ETHDenver - A Reputation System For Bounty Platforms

[vs77bb]

This is an ETHDenver Bounties Track Issue.

Bounty submission idea by Bryant:

With people moving increasing into remote working contract roles, and especially for the decentralized app economy (e.g. the free-agent "blockchain man" moving between dapp projects), the need to identify and trust partners you might not be able to meet in person will become increasingly important.

We do this currently through a mish-mash of different platform-specific reputation systems that get leveraged for any major app in the gig economy (Uber, AirBnB, TaskRabbit, eBay, Amazon, etc.). You have to build reputation on each one of these systems separately, and it's hard to get paying clients until you do. There's no relationship between all these systems that builds a reputation for the long term, and no way to leverage that long term reputation for anything besides these specific gigs.

Similarly, the means to determine reputation for longer term gigs (aka jobs) is via the resume, a one page distillation of your personal skills and talents that has been a terrible but ubiquitous piece of technology from the get-go. Even worse, Applicant Tracking Systems, built to contend with the needs of managing 1000s of applications in your organization's applicant silo, distill this already-terrible representation of who you are down to a percentage score "representing" how well you fit a job posting. This hugely complex and uncoordinated system is meant to help you understand a very human need: whether a potential partner's reputation and skillset fits the needs you might have in your organization.

The solution is to build a "reputation economy" that trustfully shares tens of thousands of datapoints related to the context in which a person works and how they are successful, built up over time via the personal interactions of those we work with on a daily basis. It needs to share this data in a secure and open way, preventing any one group from being marginalized and any one metric from becoming the Social Credit Score through which all citizens are judged.

There are many different issues with this approach relating to trust, accuracy, privacy, and adaptability, but the hope is a blockchain-based dApp can solve these issues using encryption, off-chain/on-chain interactions, and tokenomics to build an economy for this highly-sensitive data to change how we find and maintain our working relationships for the better. Most importantly, this dApp must move a portion of the human experience into the digital age, as the existing infrastructure around this economy is woefully inadequate and leaves millions of people stuck in unfulfilling work with little chance of advancement.

To sum it up, "Our reputation is our most important asset"

[vs77bb]

What

A reputation system tracks data points across platforms (such as Gitcoin and Bounties Network) and includes cross-functional reputation scores.

Requirements

• Create reputation profiles which proliferate across at least two platforms (like Gitcoin and Bounties Network)
• Track reputation across at least three categories

[gitcoinbot]

This issue now has a funding of 1.0 ETH (910.49 USD) attached to it.

• If you would like to work on this issue you can claim it here.
• If you've completed this issue and want to claim the bounty you can do so here
• Questions? Get help on the Gitcoin Slack
• $10461.41 more Funded OSS Work Available at: https://gitcoin.co/explorer

[tra38]

I'm not sure if the hackathon is still continuing, but here's my thoughts on the matter.

Most of the time, only one person applies to work on a bounty. It doesn't matter how much (or how little) "reputation" a person has; they would be the only person who wants to work on it, so they would get the gig by default. Reputation systems only really come into play if the bounty owner have two or more people who want to work on a bounty...and neither of them are willing to share the work/reward in question. In that scenario, I would think that the bounty owner gets to decide which one gets to work on the bounty, and then keep the other people as backup. I'm not sure how common this scenario would be though.

What I would find interesting is a way to get a list of all bounties a person ever worked on. Essentially, this would be duplicating the "resume", but the difference here is that due to the open-source nature of the work, the bounty holder could look at the exact details of the bounties people worked on, see how those people worked on the bounty, and use those data points to choose the best candidate. I don't need to rely on text on a single piece of paper. I can verify the work myself.

This information would be very useful even if one person applied to the bounty, as it would give me some data points to understand how long that person would take and the expected quality of the work.

Reading how people behaved at different bounties does require some effort on the part of the bounty owner though, but this might be justified...I might only care about Alice behaved on Bounties Y and Z, and not care how Alice worked on Bounty X (that work isn't really relevant). I suppose a bounty owner could pay an "HR Screener" (a third-party bounty hunter) to scan the histories of all people who applied to a specific bounty and then to pick one person to work on the bounty, but the bounty owner still have to find the "HR Screener" to begin with. Ultimately, the bounty owner must put in some time and effort into researching the candidates (whether that would be candidates for the "HR Screener" role or for candidates for the bounty itself).

