This video from November 22, 2023 provides a general introduction to Large Language Models (LLMs) suitable for general understanding. The video continues to examples, possible upcoming paradigms such as the view that LLMs have similarities with operating systems. The final part details potentials threats, including jailbreak examples.
What’s wrong with LLMs and what we should be building instead” – Tom Dietterich
This is the keynote video from Thomas G. Dietterich, emeritus professor of computer science at Oregon State University. He is one of the pioneers of the field of machine learning.
The main subject is shortcomings of LLMs (expensive to train and to update; poor non-linguistic knowledge, make false and self-contradictory statements which can be socially and ethically inappropriate.) Suggests a more modular architecture that decomposes the functions of existing LLMs and adds several additional components.
One slide shows:
- perplexity.ai – Click at the bottom of the page to the “Playground” link which gives access to these LLMs:
- mixtral-8x7b-instruct mistral-medium