How To Learn AI For Free For Beginners (30-Day Roadmap)

How to learn AI for free as a complete beginner is the question I want to answer properly today, because most "beginner guides" online treat you like you already know what an LLM is.

You don't need to know what an LLM is to start.

You need a calendar, 20 minutes a day, and a clear sequence of free things to click on.

Why a 30-day roadmap beats a "learn AI" Pinterest board

I want to set the frame before we get into the days.

When you Google "how to learn AI for free", you get a hundred lists of resources.

A hundred lists is not a learning plan.

A hundred lists is a procrastination buffet.

What I'm going to give you instead is a 30-day calendar — week by week, day by day — that takes you from "I've never opened ChatGPT seriously" to "I've built and shipped my first real AI workflow".

Everything in here is free.

Everything in here is doable in 20-30 minutes a day.

And there's ONE free community I'll keep pointing you back to — the AI Money Lab on Skool — because it removes the "what do I do next" decision fatigue that kills 90% of self-learners.

🆓 Free AI Money Lab — your 30-day base camp Join 75,200+ free members — 50+ AI tools, 200+ ChatGPT prompts, 1,000+ n8n workflows, free AI courses. No credit card.

Week 1 — Get fluent in plain ChatGPT (days 1-7)

This is the most important week of the whole 30 days.

If you skip this week and go straight to agents or workflows, nothing later will stick.

The whole job this week is to make ChatGPT feel as natural as Google.

Day 1 — Make an account and do ONE real task

Open chatgpt.com, sign up free, and pick one real task from your actual life.

Write an email you've been putting off.

Draft a proposal.

Summarise a long PDF you've been avoiding.

Plan your week.

Doing one real task beats watching 5 tutorials.

Day 2 — Learn the 5-part prompt structure

The structure I teach inside the free AI Money Lab vault is — Role, Context, Task, Constraints, Format.

Tell the model who it is, what it knows, what to do, what NOT to do, and how to format the output.

That's 80% of prompt engineering done in five lines.

Day 3 — Steal 10 prompts and adapt them

Inside the free AI Money Lab vault there are 200+ prompts.

Grab 10 that match what you do every day and adapt them with your own context.

Stealing and adapting is 10X faster than writing from scratch.

Day 4 — Try Claude.ai for the same task

Notice the difference in tone, depth, and structure.

Knowing when to use which model is half the battle.

Day 5 — Try Gemini and Perplexity

Use Gemini for image work, Perplexity for any research question.

By end of day 5 you've used the 4 major free models on real work.

Day 6 — Build a "prompt swipe file"

Open a Google Doc and call it "My Prompt Swipe File".

Every prompt that worked this week, paste it in.

Going forward, you copy from this doc instead of writing fresh.

Day 7 — Reflect and post

Post a 3-line summary on Twitter, LinkedIn, or inside the AI Money Lab Skool community about what you learned this week.

Public reflection cements the learning.

That's week 1.

You're already ahead of 80% of people who say they "want to learn AI".

Week 2 — Learn how AI actually works (days 8-14)

Now you've got fluency, you need the why.

This is the week you do the free theory layer.

Day 8 — Start Andrew Ng's "AI For Everyone" on Coursera

Audit it for free.

Watch the first 2 lectures (about 30 mins).

This is the best free intro on the internet.

Day 9 — Continue Andrew Ng's course

Watch the next 2 lectures.

You're not trying to memorise — you're trying to absorb the vocabulary.

Day 10 — Watch a 3Blue1Brown "Neural Networks" video

Search YouTube for "3Blue1Brown neural networks".

Watch the first one (about 19 minutes).

Most beautiful free explainer of how neural nets actually work.

Day 11 — Finish Andrew Ng's course

Power through the rest.

You'll have a real mental model of what AI is by end of today.

Day 12 — Watch Andrej Karpathy "Intro to Large Language Models"

Free on YouTube, about 1 hour.

The single best free explanation of LLMs from one of the people who built them.

Day 13 — Pick ONE technical rabbit hole

By now you'll have a topic you're curious about — RAG, fine-tuning, agents, image gen.

Pick ONE and watch ONE free video on it.

ONE.

Not seven.

Day 14 — Reflect and post

Same as day 7 — 3-line public summary.

What did you learn?

What's still confusing?

What do you want to build next?

Week 3 — Learn free AI tools beyond chat (days 15-21)

Now we leave chat windows and get into the real toolkit.

Day 15 — Set up Ollama for local models

Download ollama.com (free).

Run ollama run llama3 from your terminal.

You now have a free local LLM running on your laptop.

This alone is mind-blowing for most beginners.

Day 16 — Try n8n free self-hosted

Build one workflow — "when I get a new email, summarise it with AI".

You've now built your first agent.

Day 17 — Try Claude Code (free tier)

The free Claude Code walkthrough I wrote gives you the full setup.

Build one micro-tool — e.g. "rename all these files based on their content".

You've now coded with an AI agent for free.

Day 18 — Try a free image model

Use Gemini or DALL-E inside ChatGPT free tier.

Generate 10 images for a real project — a blog header, a social post, a thumbnail.

