Relational Intelligence is a smarter, more human approach to artificial intelligence. It is a new data and toolset which enables people to objectively understand, predict, and trigger change in complex adaptive systems. Its purpose is to understand what triggers change in relationships of components in the same layer (I.e. intra-relationships and between layers (I.e. inter-relationships) to create emergent micro- and macroscopic patterns. Existing approaches were limited by their geometries,

- Euclidean geometry which has 1 infinite plane reaches limitations at edge cases producing false positives and false negatives
- Spherical geometry which has 1 finite manifold only enables understanding of intra-patterns

Relational Intelligence therefore was created based on the Thurston Geometrization Conjecture which inter-connected the 8 different geometries (including Euclidean and Spherical) which up until this point existed as independent paradigms.

It is not to be confused with artificial intelligence, many of which have evolved from Geoffrey Hinton and similar neural network based, nor should it be confused with agent-based modelling approaches to understanding CAS. Both fall into Euclidean, and sometimes into spherical geometry categories.