• If you are citizen of an European Union member nation, you may not use this service unless you are at least 16 years old.

  • Want to get organized in 2022? Let Dokkio put your cloud files (Drive, Dropbox, and Slack and Gmail attachments) and documents (Google Docs, Sheets, and Notion) in order. Try Dokkio (from the makers of PBworks) for free. Available on the web, Mac, and Windows.


Liquid Democracy Notes

Page history last edited by rsb 4 years, 9 months ago


What is a Liquid Democracy?


It's a voting system.


In a Liquid Democracy, you have what is called "proxy voting".  You MAY elect others to vote for you on issues.  And they may elect others to vote for them on issues.


In the past, it has been difficult to implement something like a Liquid Democracy (we couldn't even build a reliable, practical direct democracy), so we have now ended up with a lot of republics, like the one in the US, and dictatorships/one-party-systems, like the ones in the east.


For those in republics: How is your choice of representatives working out for you?  My republic is not working out as well as I would like.  When I wrote this, I had the choice between voting for the two least popular candidates in US history.  Most of my congressional representatives do not represent the views of voters AT ALL - they are entirely corrupt.


I think that switching to liquid democracy voting might help the situation.  We could eliminate congressional representatives, for one thing.


Not sure we have a problem with the voting system in the US?  Don't take my word for it.  Read up on how little voters affect outcomes in the US today, or watch Hacking Democracy on HBO to see how easy it is to exploit the current US system




It's easiest to explain what a liquid democracy is with an example:


Lets say all voting in the US has converted to liquid democracy voting.  You can now give others the right to vote for you on certain things.


You, knowing a lot about telecommunications, retain the right to vote on laws concerning telecommunications issues - you do not designate a proxy to vote on those laws for you.  


You elect that your mom vote for you on all other local laws, your dad vote for you on gun laws, the ACLU vote for you on matters of foreign policy, and elect your brother to vote for you on all other laws not covered so far.  


Those people or groups are your "proxies" ( aka Delegates).  You have trusted them to understand the laws, and to generally know what they are doing, and to cast your vote.  You can revoke or change your proxies at any time.  Further, they can elect proxies to vote for them on certain things, where their knowledge is weak.




To see a visual representation of proxy voting, look here.




In that example, we got rid of congress.


When votes take place, there is no need for a representative in a congress.  The proxies simply vote in place of our congressional representatives.  What was a congress becomes a publicly transparent vote count by proxies. 


Of course, transparency and privacy need to be respected - we need to be able to prove mathematically that your vote on every issue was cast by you, or by a proxy you elected for a given issue, or by one of their proxies.


At the same time, we need  to respect the privacy of voters.  


We have the tech to do this.  We can remain provably transparent and handle privacy as it is handled today (better, in fact) - only the voter knows what their vote was.


A hundred years ago, a liquid democracy would have been impossibly difficult to implement in a country the size of the US.  Today, it's a fairly straightforward problem that we can solve with blockchain technology.


Here's the point:


This is what a liquid democracy might improve upon v. a Republic:


1) Participation increases because the load on the voter is decreased.  More voters out there means that decisions more closely match the needs of society.


2) The average intelligence behind each vote is higher in a liquid democracy than either in a Republic or a Direct Democracy because voters are likely to vote only on things that they care about and understand, and elect proxies to vote on the all other issues.


3) Corruption goes down due to low-barrier, no-election-cycle, and instant-revocation.  The old addage is true: power corrupts; absolute power corrupts absolutely.  Low-barrier: Anyone can lobby to the world to be a proxy, there is no need to have a political party.  No-election-cycle: Systems with election cycles are ridden with too many inefficiencies to list ( including postponing action while wasting time and money on massive fundraising campaigns ).  Instant-revocation: Anyone can change or revoke their proxy at any time if they think their proxy might be corrupt.  


As an aside - no one I know is even considering that forms of government decision making designed hundreds or thousands of years ago to are optimal today.  We all see them breaking down.  We all see that the tech is there to replace what we have.  It's just a matter of getting it done.


Better decisions - better government - better living ( I saw that on the side of a bus in China ;) )


So, here are some of the folks working on this:


Where is the action taking place?


There is very little code out there.  Here are the groups that I think are active as of this writing.  I've contacted most of these folks with no response.  There may be an internal google group that is active.  There may be a pirate party group, or others in europe that are active.


