6.1 Short-term: Improved Mechanisms for Reputation Systems and Applications

Novel Mechanism for Reputation Systems

The reputation mechanism we develop explores new interplays of foundational elements of classical networks theory, e.g., the well-known PageRank algorithm, with game the- ory. We anticipate a major impact in the design of stable reputation systems that incentivize honest behavior and allow the system to discover the ground truth with respect to the honesty of the ranked parties. This can revolutionize classical recommender systems, e.g., service-review platforms like Yelp, Amazon reviews, etc. The potential impact of such an outcome will bring value to the project demonstrating its dominant place within the blockchain ecosystem.

Partnering with Real-Wold Recommender Systems

Following the theoretical investigation of our new mecha- nism, we plan to seek partnership with regional recommender systems, e.g., small-scale goods and/or content distribution services (think of regional alternatives to Amazon). Our contribution will be to offer an engineering team that will interface their recommender system with Möbby, allowing its users to obtain Möbby tokens by using the recommender system. This will give a solution to two of the biggest plagues of such systems: (1) The very low turnout of reviews (this is due to the current very weak incentives to review—effectively relying in user’s altruism); (2) High polarization of reviews—a typical situation in such systems is to have either reviews praising or strongly complaining about services. Instead, our Möbby-induced incentive mechanism will increase the incentive to review in the first place, and with the aforementioned truthfulness-supporting structure, the system will be able to compute rankings based on more and better data. We anticipate success with such smaller systems to open the door to partnership with also the big players in the landscape.

6.2 Mid-term: Built-in (Layer 1) De-Fi operations

As discussed in the whitepaper, the tiered-blockchain architecture of Möbby allows us to enjoy the speed and scalability of an optimal parameters blockchain with the security (safety and liveness) of an assumption with a strong fallback. In addition to the above benefits, our architecture is an enabler for De-Fi operations that are built on Layer 1 and/or make novel use of the reputation system developed and maintained by the Möbby chain. In the following we discuss two examples that are part of our original implementation and deployment roadmap:

1.

Safe transaction at near network speed: The first one is a mechanism that allows to pledge Möbby towards specific receivers (vendors) in a way that they can be used towards payment to these vendors without the need to wait for it to settle on the (slower) backup chain, but where they are not locked to these vendors (they can be used towards anyone. A novel UI and protocol design mechanism ensures that double spending is impossible!

2.

Using users’ reputation as their creditworthiness: The second one, takes advantage of the way Möbby maintains and uses reputation, to allow the system to give low-risk loans to its users, according to their reputation. This can be thought of as a transparent and fair version of the classical credit-score system.

Both the above mechanisms will be included (described and proven secure) in our research team’s upcoming technical report and will once again demonstrate the superiority of the Möbby blockchain.

6.3 Long-term Reputation Driven, Personalized, and Trustworthy AI

Notwithstanding, the overarching goal of the Möbby project is to leverage the transparency of a DLT jointly with the economic incentives of a token to allow its users to converge to a trustworthy reputation system. This can be used to further develop personalized and more robust AI mechanisms that will allow users to make decisions based on transparent, trustworthy sources of data, according to their beliefs, and identify other users with similar beliefs. Such a system can give our digital society an important tool to cope with the dangers posed by the current and projected state of the art in AI. Möbby ’s long-term plan is to facilitate such on-chain AI mechanisms by offering a scripting language (akin to special-purpose smart contracts) that allows its users deployment and use of basic and advanced machine learning functions.