointed because I saw it as a downgrade from Bitshares 2.0, which is still the state of the art of blockchain governance.
In Bitshares there are three different roles:
- Witness: the job is to make blocks. This fits for people who can run servers reliably but don't want to get involved in politics.
- Committee: members vote for blockchain parameter changes like transaction fees. Most of the shareholders don't follow very closely what is happening, so they vote people in committee to do the adjustment of parameters.
- Worker: does anything that shareholders want to be done. Coding, marketing, documentation, etc. A worker publishes a proposal on the blockchain and shareholders vote if they want to get the job done.
Separation of powers is usually a good thing, especially when decentralization is the goal. I really like the Bitshares model because different roles are clearly separated.
When the term witness was invented for Bitshares 2.0, the point was to clearly separate block production from everything else. Their job is only to witness transactions and put them into a block. This makes the job easy and also relatively risk-free from a legal standpoint because blockproducers are just working for the blockchain and have no decision making power.
That's why I think the first change to governance model of Steem should be to make block production to non-political job. Currently witnesses are part of the decision making process, they can influence whether or not a hardfork should be approved.
We should also pay attention to the set of witnesses and not only individual witnesses. From a decentralization perspective we need to have witnesses from all over the world. It might be a good idea to let the stakeholders to decide how many witnesses there should be. My intuition says that we should have a few more than we have now.
In the long term we should also have a worker system where the blockchain pays for people to do stuff that stakeholders want. This means mostly coding new features and fixing bugs. This is not a critical because Steemit Inc has so big stake and they can use it to develop Steem. But we should be aware that it's not going to last forever and make necessary changes to the blockchain well before it runs out.
Here is a thought that I'll let you ponder: Should we separate voting powers over reward distribution and blockchain governance?
Somebody who is good at evaluating blogposts is not necessarily qualified to decide who should be hired to produce blocks for the blockchain. In the future, hopefully, governance system will also include voting for blockchain parameters and workers. Then it will be even bigger job to keep up with everything. An average blogger shouldn't be forced to take part in the governance – which means that they shouldn't have power over things they are not interested in. Voter apathy is dangerous for a DAO.
I haven't thought this much yet, but probably the solution would be to create two kinds of Steem Power: one for rewarding writers and other for blockchain governance.
Keeping the KISS-principle on mind, at first this might seem to be a change for more complex system. But actually it's the opposite, it makes life easier for most users. Giving all powers to everybody makes everybody's job harder. By separating powers we can let different individuals to focus on one of few things and forget everything else.
This is a quickly written, I'm little bit busy right now but wanted to raise this issue to be discussed. I'll probably write more about the subject in the near future. You'll might want to check also the discussion I had with @smooth yesterday.
Please check out also my older post: How to design efficient and resilient DAO
I think it depends what type of governance. For example if we want to decide between 2 locations for steemfest and make a poll for users to vote on, this is also a form of governance because it directly affects steem but in this case I think everyone should be able to vote.
For decision regarding protocol changes, i think maybe only people with a minimum amount of stake could vote.
How else are users going to be selected?
I agree that governance is very important and Im glad you brought up this topic.
I also like the worker thing, could this be implemented in steem easily?
One simple way would be to let users have two kind of SP. When they power up, they can decide which one they want. Most users would like to use it for writing rewards and only a minority would like to have the governance power.
This might be a way to equalize the rewarding system. Whales have the most power and they have the most skin in the game financially. If they want to take care of their investment, they need to have governance power. That means they will have to transfer the rewarding power to governance power, which makes them less powerful to determine who gets rewarded.
Many in the community have been discussing about removing steem power entirely as its not necessary, creates extra complexity and also lock people's fund . Basically people would be able to vote with the steem in their wallet, but only the steem that has been in the wallet for more than 7 days to prevent double voting. This will allow people to withdraw funds immediately and not have to learn about steem power. Many people agree with this included developers at busy.org, simplicity is key to create a mainstream product
I think governance power could be a good incentives to invest and for minnow/dolphins to be a kind of holy grail to achieve.
The idea of having investors class who are not able to vote has been discussed a lot too, there has been many proposal for this and i think its a good idea and it keeps the influence accessible to users regardless of the price of steem
Perhaps I should have been a little more clear. By DAO governance I mean the things that are functioning on the blockchain level. DAO is basically a collection of smartcontracts.
Things like festivals are not happening in the blockchain level, so they do not need to be decided on the blockchain. But of course it's possible, too. Easiest way would be to set up different workers (with zero pay) and let stakeholders vote which one of them they prefer the most.
It might be a good idea to design the worker system to be easy to use for asking (stake-weighted) opinions of stakeholders. That would be a helpful tool for hardfork debates. Stakeholders could vote on controversial topics.
Well stated. This is the exact reason i gave dan for why i still love bitshares when inwas asked to join ned and him for our own beyondbitcoin "upgrade" for steem.
The value of bts serparations of power is stunning and all too often overlooked.
I dont think it can be for long though....
Remember that people who are unhappy with Steemit at the moment can migrate to Golos, which is at the early adoption stage Steemit was last May (and you can thererfore power up cheaply). See
https://steemit.com/experiment/@teatree/suggestion-for-those-who-are-upset-with-the-experiment
Golos has the same governance model so it must make some changes, too.
Golos doesn't have to do anything.
The beauty of having two separate sites with their own currencies is that they can each choose to go down their own paths. Steemit can go down the "downvoting" path, and the meme/throwaway short posts/cat pictures path, and Golos can be a repository for longer form articles that take some time to write.
The governance model is different thing than the service. A DAO can be compared to a normal company. The governance model correlates to what kind of structure the company has. The service correlates to what kind of products the company is producing.
Good points raised. I'd like to see more posts such as this, from all parties invested in the platform.