Day 19 — Try a free transcription tool

Use Whisper on your laptop (free) OR upload a video to a free Whisper webapp.

Get a real transcript of a real meeting.

Day 20 — Try a free RAG tool

Upload 5 PDFs and ask questions of them.

You've now built a mini-knowledge-base.

Day 21 — Reflect and post

Same drill.

What worked, what didn't, what you'll build next.

Week 4 — Build and ship ONE thing (days 22-30)

This is where most courses fail you — they never make you ship.

This week you ship.

Day 22 — Pick ONE problem worth solving

Look at your week and pick the most boring repetitive thing you do.

That's your project.

Day 23 — Design the workflow on paper

3 boxes max — input, AI step, output.

Don't overthink it.

Day 24 — Build the first version in n8n

Wire up the 3 boxes.

It will be ugly.

Ship ugly.

Day 25 — Test it on real data

Run it on 5 real examples.

Note what breaks.

Day 26 — Fix the top 2 issues

Don't try to fix everything — just the top 2.

Day 27 — Document what it does

Write 5 sentences explaining what your workflow does and how someone else could use it.

Day 28 — Share it in the community

Drop a post inside the free AI Money Lab showing what you built.

You'll get feedback within hours from 75,200+ other members.

Day 29 — Learn from the feedback

Pick the 1-2 best suggestions and apply them.

Day 30 — Plan the next 30 days

Look back at what you built.

Pick the next problem worth solving.

Repeat the cycle.

That's the full 30-day roadmap.

You now have free fluency, free theory, free tools, and ONE shipped project.

Most people never get here.

🆓 The 30-day roadmap lives inside the Free AI Money Lab Join 75,200+ members free — 200+ prompts, 1,000+ workflows, the full course vault, daily content. No credit card.

How to learn AI for free — beginner gotchas to avoid

A few mistakes I see almost every beginner make.

The first is buying a paid course in week 1.

You don't need it.

The free tier of literally every tool is enough for 30 days.

The second is jumping to "AI agents" before they can prompt.

Agents = chained prompts.

If your prompts are broken, your agents are broken.

The third is joining too many communities.

ONE community.

ONE.

Pick the AI Money Lab if you want the structured path I've described, or pick a focused subreddit, but ONE.

The fourth is consuming without applying.

Every 1 hour of consumption needs at least 30 minutes of doing.

The fifth is grinding too hard for 3 days and then ghosting.

20 minutes a day beats a weekend binge every time.

Free tools every beginner should know exist

A quick reference list, all free, all beginner-friendly.

ChatGPT free tier — general use, prompting practice.

Claude.ai free tier — writing, reasoning, long docs.

Gemini free tier — images and quick research.

Perplexity — AI-powered search.

Ollama — local LLMs on your laptop.

n8n free tier — automation and agents.

Claude Code free tier — coding agent.

Whisper — transcription.

Supabase free tier — vector database for RAG.

Hugging Face — free model hub.

That stack costs you £0 and is more powerful than any enterprise software from 5 years ago.

When to consider going paid

After 30-60 days of the free path, you'll know if AI is for you.

If it is, and you want to go deeper, the upgrade I recommend is the AI Profit Boardroom.

It's $59/mo locked forever with a twin guarantee (7-day refund + 30-day ROI).

You get 5 weekly live coaching calls, 1,000+ done-for-you workflows, daily Q&A with me, and a paid community of 3,000+ operators who are actively running AI businesses.

But ONLY after you've done the free 30 days.

Free first.

Always.

How to learn AI for free — beginner FAQs

How long does the 30-day roadmap really take?

About 10-15 hours total — 20-30 minutes a day for 30 days.

You can compress it into 2 weeks if you do 1 hour a day.

You can stretch it to 60 days if life is hectic.

The structure is what matters, not the speed.

Do I need a powerful computer to learn AI for free?

No.

Most of the free tools (ChatGPT, Claude, Gemini, Perplexity, n8n.cloud) run in the browser.

Only local models (Ollama, Whisper) need a decent laptop — but even a 5-year-old MacBook can run small models.

What if I miss a day on the roadmap?

Pick up where you left off.

Don't restart.

Don't double up.

The whole point is consistent low-grade reps over weeks, not heroic binges.

Should I get a certificate?

For learning purposes, no.

Certificates from free audit-mode courses cost money and add nothing to your skills.

What gets you hired (or paid as a freelancer) is the project you shipped in week 4, not a Coursera PDF.

How to learn AI for free for complete beginners — what's the single best starting point?

ChatGPT free tier + Andrew Ng's "AI For Everyone" Coursera audit + the free AI Money Lab community.

That trio covers fluency, theory, and community in one stack.

All three are free.

Start today.

About Julian

I'm Julian Goldie — AI entrepreneur, SEO expert, and founder of the AI Money Lab (75,200+ free members) and the paid AI Profit Boardroom. I help beginners go from zero to shipping real AI workflows.

282K+ YouTube subscribers
7-figure AI agency (Goldie Agency)
Daily content inside the free AI Money Lab
Founder of the structured 30-day path

→ Start the 30-day path free inside AI Money Lab