Active Graph Agent Runtime (BabyAGI 4) — Yohei Nakajima, Untapped Capital

Summary

Yohei Nakajima presenta ActiveGraph, un enfoque experimental de código abierto para construir agentes autónomos que se aleja de la estructura centrada puramente en LLMs. En lugar de eso, propone una arquitectura basada en un registro de eventos inmutable (log) que actúa como la fuente de verdad del agente. Al tratar el log como el núcleo, los agentes pueden ser auditables, recuperables y capaces de auto-modificarse mediante un sistema de 'comportamientos' y 'políticas'. El sistema facilita que los propios agentes construyan otros agentes, gestionen su memoria y experimenten para mejorar su rendimiento sin necesidad de reiniciar procesos desde cero cuando ocurren fallos.

Topics covered

• Evolución de BabyAGI
• Arquitectura log-centric

+5 more topics (full content when you follow the channel)

Transcript excerpt

[music] >> Hi everybody. Thanks for coming. I'm excited to be here. AI engineer warfare has been so fun meeting everybody. Um, but I'm here to talk about active graph, which is my new open-source experimental approach to building agents, which looks a little bit different than maybe you've been building agents. Uh, it's definitely experimental. The idea is more to give you inspire you with some potentially new ideas. Um, agents are awesome, but long-running agents break. And if they're so awesome, why am I still building them? Why they should build themselves. Let's build the simplest thing that can build itself has basically been kind of my research theme for the last 3 years since I did baby AGI back in March of 2023. So that's over 3 years ago. If you were there at the time, it was crazy. It went wild like it was covered by media. People thought it was going to work. It didn't work at…