← Back to all free guides
A free guide by Mariah Brunner ↗ INSTAGRAM ↗ TIKTOK

The Best Free
Ways To Learn AI

You don’t need a bootcamp or a budget to get really good at AI. These are the free resources I’d actually use, from total beginner to going deep, plus the YouTubers worth following.

Boris Cherny, who created Claude Code at Anthropic, posted in June that engineering, product, design and data science are melting into a new kind of role. Looking at the Claude Code team, he sees five archetypes, and he notes they aren’t tied to job function. He floats it as a maybe for how roles could look, not a done deal. At Anthropic, some designers are prototypers and some are builders, and the same goes for engineers and PMs.

Whichever one you are, being really good with AI decides how far you go, and you can learn it without spending a dollar. Find your role first, then pick one resource and use it on something real this week.

The 5 Roles
  1. Prototyper. Comes up with brand new ideas and churns out many, most of which don’t ship.
  2. Builder. Quickly turns an idea or prototype into a production-grade product.
  3. Sweeper. Cleans up the UI, simplifies the system, removes what isn’t needed, makes it faster.
  4. Grower. Takes a product that exists and keeps improving it so it fits the market better.
  5. Maintainer. Owns a mature system and keeps it secure, reliable and efficient as it scales.

He says many people span two roles, sometimes three. Paste this into Claude to find yours and get a free 30-day plan built around your real job.

Find My Role + 30-Day Plan
You are my AI learning coach. Help me figure out which of these 5 roles fits me best and build a free 30-day plan to get good at AI for it. The 5 roles (from Boris Cherny, who created Claude Code at Anthropic): 1. Prototyper: comes up with lots of brand new ideas, most of which don't ship 2. Builder: turns an idea or prototype into a real, production-quality product 3. Sweeper: cleans up, simplifies, removes what's not needed, makes things faster 4. Grower: takes something that exists and keeps improving it so more people want it 5. Maintainer: owns a mature system and keeps it secure, reliable and efficient as it scales About me: - My job or business right now: [describe it] - What I spend most of my week doing: [list 3 to 5 tasks] - What I'm best at and what drains me: [be honest] - My AI level today: [never used it / use ChatGPT or Claude sometimes / use it daily] - Time I can spend learning: [e.g. 20 minutes a day] Steps: 1. Ask me up to 3 follow-up questions if you need them. Wait for my answers. 2. Tell me my main role and my second role, with one sentence of evidence from my answers for each. 3. Name the 3 AI skills that matter most for that role in my actual work (for example, a Grower needs to analyze feedback and test changes fast). 4. Build a 4-week plan using only free resources (Claude Academy, Elements of AI, Microsoft AI for Beginners, Google's Machine Learning Crash Course, the OpenAI Cookbook, YouTube). Each week: one thing to learn, one real task from my job to practice it on, and how I'll know it worked. Output: my roles, the 3 skills, then the 4-week plan as a simple table (week, learn, practice on, proof). Rules: no paid courses. Keep the plan to my real time budget. If I'm a beginner, skip anything that needs coding until week 3 or later.
Where To Actually Learn AI

Start with mine, then work down the list from gentlest to most technical. You don’t need all of them.

01

My free guides library

A few hundred plain-English guides on real AI tools and workflows. Pick one that matches something you do this week and follow it end to end.

Open it →

02

Elements of AI

The gentlest starting point. A University of Helsinki course that explains what AI actually is, with no coding or complicated math. Over 2 million people have taken it.

Open it →

03

Claude Academy

Free courses and tutorials from Anthropic’s education team, the people who build Claude. Start with AI Fluency, then the course for whichever Claude product you use most.

Open it →

04

Microsoft AI for Beginners

A free 12-week, 24-lesson curriculum on GitHub with quizzes and labs. It uses Python with PyTorch and TensorFlow, so it’s for when you want to understand how the field works.

Open it →

05

Google Machine Learning Crash Course

Google’s course on how machine learning really works, with videos and exercises, including a section on large language models. Technical, for when you want to look under the hood.

Open it →

06

OpenAI Cookbook

The most advanced pick. Real code recipes for building with AI, for when you’re ready to build apps and automations.

Open it →

07

YouTubers worth following

Free, beginner-friendly AI videos from Matt Wolfe, The AI Advantage, Jeff Su and Riley Brown. Subscribe to one or two and you’ll pick this up without ever sitting down to study.

The One Rule

Don’t collect resources you never open. Pick one thing from this page, use it on a real task today, and build from there.