Joscha Bach Λ Ben Goertzel: Conscious Ai, LLMs, AGI
Theories of Everything with Curt Jaimungal
Expanding Dimensions and Capabilities in Natural Language Models
The training data for LLM relies on dimensions of meaning, which can be difficult to extend, but new combinations can be created. It is possible to set up LLM to explore more dimensions, verify new dimensions through tests, and make experiments in the world using plugins, code, cameras, sensors, and actuators. The transformer algorithm focuses on finding the most likely token, which can lead to suboptimal moves after making mistakes, assuming it's playing against a bad opponent, while the LLM aims to find the best possible token next.
00:00
Transcript
Play full episode
Remember Everything You Learn from Podcasts
Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.
