The future is neuro-symbolic: Expressiveness of ChatGPT and generalizability of symbols (SymbolicAI)
15What if we could process information through graphs?
What if these graphs can be represented by symbols? Each symbol can be the root or can be expanded. If a symbol can be expanded, we call it an expression. We can take in a query as an expression, and try to process it using engines such as web search (Google), memory retrieval (Pinecone), speech to text (Whispr) (similar to LangChain with tools), and then try to decompose it all the way to the root node symbols. That way, the entire chain of evaluation will be explanable and also error-traceable!
Each symbol is actually very flexible. Whatever you can express as a string in Python is a symbol. They can also be manipulated and compared with each other using fuzzy logic (powered by GPT). At the lower level, symbols store information. At the higher level, symbols represent operations.
Join Leo and Marius and explore the world of neurosymbolic AI - using symbols to represent everything! Cool applications of this include web crawlers, chatbots, casual reasoning agents and many more!
~~~~~~~~~~~~~~~
Repository: https://github.com/Xpitfire/symbolicai
~~~~~~~~~~~~~~~
0:00 Introduction
1:55 Theoretical Introduction of Information Process and Symbols by Leo
43:04 SymbolicAI Architecture by Marius
1:22:40 Q&A
~~~~~~~~~~~~~~~
AI and ML enthusiast. Likes to think about the essences behind breakthroughs of AI and explain it in a simple and relatable way. Also, I am an avid game creator.
Discord: https://discord.gg/fXCZCPYs
LinkedIn: https://www.linkedin.com/in/chong-min-tan-94652288/
Online AI blog: https://delvingintotech.wordpress.com/
Twitter: https://twitter.com/johntanchongmin
Try out my games here: https://simmer.io/@chongmin