Expert Network - request for funds

Iā€™ve updated the Nest funding proposal for the expert network on Github. Also including a copy here.

Summary: The proposed expert network is a directory of all Aragon DAO members, together with their skills and associated reputation. The following opportunities and challenges were identified during the group discussion:*

  • A freelancer marketplace like Bounties.Network or Upwork could be created as part of the expert network. Alternatively, we could just integrate or fork bounties.network. This could be used for hiring proven talent, which is typically the largest challenge that founders have when building a startup.
  • It was suggested that the expert network could also provide a critical missing part of the architecture for Aragonā€™s upcoming arbitration court. Specifically, it would enable a fairer system to decide who is qualified to arbitrate a specific issue.
  • It was suggested that ā€œreputation and tasksā€ could be a strong motivator to drive users to the Aragon network. Seeing what tasks you can perform for Aragon DAOs and knowing that you will earn a reputation score for doing them would likely increase the number of people wanting to work for Aragon DAOs.
  • It could be used as a filtering system to reduce the amount of notifications to users and ā€œInformation Overloadā€ through more targeted filtering of tasks based on a userā€™s skill sets. For example, when assigning new arbitration requests or other tasks created from within other Aragon DAOs (such as with the Open Enterprise ā€œProjectsā€ app), related notifications could be sent only to users with matching skills in that domain. This would also be useful in helping the overall communication and proposal making systems to scale through more targeted notifications.
  • It was suggested to integrate Sourcecred, a reputation protocol for open collaboration. Sourcerank uses the PageRank algorithm to generate a score, called cred for every contribution. By integrating Sourcecred into the Expert Network, it would enable us to understand how worthy member contributions were to Github repos, Twitter, Discourse forums etc. This would generate valuable reputation-based data points for both technical and non-technical communities. It is envisaged that Sourcecred would just be one data source of many that makes up a usersā€™ weighted reputation score. Other potential data sources include:
    • a) The task rating given to a user upon task completion within the Open Enterprise ā€œProjectā€ app.
    • b) Member reputations on Bounties.Network.
  • It was also suggested to integrate 3box (already ongoing by Autark) to view more friendly user profiles together with their native user identity validation (Facebook, Twitter, Github etc). The idea would be to then display the userā€™s skills and reputation score next to their 3box generated profile.
  • The exact formula to calculate reputation requires further research and continuous optimisation. Some members thought it was unfair to use ā€œamount paidā€ to an individual as a reputation proxy, given that pay asymmetries exist based on peopleā€™s geographic location, discrimination etc.
  • Another idea was to use ā€œmember reputationā€ as an optional method of distributing DAO proceeds between its members instead of existing distribution methods such as proportion of tokens held.
1 Like

Been thinking about the TAM for Aragon. Something like the ā€œexpert networkā€ could expand the value of the Aragon network considerably. It would lower barriers to launch quality products, increase the appeal of platform to talent, and reduce the cognitive overhead required to find relevant actionable information. In short, something like this needs to exist.

Looking at your Nest Grant application, it seems like thereā€™s still some open design questions (SourceCred vs Colony, technical challenges, etcā€¦). Iā€™ve talked to the Colony team a little and the SourceCred team a lot so happy to brainstorm on this.

Also, In the Nest Grant I did not see a lot of discussion regarding the UX/UI of this app/protocol. While the timeline and financing seem appropriate for a prototype, making something that people actually use will probably be 10X more difficult. Since the ā€œexpert networkā€ is a network (of people), UX/UI seems essential to bake into the DNA of the project vs being an after thought. Would you consider expanding the ask in order to bringing a front-end designer onto the team?
(if the expert network was already setup this would be way easier! lol)

Agreed on the TAM. There are quite a few comparables to draw from:

  1. Upwork (2019 revenue expected to be $300m)
  2. Guidepoint (Does circa $140m revenue annually)
  3. Toptal (Does $100m+ year)

I did not see a lot of discussion regarding the UX/UI of this app/protocol. While the timeline and financing seem appropriate for a prototype, making something that people actually use will probably be 10X more difficult. Since the ā€œexpert networkā€ is a network (of people), UX/UI seems essential to bake into the DNA of the project vs being an after thought.

What are you thoughts on forking bounties.network? That might considerably reduce the overall workload required. And we could add additional filters to it so that users can filter jobs by a) All Aragon DAOs, b) Aragon DAOs they are part of.

Would you consider expanding the ask in order to bringing a front-end designer onto the team?

The financing amount I put in was more of a placeholder, pending a more in-depth technical review and discussion with other Aragon teams to better understand how this might compliment/overlap with other projects. I had a call with @lkngtn earlier this week to discuss plans for Apiary/DAOlist (as it seems like an obvious future feature of DAOlist to want to also display the people behind each Aragon DAO and make it more social). The work that Luke did at Eth Boston for Credao/SourceCred is also potentially very relevant here for a v1 form of reputation tracking.

Separately to this, @chrishobcroft is working on MyDaoDashboard. I understand that is also using an Apiary API which is great. I would like to better understand from @chrishobcroft and @lkngtn what the future co-ordination plans are here to try and avoid feature overlaps.

