What is the proof of intelligence algorithm ?

What is the proof of intelligence algorithm ?

One of the most important feature of the network and that will power the KPY token is the POI or Proof of intelligence.

To explain it let’s take a search engine like google and explain how it works (very briefly) for those who don’t know yet.

  • Google is ranking your website based on scoring;
  • It uses an algorithm that includes more than 200 different criterias from simple to complex;
  • Based on the result it provides a score which will determine where your page is ranked in the search results.

This is a very basic overview of how Google works to sort its content. Now how does that apply to the Kryptopy Network ?

The Proof of intelligence algorithm is similar in the way that it will score content to give the rewards and avoid spam with a scoring system that is having various criterias that are not necessarily known to the wide public in general to keep the algorithm secret and efficient (if everyone knew about it they could try to trick it ;)). Let’s give a very simple example to show you how it works :

JOHN POSTS : I really love cryptocurrency please share with me what your favorite ones are !

ALICE REPLIES: Oh I really love it too KPY is my favorite because of X Y Z

JACK REPLIES: Yea me too

GIGI REPLIES: Check this out http://www.somecryptosite.com/ you will like it

This is a simple scenario where :

JOHN will get per example 0.0005 KPY (or 5 mKPY) for initiating a positive post that includes an invitation to participate which will hopefully lead to more actions for the users on the network.

ALICE gets 0.0002 KPY ( or 2 mKPY) for replying something very pertinent and explaining why it is in a positive way. She proves she has knowledge of the topic and she probably also owns the currency she is talking about (another criteria that will be checked).

JACK is relevant but doesn’t show any proof of his intelligence by posting such a short sentence, 3 words sentences like that are typically use by followers and psychologically talking does not show a lot of involvement. Such a comment would have no reward as it doesn’t bring or remove any value from the network.

GIGI is definitely looking like a “spammer” maybe in all honesty she just wanted to share a link. If that’s the first time she does, it wont be penalized however she won’t be rewarded unless the link she did share is very relevant to the topic which will also be determined by text analysis.


As said before this is a very simplistic example to show how the Proof of intelligence algorithm is working with text analysis.

Proof of intelligence VS decentralization and data privacy ?

Data privacy is the #1 priority for Kryptopy. However we will need a “centralized” database to process data where no information will be readable “in context” or accessible by anything else than our machine learning engine. In machine learning, you can evaluate patterns which are stored as words with 0 and 1 in our case. In the very unfortunate event the system would be compromised, the data in that database would not be very valuable to someone who does not have access to the underlying algorithm.


No Comments

Leave a Reply