Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A new solution and concrete implementation steps for Artificial General Intelligence

Yongcong Chen, Ting Zeng, Xingyue Chen

TL;DR

This work tackles the challenge of achieving artificial general intelligence by proposing a human-inspired framework that endows machines with self-needs, emotions, and a directly learnable world model. It replaces exclusive reliance on reinforcement learning with a memory-centered architecture built around a memory bank of tokens, chain association activation, and a dynamic attention mechanism that preserves spatiotemporal token ordering. Key contributions include the seven-step attention/activation pipeline, an innate knowledge preset, and a fully connected knowledge network that enables real-time lifelong learning and explicit decision-making grounded in reward/punishment signals. The approach aims to deliver safe, transparent, and general-purpose AI capable of learning from language and interacting with the world without requiring extensive trial-and-error training for every new task, potentially enabling broad-domain autonomy and safer integration with human goals.

Abstract

In this paper, we propose a new approach to building a artificial general intelligence with self awareness, which includes: (1) a new method to implement attention mechanisms; (2) a way to give machines self-demands; (3) how to form a value evaluation system compatible with the network; (4) a way to create the world models; (5) how to realize a top-down, hierarchical thinking decision-making chain; (6) a way to achieve general decision-making and response capabilities; (7) a way for a machine to directly obtain human experience through language. In the paper, we first analyze some of the shortcomings of current LLMs (Large Language Model) and propose ideas for improvement. Then we analyze why our scheme can solve the above problems and provide detailed steps for implementing our scheme. In chapter 4, we have presented a step-by-step mplementation roadmap. And in chapter 5, we have presented a specific implementation demonstration. In chapter 6, we analyze the advantages and disadvantages of our scheme and propose further research directions. In this article, we have put forward how to create genuine artificial general intelligence step by step. It can handle data of all modalities in a unified form and can directly understand the experience that humans already possess through language, thus avoiding the problem that reinforcement learning is required for every decision-making process.

A new solution and concrete implementation steps for Artificial General Intelligence

TL;DR

This work tackles the challenge of achieving artificial general intelligence by proposing a human-inspired framework that endows machines with self-needs, emotions, and a directly learnable world model. It replaces exclusive reliance on reinforcement learning with a memory-centered architecture built around a memory bank of tokens, chain association activation, and a dynamic attention mechanism that preserves spatiotemporal token ordering. Key contributions include the seven-step attention/activation pipeline, an innate knowledge preset, and a fully connected knowledge network that enables real-time lifelong learning and explicit decision-making grounded in reward/punishment signals. The approach aims to deliver safe, transparent, and general-purpose AI capable of learning from language and interacting with the world without requiring extensive trial-and-error training for every new task, potentially enabling broad-domain autonomy and safer integration with human goals.

Abstract

In this paper, we propose a new approach to building a artificial general intelligence with self awareness, which includes: (1) a new method to implement attention mechanisms; (2) a way to give machines self-demands; (3) how to form a value evaluation system compatible with the network; (4) a way to create the world models; (5) how to realize a top-down, hierarchical thinking decision-making chain; (6) a way to achieve general decision-making and response capabilities; (7) a way for a machine to directly obtain human experience through language. In the paper, we first analyze some of the shortcomings of current LLMs (Large Language Model) and propose ideas for improvement. Then we analyze why our scheme can solve the above problems and provide detailed steps for implementing our scheme. In chapter 4, we have presented a step-by-step mplementation roadmap. And in chapter 5, we have presented a specific implementation demonstration. In chapter 6, we analyze the advantages and disadvantages of our scheme and propose further research directions. In this article, we have put forward how to create genuine artificial general intelligence step by step. It can handle data of all modalities in a unified form and can directly understand the experience that humans already possess through language, thus avoiding the problem that reinforcement learning is required for every decision-making process.
Paper Structure (35 sections, 2 figures, 2 tables)

This paper contains 35 sections, 2 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Characteristics of human-like intelligence
  3. A Path to Realize Human-like Intelligence
  4. Brief description
  5. Implementation details
  6. Step 1. Tokenize information
  7. Step 2. Matrixize tokens
  8. Step 3. Perform similarity activation from input tokens to tokens in the memory bank
  9. Step4, Perform proximity activation for all activated tokens
  10. Step5, Chain Activation Process
  11. Step6, Accumulate activation values
  12. Step7, Activation values fade over time
  13. Step8, Obtain Pre Training Weight Matrix
  14. Step 9, Preset Minimum Machine Requirements
  15. Step 10, Form a Fully Connected Knowledge Network
  16. ...and 20 more sections

Figures (2)

  • Figure 1: The method To preset innate knowledge
  • Figure 2: A workflow diagram