It also feels like @stellarmagnet and Autark should be involved in this discussion. The freelancer marketplace / expert network should ideally integrate well with the Projects app, following the same skills taxonomy (I donā€™t think the Projects app currently lets you tag a task to a skill set) and various other task parameters such as completion status. Once a user has created a task from within the Projects app, the obvious next step is finding the best person in your immediate network or beyond to complete it.

All of this I think raises an important point around effective project coordination across the various Flock & Nest teams. In short, Iā€™d be more than happy to increase the requested budget for this if necessary but would like to first get more feedback from the community. I think @lkngtn, @stellarmagnet and @DanielS will all likely have a valuable perspective to share on this.

1 Like

Our general goal at Autark is to have user research drive which features we focus on ā€“ and much of what is mentioned here is aligned with the types of products we plan to build or already have built.

The Projects application that Autark has developed interfaces with the StandardBounties.sol contract, which is the contract that Bounties Network uses. The current strategy has been to interface with the same contract they have deployed instead of forking it.

The Autark dev team needs to evaluate what you have proposed here, look into Sourcecred more, and validate if it will technically work as far as our ā€˜Expanding Projectsā€™ initiative goes, without requiring a new contract all together.

But most importantly, we want to prioritize the features for Projects using a community curation process that we will be announcing soon.

As part of our Sustainability Initiative, we have also mentioned the idea of a professional services network, where on top of the general offering, discounts can be provided when paying with ANT. People with ANT holdings may prefer to be paid in ANT. Autarkā€™s Research Topics - HackMD

This can be done using our Rewards application, as reputation is a non-transferable token as far as Aragon is concerned (hence it is in proportion to your reputation earned, which is the same as number of tokens held)

I think this will be one of the most challenging aspects to solve, I donā€™t think there will be a one-size fits all solution.

Yeah, this enhancement would fall under our ā€˜Expanding Socialā€™ initiative.

We have plans in the pipeline for an organization to further classify the token managers installed in their DAO, so we can add a field for the DAO to classify whether itā€™s reputation, membership, or equity token etc. Then when we display the tokens on the userā€™s profile it can be displayed in the portion of the profile reserved for reputation. But right now reputation in Aragon is one-dimensional, so thatā€™s the part that would need more thought & architecture.

Hence if a Nest grant were to be created, I would say it should be focused on the technical design & architecture of the reputation system ā€“

  • What changes do we need to make to the token manager, if any?
  • Should the Projects app be used for this use case, or is a new standalone application better for UX?
  • If Projects, what changes do we need to make to the Projects contracts, if any?
  • What is the optimal UX for reputation management ā€“ from a perspective of managing formulas, distribution schedules, delay etc?

It wasnā€™t on our agenda to make reputation multi-dimensional, so if others want to take on the technical design & research, it will help. But I agree with @burrrata that solving this is in large parts a UX endeavor.

1 Like

Great! The more I think about it, the more having a ā€œblock explorer,ā€ but for Aragon DAOs is essential. Unlike a block explorer, however, itā€™s an open design space. Weā€™re currently brainstorming lots of ideas for how to build and evolve Apiary. Can you please open an Issue detailing exactly what you mean here so that we can discuss and potentially integrate it into the roadmap?

What about instead of forking a system like bounties network, i would fork https://erasurebay.org/

Information finds you :crystal_ball:

Lock up a DAI reward. Anyone in the world can fulfill it. They must stake cryptocurrency and upload a file containing the requested information.

Release the reward if you are satisfied with the upload. Destroy their stake if you are dissatisfied :fire:
The destroyed stake will be converted into ANT and ANT will be burned.

Erasure Bay is decentralized, encrypted and unstoppable. All requests are tweeted.

Make a request :mega:

Description

Enter a short explanation of what youā€™re looking for. This can include links, Twitter handles and hastags. Make your descriptions as clear as possible.

Reward

An amount of DAI cryptocurrency you are locking up as a reward. This will be transferred into an escrow when you make the request, you make sure you have this in your wallet. Like this fulfillers can see you really have the money and will take your request seriously. (Once someone fulfills your request it is added to their stake and you will not get it back, you can only punish it.)

Fulfiller stake

This is what makes Erasure Bay powerful. This is how much DAI cryptocurrency someone will need to deposit when fulfilling your request. You can destroy a fraction or all of their staked money if you are dissatisfied with what they provide. This will stop people responding with spam or bad information. It usually makes sense to have this be roughly 10% - 50% of the reward.

Punish ratio

How many dollars it will cost you to destroy one dollar of the fulfillerā€™s stake. For example; if you set the ratio to 0.1 and punish a fulfiller who staked $100, it will cost you $10 to destroy their entire stake. This protects the fulfiller from reckless punishment. The default value is good for most requests.

Punish period

How many days after your request is fulfilled you have to verify the quality of the information provided. Within this window, you may punish the fulfiller. After this time their stake and reward are released. You may decide to release it early if you are satisfied with the submission. The default value is good for most requests.