Creating Scalable AGI: the Open General Intelligence Framework

Daniel A. Dollinger,Michael Singleton

2024-11-24

Abstract:This paper introduces a novel general artificial intelligence systems architecture that provides generalized flexibility and solves current scalability issues plaguing the field. The architecture, OGI (Open General Intelligence), utilizes a dynamic processing system to control and delegate across specialized artificial intelligence modules. It is intended to be used as a reference design for intelligent systems, providing human-like cognitive flexibility for generalized artificial intelligence across various real-world applications.

Artificial Intelligence

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the limitations of current artificial intelligence systems in terms of reliability and scalability, especially their insufficient ability to handle multimodal data and solve complex real - world problems. Although existing large - language models (LLMs) have made significant progress in specific tasks such as image classification, language translation, programming, and writing, their design is usually unimodal, that is, they can only process one type of data (such as text, image, or audio) at a time, which limits their application ability in complex scenarios that require the integration of multiple data types, such as medical diagnosis, quality assurance, equipment troubleshooting, and financial decision - making. These problems cannot be solved simply by increasing computing power, but require a more comprehensive architecture that can handle multiple data modalities.

To address these challenges, the paper proposes the Open General Intelligence (OGI) framework, a new system architecture aimed at achieving artificial general intelligence (AGI). The OGI framework achieves real - time adaptability, multimodal integration, and scalable processing through a modular design, combined with a dynamic processing system and an interconnect structure. Its goal is to overcome the limitations of existing intelligent systems by drawing on human cognitive principles, thereby providing more comprehensive and situation - aware problem - solving capabilities.

Specifically, the OGI framework consists of three main components:
1. **Overall Macro - Design Guidance**: Guides operational design and processing.
2. **Dynamic Processing System**: Controls routing, main objectives, instructions, and weights.
3. **Framework Areas**: A series of specialized functional modules that work together to form a unified cognitive system.

These components work together to enable the OGI framework to support multiple data types, utilize multiple specialized processing modules, and achieve real - time cognitive processing and cognitive process switching through an interconnected processing structure, thereby better simulating the flexibility and processing ability of the human brain.