How to learn AI (LLMs) - terrytaylorbonn/auxdrone GitHub Wiki

25.0626 Lab notes (Gdrive), Git

The AI (LLM) stack half of the sandbox is all about showing how to use basic AI tools (LLMs) to solve problems. More of a practical hands-on approach (no hype) than visionary (the hands-on makes it much easier to recognize the hype). Kind of the Linus Torvaldis philosophy (he prefers fixing potholes to being a visionary).

As you get more hands-on experience, you'll discover that AI frees you from dependence on lousy documentation and poor product design. Rather than searching for hours for the just the right command input, rather than rummaging through docs that tell you everything except what you need to get things working, AI will be answering your questions and you'll be creating working solutions. Fast. Not just another iteration in the constant evolution of software dev, but a revolution.

These are the main diagrams I use throughtout this wiki (these are my own diagrams; they represent my understanding (could be wrong) after 6 months of hands-on). While doing hands-on (the only way you can really understand such diagrams), always analyze what you're doing and where it fits in with such diagrams.

OVERVIEW

DETAILED (WIP)