[AWS UG] Theory of an LLM
Details
Most of us use AI through a chat application. You type, something happens somewhere, and an answer comes back. At the end of the session, you’ll have a mental map LLM through analogies. Take home a model running a real model locally on your iPhone with no internet and no server behind it.
We'll be joined by Hariprasad Sudharshan, who will build up from first principles what a Large Language Model actually is, then take us through the concepts of open model with practical insights.
- What open-weights models are, and why that's not the same as open source
- One division you can do in your head that predicts how fast any model runs on any hardware
- The same model file measured on an iPhone, a laptop, and AWS Bedrock, where the maths holds and where it breaks
What to expect
- A beginner-friendly first half that enhances you AI vocabulary and build intution.
- Hands-on: you'll leave with a model running offline on your own iPhone, free and yours to keep
- Bring a charged iPhone with ~3 GB free storage
- Q&A and networking
- Refreshments
Join the Community
Get involved in the conversation on-demand through our WhatsApp community.
Get Involved
Interested in speaking at a future event or partnering with the group? Submit this interest form.
Event Details
- Thursday, September 24, 2026
- 6:00 PM – 7:30 PM
- Frisco Library - Community Event Room B
We hope you’ll join us for this engaging conversation.