Curation Markets Update: 31 August 2017
Curation Markets is an experiment in
- using tokens to curate information &
- exploring ways to mint/spawn these tokenized, curatorial markets (eg continuous token models).
Curation Markets is a broad concept that will ultimately allow more groups to coordinate globally around shared goals.
Many of you want to help, and I’m incredibly thankful for that! Curation Markets *is* about enabling many more communities to work together and earn from the value they co-create. I will thus be doing an update every 4 weeks on what’s happening, what’s transpiring, and what’s next!
Updates!
Not much updates in the past 4 weeks unfortunately. I’ve been heads down working on Ujo-related work & also traveling.
The Autonomous Artist
I finished a draft post and design for an autonomous artist using Curation Markets, but I want to finish up the designs of a new variation of Curation Markets that could be used instead. When this is done, I will release the bounty! In essence, the project will aim to curate software that produces art. This art bot then auctions a new, unique digital art work every week and rewards the curators for choosing the right software that produces this unique artwork.
Funding Public Goods
Fred Ehrsam posted a new, useful post on funding the evolution of blockchains. It really reminded me that I want to play around with Curation Clubs.
It is essentially a Curation Market whose only goal is to help curate cashflow effectively. This would mostly be useful in the context of public goods funding, where people would want to invest in education, but don’t know whom to donate towards.
District0x!
Perhaps the biggest news is that District0x’s Meme Factory that will utilise Curation Market designs are partnering with gifs.com. Exciting to see this eventually in action!
Additionally, another District0x proposal from Luke Duncan, 1Hive, got a spotlight:
I’ve been at work on an iteration on Curation Markets that mostly gets rid of the reserve and more directly rewards curators. It disincentivizes speculators and more directly rewards curators for being good curators. With Luke, we’ve been discussing it in the Gitter room: https://gitter.im/Curation-Markets/Lobby (please join and say hi!).
I’m copy and pasting a part of the discussion here:
Luke Duncan @lkngtn Aug 22 03:35
I saw a note on the whitepaper about splitting the concept into two pieces, continous token minting and the actual mechanism for curation that utilizes those tokens. Is that something that is being pursued?
Based on the stuff I’m working on for 1Hive I think that seperation makes sense — You essentially have 1 variable which is which token minting/pricing funtion to use for “buy in” to a community, and then you have a second variable which is what scheme you use for curation, i.e. token balance voting, quadratic burn voting, etc. Different combinations of those two variables will result in “curation markets” with different properties.
Simon de la Rouviere @simondlr Aug 22 16:58
@lkngtn Yes. I have been exploring that. Just as you say: The token is minted and destroyed in some way (ideally through a continuous model) and then once the token is in circulation, it used for curation/voting/swarming.
…but, I’ve been working on a new iteration that gets rid of the pool/reserve. When minting, the current set of staked coins basically get a “dividend” directly. It’s a way of saying: “This current set of curated items has caused me to want to join.” — Think of it like seeing a set of posts on a sub-reddit that causes you to hit the subscribe button.
It’s not necessarily going to be applied everywhere, since in your case, have the reserve is a useful thing. With the new iteration I wanted to make more sure that you earn by being active participant in the curation and reduce less deadweight, where users buy tokens just to hang around and do nothing. With the newer iteration, token holders are incentivized to stake.
Still busy with this variation. One or 2 design decisions to make still.
Luke Duncan @lkngtn Aug 23 01:31
Thats really interesting. Curious to see how that ends up working, I would worry that distributing a dividend at the time of minting would not scale very well (though I’m not really a developer so perhaps I’m offbase there). I do like the idea of a more direct incentive for curators/community members who are building value, though how do you differentiate between those who are actively contributing versus those who are just holding tokens (speculators)?
Simon de la Rouviere @simondlr Aug 23 04:14
> would not scale very well
Yes, still uncertain how this scales. Perhaps it’s likely that it’s only say the top 100 staked pieces of information. But there might be clever ways to make this work (for all staked information) without too much gas overhead.
