Knowledge Structures in AI
Jamie discusses the limitations of current AI systems in creating symbols that align with non-linguistic conceptual representations. He suggests that by focusing on better knowledge structures and reasoning systems, even without machine learning, we could enhance the effectiveness of expert systems. This approach might have prevented past failures and AI winters.In this clip
From this podcast
The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence)
Creating Robust Language Representations with Jamie Macbeth - #477
Related Questions
How does symbolic learning work in the context of the episode AI Today Podcast: AI Glossary Series: Symbolic Systems & Expert Systems and the clip Symbolic vs. Statistical AI?
How is symbolic reasoning used in natural language processing (NLP) as discussed in the episode #103 - Prof. Edward Grefenstette - Language, Semantics, Philosophy and the clip Language Model Conceptual Abstraction?
How does symbolic reasoning compare to neural networks in the context of the episode Human-AI Collaboration for Creativity with Devi Parikh - #399 and the clip Neurosymbolic Generative Art?