Diving into Reputation

#1

I haven’t seen much detail on reputation systems in DAOstack talk or around other DAO projects. I think it’s a somewhat dangerous concept, so I’d love to get folks’ thoughts and find some solid links to more detailed plans around it.

From what I understand, most DAO builders, including DAOstack, are planning to support reputation currencies that are earned by individuals according to the value of their work, giving them more decision-making power. The purpose is to increase the influence of good actors and reduce that of bad ones, which sounds great and makes sense, but I see a couple frightening problems with it.

First, in most cases we don’t have a widely accepted definition of what makes work valuable. The only case where we do have a good definition is profitability. If your goal is to make as much money as possible, then value equals profit. But how much reputation was my contribution to that profitable project worth compared to others’ contributions? How much reputation do I get for fixing a purely internal problem that didn’t directly make any money at all? If my DAO is a non-profit charity, how will it assign reputation? If reputation is assigned via peer feedback, does that invite corrupt internal politics? These seem like tough questions that could make a lot of angry people and waste a lot of time for an organization.

Second, reputation seems like an endless positive feedback loop: reputation-rich people will be incentivized to vote for decisions that will grant themselves even more reputation, and they’ll have the power to do it. This might be the most serious problem. Too much lopsidedness here could even endanger the decentralization of a DAO, as it would be in effect centralized around the few super rich reputation holders.

A simple solution to most of these questions could be limiting how much power reputation can provide. Is that something that’s in the plans?

3 Likes
#2

I think the answer to this is: it depends. DAOstack is a modular framework, so people will be writing a variety of different rulesets (schemes and restraints). In regards to your point about “what makes work valuable,” I think this depends on the type of DAO: a decentralized hedge fund, for instance, might benefit from exclusively from a 100% profit motive – but even then, I don’t think this would be the case, as most fund managers understand that rewarding risk averse trading is better for the long run management of the fund. In fact, I could see some sort of conditionality applied to larger trades that would require a coalition of high reputation indvidiuals to approve – not just one.

I think the trick here is to start brainstorming what sorts of schemes and restraints would be best for curtailing opportunism and transaction costs created from reputation. In the white paper it’s suggested that reputation could have varying decay rates, for instance, so high reputation individuals are incentivized to continue doing good work for the DAO.

As far as what’s in or not in the plans – it depends on what people build, and what sorts of schemes DAOs choose to implement over time. One of the most interesting aspect of DAOs is that if enough members don’t like the current ruleset, or feel that reputation is lopsided in the hands of opportunistic actors, they could fork into a new DAO with a different ruleset. I imagine “exit clauses” will be a part of some DAOs.

3 Likes
#3

Ah, right, I’d forgotten about the reputation decay! That’s a good way of limiting it.

I guess more details here are really for specific organizations to discuss.

1 Like
#4

Problem number 2 is most worrying, but we shouldn’t blind ourselves: most by default DAO projects such as Bitcoin and to a certain extent Ethereum already have this in a non quantifiable way: be it by number of followers (Vitalik’s place in ETH for example), hashing power, staking ability or importance in the ecosystem (such as Coinbase). In a way DAOstack will allow to quantify and play with this in ways which haven’t been seen before, but it will definitely not be a panacea for all the governance and inequality issues that plague the world.
Nevertheless I would love to see a comparative study of different governance systems and the multivariate analysis of their effects on such metrics such as made in Lijphart’s “Patterns of Democracy”. We probably don’t have such a study because we can only deal with a limited number of large structures of a certain kind worldwide be they states/corporations/ngo’s etc, DAOstack would allow to compare many similar projects that differ by governance and that can be quickly compared.

1 Like
#5

@cryptodani This is definitely one of the reasons to shy away from formal reputation systems – in systems where people can already get clout in other ways (followers, regular social reputation, etc.), adding a formal system might emphasize reputation too much. It’s also pretty easy to imagine times when you really want a reputation system, though: imagine a DAO with fairly open membership where someone can join and immediately have just as much influence as experienced members. That could be pretty disruptive.

Reputation is pretty comparable to seniority in a brick-and-mortar organization. It can cause problems if it’s too rigid, but if you strike the right balance it makes work and decision-making easier.