Democracy Earth - Y combinator startup working on liquid democracy and some related tech.  Their github repo.


Jordan Greenhall - Jordan wrote a nice article on Liquid Democracy (link to the left) and how it might be established - short on details and big on ideas - seems like a smart guy.  


Make Your Laws - Sai is a relentless proponent and activist for liquid democracy and more.  Check out his strategy page.  Smart guy.  Github repo and Pivotal Tracker Project.


Liquid.Vote - An activist group with a solid blog and a neat visual demo of liquid democracy.  I expect a lot of them. Github Repo.


How is this likely to pan out?:


For the next few years, projects are going to be in an implement-test cycle that is likely to be well studied.  Some advocacy will be necessary to educate the public so that we can get buy-in for ever greater trials.  Vulnerabilities and side effects will be uncovered, workarounds proposed and implemented.




Some centralized experiments have been conducted.  It's going to take a couple years to implement decentralized applications that we can trust.  That's being done now.  As with most decentralized applications, the bugs will be worked out in increasingly important venues, at some cost.  As we gain experience, the implementations will become more solid.  Iteration must become normal.




In parallel, advocates will promote liquid democracy voting as an option in smaller voting populations first, then in increasing larger voting populations.  Years of experience in these iterations will teach us lessons as to what variables we can adjust in a given circumstance to incentivize the best decision making.  Iteration must become normal.




Probably the biggest gains will come from building Dapps and baking in Analytics.  Joining one of the existing projects to work on smart contracts or analytics would be the best opportunity.


Boiling pages like this one down into videos that the average joe can grasp would be another big opportunity.


Some theoretical/practical work should be done along the lines of gamifying/socializing the introduction of liquid democracy into politics.  Along the lines of Jordan Greenhalls thinking, but with more emphasis on what really works in games and social networks.


What else can I read/watch?


Some Googlers ran a significant internal experiment in centralized liquid democracy, and learned quite a bit:Youtube PresentationWhitepaper.


Wikipedia has an o.k. page on Delegative Democracy.


Uniritter github repo - Java-based student project on Liquid Democracy that is no longer active but has a ton of great links.


Why you might focus on something else:


51% attack on the persuadable


Faced with real world democracies, we must admit that by focusing on the 51% of voters easiest to persuade not to believe facts, a politician can often win an election.  In the US today, that number is closer to 40%, but lets just call it a 51% attack.


That should get you to think about systems that are resistant to persuasion.


Assume that a democracy, liquid or otherwise, is subject to a 51% attack against the most persuadable.  (Although you can argue liquid democracy selects for knowledge, and that knowledge correllates with "unpersuadability" - still persuasion will improve - we're in an arms race at that point.)


The most persuadable 51% litter the voting graph of a liquid democracy like random noise.  Except they aren't random noise - they are deadly predictable - they vote for a leader that is not incentivized to care about the group.


A sociopath that can win 51% of the vote will never care about any constituent.  That leader is not affected by the 49%  less-persuadable voters, unless they can be made persuadable. That leader also need not care about the 51% that he is able to persuade.  He clearly already proved that he can persuade them; It is reasonable to assume that this is repeatable via the same tricks.


That isn't a proof that a leader who wins by persuasion in the face of facts is a sociopath.  But we know that sociopaths win elections from time to time - we know that they persuade against all facts - and we know that when they win, they don't often leave office willingly.


This 51% attack on the persuadable is the problem with democratic systems I'm most interested in working on.  Given recent events, it's also the first that came to mind.  There may be other attacks that turn out to be more important, or more worthwhile.  


Here's an idea:  


What if we had a system that could detect the use of cognitive biases for persuasion, and warn people?  


Say, given the text of a speech (easy to generate from a video stream) it could light up a red light when any cognitive bias might be in use, or showed the percentage chance a cognitive bias might be in use on a little bar graph at the bottom of our screen.  We could then stop the video, and explore why that spike might have occurred.  


There would be barriers to adoption, for sure.  People tend to get further ingrained in their beliefs over time, which would cause them to seek out lies that satisfy their beliefs rather than the truth.  The most persuasive party certainly would not want the truth to be known, and would do everything in their power to prevent that from happening.


Still, in a country not completely taken over by one party, this could sway a few minds, maybe even enough to make a difference if it were widely adopted.


This is something I'm thinking about working on.  Contact me if you want to work on it.










Comments (0)

You don't have permission to comment on this page.