Hear a lot about LLMs and don't know where to start? This talk will guide you gently through the world of LLMs and how to interact with them. I will start with one of the seminal papers that defined a novel model architecture that got this started, and go through the following questions:

1. What is an LLM?
2. How do LLMs work? What do they do?
3. What tasks are LLMs well suited for?
4. How can I deploy an LLM? Should I use LLMs for my given business use case?
5. What does a typical LLM application development cycle look like?
6. What is zero-shot, one-shot, or few-shot learning?
7. What is in-context learning?

I will also demo some cool concepts, like

1. Chat with your PDF (Retrieval Augmented Generation)
2. Neural Machine Translation
3. Summarization / Sentiment.
4. ChatGPT