The goal is to reduce speculators and have the system more directly reward curation.
> how do you differentiate between those who are actively contributing versus those who are just holding tokens (speculators)?
Minting sends funds only to current stakers. This is divided by the most staked information. eg, the most staked information can be, for example, 10% of all staked tokens, receives %10 of new funds flowing in. If you aren’t staking (just holding), you aren’t getting of the new funds flowing in.
Luke Duncan @lkngtn Aug 23 05:09
> This is divided by the most staked information. eg, the most staked information can be, for example, 10% of all staked tokens, receives %10 of new funds flowing in. If you aren’t staking (just holding), you aren’t getting of the new funds flowing in
Would this potentially encourage people to negligently stake to already popular posts just to be eligible for the reward?
Luke Duncan @lkngtn Aug 23 05:15
It seems like sort of a tragedy of the commons situation, where it makes sense for an individual to simply stake to already popular content to be eligible to the reward. This may have a negative effect on the community as a whole as that behavior would not actually be curating content, fewer people will join, but the cost of that action is spread among the entire community, not the individual speculator who just stakes in order to be eligible for rewards.
Simon de la Rouviere @simondlr Aug 23 15:51
> Would this potentially encourage people to negligently stake to already popular posts just to be eligible for the reward?
Yip. It’s something that’s a problem. Which is why this variation includes another “smaller game”, which is basically likely a tiny prediction market in the stakes themselves. Early stakers to an eventually popular piece of information gets rewarded with more tokens when they leave. Late stakes leaving too late gets burned.
So, it incentivises people to go find other information to stake towards. It’s a trade-off between earning more tokens and earning ETH from minting.
It adds more complexity though, since there is now a new game being played inside the larger game of community curation. But both might offset each other to a nice outcome.
The smaller game (surfacing good content early) is useful in surfacing relevant information. But this game is insular and not necessarily aware of the global game (doing good community curation). When new people enter, it’s a way to reward the smaller games to conform to community curation standards. Additionally, if there’s only global curation, then speculators can earn without participating, since the act of curating is only about getting more people in, which isn’t necessarily optimal. Hence it is own tragedy of the commons: hoping there are at least some people who curate.
What do you think?
Luke Duncan @lkngtn Aug 23 23:53
I’ve explored (in theory) a similar concept in relationship to trying to incentivize early and active debate prior to a vote. I called it a consensus prediction market and a lot like what you are describing. I think it will definitely have a positive effect (though does add complexity), but I don’t think it really addresses the issue of negligent voting at the end. What if instead of providing the dividend to only those who actively stake, you have a part that goes to active participants in the prediction market, and a part that goes to everyone who is in the market regardless of if they have actively staked? You still have deadweight loss, but you don’t encourage people to stake for the sake of voting/rewards.
Simon de la Rouviere @simondlr Aug 24 17:18
Ah hmm. That’s interesting. So, the ETH paid is split between the stakers and non-stakers, with the stakers ideally receiving more of the ETH? Is that a correct assumption? If so, what precentage should the split be? Interesting thinking.
Luke Duncan @lkngtn Aug 25 01:37
I’m not sure about what percentage for the split, though I think its a matter of balancing the following two factors — If you create a continous token issue/market maker for a community token and its sole purpose is to be used in the prediction market game, then I think you lose a lot of the insulation of the community from outside forces since people would just mint more tokens to control the prediction market outcome. By having a portion either remain in a reserve (like the original concept) or otherwise be distributed irrespective of staking then you create buffer between the community and outside influence (since the price of the token would also include speculators rather than only active participants in that market).
Thanks for the discussions! At the moment, I’m still travelling and working on Ujo, so Curation Markets work has slowed down. Don’t have that much extra time atm. It still something I think about frequently. Hopefully we get to test some of these assumptions in the wild, soon!
See you again in a few weeks. :)
