Embodied Asynchronous Multi-Tier Artificial General Intelligence Architecture v1.0.0

I have proposed a new theoretical AGI Architecture that explains how current AI models are inadequate for imitating Human Intelligence and also provides an architecture that is theoretically capable of such a feat.

This Architecture introduces many ideas simultaneously and makes a lot of attempts at uniting many research ideas popular currently. Some such Ideas are:

  • Asynchronous Sparse Mesh (ASM) Architecture

  • Multi-Tier Learning

  • Introduction of Emotions to an AGI model

  • Evolvable Intelligence

I gladly invite any technical and biological critiques on this theory, especially on:

  • Hardware Assumptions

  • Training Methods

  • Philosophical Claims

  • Feasibility

For the full proposal, visit this GitHub Repository;

Embodied-Asynchronous-Multi-Tier-AGI

Looking forward to a serious discussion.

No comments.