Today is officially day one. I spent the morning figuring out where to actually start, because "learn AI" is about as vague as "learn programming."

What I decided

After reading a bunch of threads and asking people smarter than me, here's the plan for week one:

Courses

• AI for Everyone (Coursera) — non-technical overview to set context
• Fast.ai — practical deep learning, code-first approach

I'll do both in parallel: one for intuition, one for hands-on.

Tools to set up

• Python 3.12 with a clean virtual environment
• Jupyter notebooks for experiments
• Ollama for running local models
• A simple Next.js app for anything I want to put on the web

First tiny project

Build a CLI tool that takes a markdown file and summarizes it using a local LLM. Nothing fancy — just prove I can go from "install Ollama" to "useful output" in one sitting.

Resources I'm bookmarking

You can find the full list at ai.charan.dev/resources. The highlights for day one:

• Elements of AI — free, well-written intro
• Vercel AI SDK docs — for when I start building web stuff
• Ollama — run models locally without a PhD

What I learned today

The hardest part wasn't the technology — it was choosing one path and committing to it. There are a hundred "best" courses and a thousand hot takes on which framework to use.

I'm going with: learn by doing, document everything, and adjust as I go.

More tomorrow.