Content Mining Mechanism

This will be updated to filter conversation about content mining once I have more time to get this set up

Apart from the obvious (illegal content), what would be an incentive to vote down content?

Why wouldn’t you want to have as much content on the platform as people want to upload? Cost of infrastructure is one debatable point that comes to mind, but then again it doesn’t really affect me personally. Or does it? Another reason would be the fact that you want to maintain only quality content on the platform. But having “low quality” content isn’t really a problem as long as there are recommender systems in place. Or curation services. It works with Netflix and YouTube.

Apart from the obvious (illegal content), what would be an incentive to vote down content?
No you are right, this is the main reason.

Why wouldn’t you want to have as much content on the platform as people want to upload?
Personally I expect most content to end up on the protocol, it will be up to the UI’s built on top to do the filtering for what they want on their UI. Like say someone wants to open a stock footage storefront they wouldn’t also have feature length movies and vice versa…I don’t see BREAKER having stock footage for sale on the current dapp.

1 Like

In my very personal opinion, this enables the community and all the people in the DAO create a standard so others who come to the platform know and understand the Quality of the content.

i.e I would vote against all Ben Affleck related things and all Trap Music, Reggeaton, but would vote yes to Good Rock and Metal.Vote down on things with no class.* Ok, this is all a joke, but yeah If I don’t think something goes along with the tone of how the content curation is going, then yeah that be a down vote.

Just like making a playlist, you want everything be in sync with what you have and are headed.

Another thing that might be plausible would be content that has an agenda against the main idea/purpose of the community.

Is this centralization resistant ? That is to say, perhaps different weights might be needed to guard against simply having large whale influence dictate all the views/promotional content. The cost much be sufficient to on the one hand, reward content validators, while still providing a counterbalance to outright whale manipulation. I’ve seen issues with voting up/down governance issues with both steemit and EOS, so I’m mentioning all this with that experience/history in mind.

One aspect to consider is not only the “cost” might involve the aspect of time itself. The upvote/downvote for content might want to explore the aspect of time-domain based weights to voting. A user may have an option on how long they intent to “upvote” the content, which would both impact his/her remaining vote strength for other content, but also the “weight” of the upvote on that content itself. If a user is willing to lock-up his/her vote for say 1-year on the sngls platform, this could suggest they are willing to take more risk as a witness to validate the content being worthy of others to consume. A shorter timeframe might suggest that while the content is good, the user doesn’t see (or “feel/emotionally”) tied to the content they are recommending. This is a form of “counterbalance” for perhaps minority users to risk more sngls stake at the cost to promote content for the network as a whole.

Is this centralization resistant ? That is to say, perhaps different weights might be needed to guard against simply having large whale influence dictate all the views/promotional content.

  • The two systems we are looking the most at are dxDAO and Aragon. The tech team has a lot of connections to Gnosis so we are thinking of using dxDAO more heavily at this point since there were the driving forces behind that and we can get inside information. The Halographic consensus mechanism (which is the safeguard against what you were originally asking) has recently been ported over to Aragon. Aragon also has finance apps that come pre-packaged so we wouldn’t need to develop something like that and people are building apps for it all the time.

If a user is willing to lock-up his/her vote for say 1-year on the sngls platform, this could suggest they are willing to take more risk as a witness to validate the content being worthy of others to consume. A shorter timeframe might suggest that while the content is good, the user doesn’t see (or “feel/emotionally”) tied to the content they are recommending. This is a form of “counterbalance” for perhaps minority users to risk more sngls stake at the cost to promote content for the network as a whole.

  • Interesting. I’m not agains this idea. We are working on the roadmap for the development now. I think we are at least a year away from starting dev on the content mining mechanism so the DAO treasury control and parameter settings for the protocol will come first. I’m working on an article that will explain all this for everyone is greater detail specifically how the protocol and the platform work.

I think that should be left up to the platform. You can use the protocol for a lot of different stuff and give it different licenses. For example, stock footage providers are a large business, would the BREAKER platform really need to have stock footage listed or would another platform come a long and build something like that or even possible we build another storefront. What IS interesting is what society will deem acceptable and if people disagree then they can fork the network.

In a nutshell - how would the system you are envisioning work for checks and balances ? I’ve provided some examples of concerns and possible mitigation to deal with issues I’ve seen. I’ve yet to hear of a solution that addresses these problems entirely, from an engineering perspective…open to learning\hearing your thoughts…

In your WP you mentioned “Once the voting is done, no matter if the entry was accepted or not, the Validators are rewarded the fee proportional to their voting weight.”. What is to stop a whale from abusing this role by either rejecting or approving all content just to make a validation reward - if they are rewarded irrespective of outcome ?

The reward should have a feedback mech, due to success (subscriptions purchased, or content purchased) on the network. The validator also needs to stake his/her tokens for validating some content vs. others during the time he/she is in the role of curating the content for the platform. May need to introduce a time-domain basis for validation of content to create a more liquid dynamic market for this process - this could be aligned with the time-domain weight concept discussed previously. Do you guys do calls on this stuff ? It would be easier to discuss it in an audio format vs. forums for brainstorming. Just an idea.

You are adding a lot more variables into the system that we aren’t 100% sure how that will effect the tokenomics. I’m not actually against the idea but to put it in the first iteration of the system seems risky. The goal was to create a basic system that could be upgrades through proposals. That’s why we took a “keep it simple stupid” approach. Adding reputation is something we are actively looking at but we would need more of a breakdown on it.

We do, do calls but haven’t opened them up to outside participants yet. You can find the recording here: https://github.com/SingularDTV/snglsdao-pm

I would imagine we are going to have an AMA soon about the WP and the mechanisms and this should be something that we talk about.

Keeping it simple is always preferred and a good approach - provided the solution accounts for likely “edge” cases. Otherwise, be ready for a world of headache as flaws become exploited by others…