Playbook: how to start learning AI

1. Understand what AI actually is
- the difference between AI, machine learning, etc.
- watch 3Blue1Brown's neural network series on YouTube
- read a plain-language explainer on how LLMs work
2. Get comfortable with Python
- python is the language of AI
- you need enough to manipulate data and run models
- focus on lists, dictionaries, loops, functions, and basic file handling
- Google Colab lets you run Python in a browser with zero setup
3. Understand how LLMs work
- read about the transformer architecture at a high level
- focus on attention, tokens, and context windows
- understand why scale matters and how to make it
4. Master prompt engineering
- study chain-of-thought prompting
- study role prompting
- study few-shot examples
- Anthropic's prompt engineering guide at is the best single resource
5. Make your first API call
- get an API key from Anthropic or OpenAI
- make a simple call in Python
- build a small tool (a summarizer, a Q&A bot, or a classifier)
6. Learn about RAG
- once you can call an API, the next unlock is giving a model your own data
- learn what retrieval-augmented generation is and why it matters
- you don't need to build full system (understanding the concept is enough for month one)
7. Ship something small
- build one complete mini-project
- make it useful and test it
- give it to your friends to try out
this is more of a practical guide, actions you can start taking right now
learning things from this list will be useful no matter what you do
it's essentially a universal guide, and each step brings you closer to real results
once you have these skills, it becomes much easier to understand what you actually want to focus on and build around it