Practical AI for Beginners
A no-coding-required roadmap for actually using LLMs in your daily life and work. We use Andrej Karpathy's free, plain-English talks as the backbone, then add small self-guided exercises that turn watching into doing.
The storyline
First, understand what an LLM is (lessons 1–2). Then learn how a power user actually uses one day to day (lesson 3). Then compare tools, sharpen your prompting, and design your own first AI workflow (lessons 4–6).
- 01
What is an LLM? (the friendly version)
Before touching any tool, build the right mental model. Karpathy's intro talk is the single best one-hour explainer of what ChatGPT-style models actually are.
[1hr Talk] Intro to Large Language ModelsAndrej Karpathy 1h 0m - 02
Deep dive: how ChatGPT really works under the hood
The follow-up deep dive. Longer, but it answers almost every 'why does ChatGPT do that?' question you'll ever have — hallucinations, refusals, tool use, and more.
Deep Dive into LLMs like ChatGPTAndrej Karpathy 3h 31m - 03
How to actually use ChatGPT in your daily life
Concrete, demo-driven walkthrough of using ChatGPT well — voice mode, files, images, tools. After this you'll have a working personal AI workflow.
How I use LLMsAndrej Karpathy 2h 25m - 04
Pick your daily driver
Try the same handful of tasks (a tricky email, a summary of a PDF, a bit of research) in ChatGPT, Claude, and Gemini. Notice the differences. Pick one as your daily driver and one as a second opinion.
Reference: see Karpathy's GPT explainer for what's behind these toolsAndrej Karpathy · supporting material Self-guided · ~1h - 05
Prompting that actually works
Once you know what an LLM is, prompting stops being magic. Re-watch the relevant chapters of 'How I use LLMs' and apply the patterns to three real tasks from your own week.
How I use LLMs (re-watch with intent)Andrej Karpathy Self-guided · ~1h - 06
Your first AI workflow
Pick one repetitive task in your week — drafting updates, summarizing meetings, prepping for calls. Design a small AI-assisted workflow for it. Run it for a week. Iterate.
Reference: Intro to LLMs (revisit chapter on capabilities)Andrej Karpathy Self-guided · ongoing
Want to know how it all works inside?
Continue with the theory course and build a GPT from scratch.
AI Fundamentals