AGI Research

The paper will be published after AGI is completed (after providing and validating the sample code).
Currently, We are only posting a memo for those who are doing the same research.
Here, instead of "intelligence = brain", it's a new paradigm that defines intelligence mathematically.
We think that most of the researchers aiming for AGI are already in an irreversible position.
We hope that there will be at least one person who aims to become an AGI, even if it means throwing everything away.
There are few people who can work on it, and the members concentrate on research, so the disclosure of the report becomes thin.


Definition of Universal AGI

Misinterpretation and Exploring the Axiom of Induction

Inductive inference of perfect quality. y=f(ε)


Deterministic inductive inference. y=f(x)


Non-deterministic inductive inference y=f(x,ε)

Inference integration

Inference with coordinate transformation

Inductive inference for all possibilities

Recursive Inference

Axiom of Induction

Complex Inference

Bias Calculation

Level1. Universal Artificial General Intelligence


Calculation Priority

Calculation Abort

Inferring Priority of Computation

Input and Output

Retention of Information


Level 2 Universal Artificial General Intelligence

Sample Program