Scale-free systems

Example

Scale-free risk & reputation systems

Society has always had implicit theories and notions of risk & reputation. What often matters to you likely doesn’t matter to the majority, at least with the same vigour. Similarly, trust is relative, and individuals trust certain entities based on their own context and experiences. As such, risk & reputation systems, and the underlying models, are social but inherently scale-free. They should strive to seek relevance to the end user.

Yet, over the last two decades, there’s been a concentrated effort to generalize risk & reputation assessment of individuals, businesses, organizations, and institutions to make it seem “explicit”.

Why does it matter?

Existing risk and reputation systems and platforms such as credit ratings, Yelp, google reviews, trustpilot etc. are based on black box models and algorithms underneath. The underlying models promote censorship, are prone to abuse, and are developed by centralized entities who control the switch. How can we trust such authoritative systems to filter information and to make informed decisions? Moreover, if we train AI agents and systems on these models, we are doomed.

So what’s the intervention?

Reppo challenges this notion of homogenous risk/reputation systems by introducing scale-free risk/reputation systems i.e. systems that enable users to assess risk & reputation for their specific context and use-case. We enable such systems using the Reppo Network that incentivizes a new class of risk & reputation models, called Reppo Models, that have relevance to specific user intents, data provenance, ownership attestations, and are built in a permissionless manner.

Reppo models and network of modellers shifts the power dynamic of risk/reputation assessment from large centralized entities to a decentralized network, enriching end-user experiences while progressing censorship resistance.