[leonprou]

Did you heard of colony? Reputation is big part of their platform

[tra38]

It appears that Colony's reputation system exist to facilitate "project management" (especially deciding what tasks even get done in the first place), and doesn't really seem to apply to scenarios where the "bounty owner" already have a task in mind and is just looking for warm bodies. This seems pretty clear even from the front page:

In an open organization, you're empowered to do the work you care about, not just what you're told to do.

That's not the case with bounty platforms. It's a pretty interesting idea though, and definitely worth exploring further.

---

While re-reading the original spec, I saw this quote...

We do this currently through a mish-mash of different platform-specific reputation systems that get leveraged for any major app in the gig economy (Uber, AirBnB, TaskRabbit, eBay, Amazon, etc.). ... There's no relationship between all these systems that builds a reputation for the long term, and no way to leverage that long term reputation for anything besides these specific gigs.

I don't see that as a problem honestly. Just because you're an awesome Uber driver doesn't mean I should rent a house from you via AirBnB. Different domains require different skillsets, and while some knowledge transfer can occur between domains, it's unlikely to be enough to justify building a whole application to represent that.

If different applications deal with the same domain (Uber and Lyft, for instance), then it makes more sense to be able to present a record of your work.

[collinvine]

Hi there, I'm part of the Colony team. Did this end up getting worked on? I'd love to talk with anyone who is interested in creating a standard reputation system for the decentralized economy. We're doing a lot of work on this now and want it to be useful to others with different use-cases. Feel free to get in touch to discuss more (collin at colony dot io).

[RhysLindmark]

This was the team that won at ETHDenver btw: https://devpost.com/software/indepact

I'll leave it to @vs77bb to incorporate the winning team into this discussion!

[vs77bb]

Thanks for poking in, @RhysLindmark. INDEPACT team (@thinkmassive, @hillpts, @inkblotty, @morgansliman) congrats on your winning project! Would you mind making a submission on Gitcoin (directing us to your repo)?

Also see note above from @collinvine if you're interested in collaborating with him, post-hackathon 🙂

[vs77bb]

What I would find interesting is a way to get a list of all bounties a person ever worked on. Essentially, this would be duplicating the "resume", but the difference here is that due to the open-source nature of the work, the bounty holder could look at the exact details of the bounties people worked on, see how those people worked on the bounty, and use those data points to choose the best candidate. I don't need to rely on text on a single piece of paper. I can verify the work myself.

@tra38 I think this is a great point and sets the stage for what reputation might look like as a v1 for Gitcoin. cc @owocki

The submission for ETHDenver was meant to be more generic in nature, though it did turn out to have some overlap with how we may think about rep going forward. Will share some more of our thoughts here soon re: Gitcoin!

[inkblotty]

Hey @vs77bb! Our project is at https://github.com/workpact -- both the web client and smart contract. Since we went at it from a different angle, I'm not sure if I should submit it as completed work directly though the Gitcoin network. Does a successful submission close a bounty?

@collinvine if you're interested, we'd get your input! INDEPACT was designed specifically for a non-tech audience (we had a team-goal to reach out to a non-tech market), so it'd could take a little tweaking depending on the use case.

[vs77bb]

Hi @inkblotty - thanks for the link! Could you provide Kevin and I a Metamask address over Slack where we can send the funds? @owocki this one is expired so we can pay out via tip

[gitcoinbot]

Work has been started on the 1.0 ETH (422.33 USD @ $422.33/ETH) funding by:

1. @vs77bb

Please work together and coordinate delivery of the issue scope. Gitcoin doesn't know enough about everyones skillsets / free time to say who should work on what, but we trust that the community is smart and well-intentioned enough to work together. As a general rule; if you start work first, youll be at the top of the above list ^^, and should have 'dibs' as long as you follow through.

On the above list? Please leave a comment to let the funder (@owocki) and the other parties involved what you're working, with respect to this issue and your plans to resolve it. If you don't leave a comment, the funder may expire your submission at their discretion.

• Learn more on the gitcoin issue page
• Questions? Get help on the Gitcoin Slack
• $3428.65 more Funded OSS Work Available at: https://gitcoin.co/explorer

[owocki]

bounty has been paid out!