Reputation systems - General
"Reputation systems ultimately need to be based on fact rather than mere opinion or faith to be effective" - Nick Szabo
Globalization and decentralization trends mean that reputation systems are becoming very important. As "attention" becomes the most valuable HUMAN-based resource in the attention economy, once again the reputation systems, which can be very different in different contexts, can greatly help to focus the attention.
Reputation is about gaining from the benefits of trust, loosing from the ommision from the social network, and collaborating in the network risk-sharing. Sharing economy is relying on the development of the reputation systems, while new social norms of trust become natural collectively.
Reputation is critical to self-actualization and self-esteem, which are the top part of Maslow diagram. Attention, new capital, and effort, labour capital, are critically coupled as discussed by Daniel Kahneman. It is yet important to appreciate the risks (and uncertainty) of the new capital, being subjective, idiosyncratic and multi-factored.
Reputation is the common view that there is asymmetric information (Nobel prize summary) and insight/effort of some participants. Signaling (signaling mechanism by Michael Spence) and screening contributes to the reputation. The markets is about the variability and the psread of information among market participants (see On the Impossibility of Informationally
Efficient Markets By SANFORD J. GROSSMAN AND JOSEPH E. STIGLITZ ). See also Alvin Roth's book about matching markets.
- Building A Better LinkedIn by Nick Hutton, while remind yourself about Understanding LinkedIn Business Model
such a platform and a decentralized (not a firm) network can foster the gig economy (see Economist article)
a clear-eyed analysis from the early days of the Internet. The context of ... how the medium transmits and translates the social experience, including its often unviewable cultural and personal cues
- Formalizing and Securing Relationships on Public Networks by Nick Szabo in the classic paper about smart contracts
see the section about Reputation Systems
Reputation systems ultimately need to be based on fact rather than mere opinion or faith to be effective. For example, if we are to have a good credit rating system, we need to be confident that the credit record assembled by the agency is sufficiently accurate.
Technical articles (without blockchain)
Economics model Trust and Social Order from 2009
In our model, network connections between individuals can be used as social collateral to secure informal borrowing. Consumption Risk-sharing in Social Networks
Wikipedia, WikiTrust, and Crowdsensus - content-driven versus user-driven
Thinking in progress
Reputation can be viewed as an illiquid asset class, which is similar to "Fixed Income and Credit". Reputation has associated risk and uncertainty, which means the typical and atypical (unknown unknowns) changes in the recognition/standing/reputation.
There are different types of reputation and different areas/pockets of reputations. Yet the reputation can be monetized (through job and attention seeking), nd therefore different types of reputation can be converted into each other. Reputation is analogous to money as well. Is the reputation closer to equity, debt, or illiquid money? See also how emotions (which can be associated with the reputation) can be classified into basic and their derivatives (Derman).
The reputation can be associated with a person or an object. Reputation can be user-driven or content-driven. So user-driven reputation means the value weight is given by "the node". Yet the value can come from the interaction/field (in physics language), similarly to PageRank. Social links, of course, mean a lot for the reputation of people.
Content-driven reputation is associated with the object (web article, fashion item, science article, etc.), and good content generate forwarding links to the content (just like th reference in science papers to previous important papers)/ Interestingly, content-driven reputation in the context of the web acquires value through the links (in PageRank).
Another point is that algorithm (like PageRank) is subjective and not unique. Static algorithm will have problems if it can not follow the changes in the values/reputation. Also a content-driven reputation is likely not transferable and can be manipulated (Ivan G). Content-driven reputation looks intrinsic and more robust - look for it. There is also the value in de-composing complex reputation (as well as complex risk) into simple components of the reputation. Is there good foundation for the core/basic skills/reputation?
Proof-of-Stake seems to be essentially a reputation system. Moreover, PoS can be value-based, not rigid, consensus. Not a sharp CS (computer science) consensus, which is required for money, but rather social or reputational or scientific consensus. Is there a convenient easy-to-use version of PoS to use for the reputation system? What essential features does PoS mechanism bring relative to a simple tokenization (as used in gaming industry for some time)? The advanced economic limits of bitcoin and the blockchain suggest economic mechanisms of sustainability and attacks, as well as the relation to the reputation systems ( see the end of the paper and Buterin's thinking in "Proof-of Stake Design Philosophy").