Discover HybridAGI, a groundbreaking Cypher-based neuro-symbolic AGI designed to make AI adaptable, reliable, and knowledge-driven. With the power of graph technology, HybridAGI goes beyond traditional Agent systems by using a graph interpreter that dynamically reads, writes, and executes graph-based data, enabling self-programming and on-the-fly adaptation.
Tune in to explore how HybridAGI combines graph and LLM capabilities, allowing users to create secure, memory-centric AI applications with deterministic behavior and extensive customization through Cypher.
HybridAGI: https://github.com/SynaLinks/HybridAGI
Notebook: https://github.com/SynaLinks/HybridAGI/blob/main/notebooks/graph_programs_copilot.ipynb
Guest: Yoan Sallami
Knowledge Graph Builder https://dev.neo4j.com/KGBuilder
0:01 - Welcome & Guest Introduction
5:06 - The inception of HybridAGI: Addressing limitations of early LLM-based agents.
6:37 - Why Cypher was chosen for HybridAGI and its benefits for AI programming.
10:12 - Explanation of the four memory types in HybridAGI and their roles.
17:06 - Demonstrating a toy example: Building a Knowledge Graph for a movie recommendation system.
22:46 - Discussing dynamic program calls and embedding program descriptions for enhanced agent capabilities.
26:01 - Importance of protected programs for safety and maintaining architecture integrity.
32:55 - Insights into effective decision-making with LLMs and optimizing prompts for clarity.
37:22 - Exploring the concept of self-programming agents and dynamic task creation in HybridAGI.
41:00 - Why graph databases are crucial in neuro-symbolic systems and their role in reducing LLM dependency.
49:16 - WrapUp
#neo4j #graphdatabase #llm #genai #aiagent