Are decentralised celebrity reputation trading markets possible? Celebdaq but for real? This is my exploration into how it may be possible to create a community lead reputation market.
“…Manfred checks into a hotel in Milton Keynes. He watches the sun set from his bedroom window, an occlusion of concrete cows blocking the horizon. The room is functional in an overly naturalistic kind of way, rattan and force-grown hardwood and hemp rugs concealing the support systems and concrete walls behind. He sits in a chair, gin and tonic at hand, absorbing the latest market news and grazing his multichannel feeds in parallel. His reputation is up two percent for no obvious reason today, he notices: Odd, that. When he pokes at it he discovers that everybody’s reputation — everybody, that is, who has a publicly traded reputation — is up a bit. It’s as if the distributed Internet reputation servers are feeling bullish about integrity. Maybe there’s a global honesty bubble forming.” — Accelerando
I read the above on the flight home from presenting blockchains at King’s internal conference and started thinking about how reputation markets could work in a decentralised world. How could we create reputation markets that are quoted in the celebrity section of the sun and the gossiped about on Twitter? You may grimace at the very idea but I see it as an interesting experiment into where decentralised technology may take us.
To make clear right from the very start — this is not an ICO and never will be, by design. This is also a theoretical discussion only unless a few people want to make it happen. It also seems that today the SEC have passed comment on celebrity endorsements of ICOs, so let’s talk about how they can even have their own but not in the way they may think!
Problems
- How can a reputation market be created without the permission of the subjects?
- Is it possible to stop individuals creating and selling other people’s reputations?
- How can this be seen as legitimate in the decentralised space?
- What gives the the reputation market value?
- How can we prevent blatant manipulation of the trading market?
The solution to 1, 2 and 3 may be token design. The solution to 4 is apparent from meme coins and projects like such as pepe, but I can think of many ways to give it value. Number 5 isn’t so easy, since it lies at the heart of valuing reputation itself — and is really the reason for writing this post following a lovely critique I had at devcon 3 today.
Token Distribution
- Create a fixed supply of a base token. Let’s call it FAN.
- Airdrop the majority to anyone on reddit with suitable mature account who expresses an interest. Accounts get an equal share.
- Retain the rest for development, bounties, whatever.
Reasoning: This should be owned by the community, other than where early stage organisation and expenditure is required. Reward for those early developers and active community members is quite reasonable, and it’s only in the token in any case.
Token Purpose
- Allow anyone to create a new reputation token. The subject is specified by linking to a Wikipedia entry or Twitter profile.
- Creating a token is free other than gas fees.
- Over a 30 day period, allow anyone to buy into the token using FAN.
- At the end of the period, tokens are evenly distributed based on FAN allocated.
- 10% of the token is airdropped to all FAN holders. However, they can only claim it by burning their FAN (alternatively, a pure airdrop of FAN or airdrop with a lock-in period?).
- After a lock-in period of X days, users can exchange their token for the FAN should they so wish, but this burns the FAN. Optional: Gradually burn the FAN. Optional: Burn a fixed % of the FAN.
Reasoning: No one should be able to monetise someone else’s reputation without everyone having a chance to do it. This is also completely decentralised — no one can stop it existing.
Additionally, no one should be able to raise money by creating a reputation, at least within the system. The token distribution is based on locking up tokens, and perhaps allowing them to burn over time if owners continue to believe on the overall token (I’m a big fan of permanent sinks).
The above creates a community incentivised to support the ecosystem.
Valuing Reputation
My initial thought was the value will simply be the market cap of each token. However, liquidity is likely to be small and hence easily manipulated. The result may be that no one trusts the data and the reputation market fails. It’s certainly a very real possibility and was the crux of an earlier conversation with Jake Goh who knows this stuff far better than I.
The alternative may be a PageRank type algorithm which occurs off-chain (and potentially is published on-chain for an incentive, using Truebit or a colony.io like system). The inputs could be:
- Established trading market prices
- Age of account holder + amount in account
- FAN stake of account holder + amount in account + time stake has been in account
A smoothing algorithm could prevent spikes occurring, this forcing anyone trying to purchase reputation to have to spend funds over a long period of time.
One of the problems here is really establishing the inputs and smoothing such that a meaningful value can be attributed, and another is trying to get a fair result whilst pushing up the cost of manipulation significantly.
It’s my understanding that Google’s PageRank is kept very secret and with a very large number of inputs to make it hard to game.
Can we do the same thing but in an open environment?
How could we do this?
There are other problems that would need solving too but I wanted to put pen to paper, or fingers to keyboard, and get the idea out there. It may be truly rubbish but it’s the type of pontificating that makes Ethereum so enjoyable.
I’m at devcon 3 so if anyone wants to talk about this or anything else then message me on Twitter@alex_amsel.
