The 5 Best Ways to Learn AI in 2026

Everyone keeps saying you need to “learn AI,” but almost nobody tells you how. There are thousands of courses, threads, and newsletters out there, and most people end up bookmarking a pile of them and never getting past the first lesson. The problem usually isn’t motivation, it’s that there’s no clear path.

I spend most of my week teaching people how to use AI, through my videos and in 1-on-1 mentorship, and the people who get good fastest almost always combine the same five things. In this post, I’m breaking down how to learn AI in 2026 using each of them, what they cost, and who they’re best for, plus the habits that will make you better at using AI once you’ve started.

Here’s what we’re covering:

1. Watch short-form AI content from creators who teach it
2. Get Claude to build you a personalized learning plan
3. Follow YouTube tutorials on AI workflows and automations
4. Play games that teach AI interactively
5. Get 1-on-1 mentorship

1. Watch short-form AI content from creators who teach it

Where to start: @sellingwithnas on TikTok and @sellingwithnas on Instagram Cost: Free Best for: Staying up to date and discovering what’s possible, 10-15 minutes a day

Short-form content is the easiest way to start learning AI because it meets you where you already are: scrolling. One 60-second video can show you a new Claude feature, a prompt trick, or a tool you’d never have found on your own, and because it’s short, you can try it straight away.

The trick is to curate your feed. Follow creators who actually show their screen and build things, not just the ones reacting to headlines. Do that, and the time you already spend scrolling turns into a daily micro-lesson.

That’s exactly what I post: quick, practical breakdowns of Claude, AI tools, and automations you can copy. You can follow me at @sellingwithnas on TikTok and @sellingwithnas on Instagram.

One tip: when you see something useful, save it and try it the same day. Watching teaches you what’s possible, but trying it yourself is what actually makes it stick.

2. Get Claude to build you a personalized learning plan

Where to start: claude.ai Cost: Free plan available, Pro from around $20/month Best for: Anyone who doesn’t know what to learn first

The biggest reason people give up on learning AI is that they don’t know what to learn next. Generic courses are built for everyone, which means they’re built for no one. Claude can fix that by building a plan around your job, your goals, and how much time you actually have.

Copy this prompt into Claude and fill in the brackets:

I want to learn AI and get better at using it in my day-to-day work.

About me:
- My job/business: [e.g. I run a Shopify store / I'm a marketing manager]
- My current level: [e.g. I've used ChatGPT a few times but never built anything]
- What I want to be able to do in 30 days: [e.g. automate my customer emails, build a simple website]
- Time I can commit: [e.g. 30 minutes a day, weekdays only]

Build me a 4-week learning plan. For each week, give me:
1. One core concept to understand
2. One hands-on project to build using AI
3. The tools I'll need (free options first)
4. How I'll know I've mastered it

Ask me any questions you need before writing the plan.

That last line matters. Claude will come back with a few follow-up questions, and the plan you get afterwards is far more specific to you.

Once you have your plan, keep it inside a Claude Project so every new chat already knows where you are in it. Then use Claude as your tutor along the way: ask it to explain anything you don’t understand in simpler terms, or to quiz you at the end of each week. If your plan points you towards building things, my beginner’s guide to Claude Code is a good next step.

3. Follow YouTube tutorials on AI workflows and automations

Where to start: @sellingwithnas on YouTube Cost: Free Best for: Actually implementing AI in your work, step by step

Short-form content shows you what’s possible. YouTube shows you how to actually do it. Long-form tutorials are where you learn to implement AI in real work: building a website with Claude, connecting your tools together, and setting up workflow automations that save you hours every week.

The best way to use them is to build along. Pause the video, do each step on your own screen, and don’t move on until yours works. One 30-minute tutorial you actually follow beats five you watch at 2x speed.

On my YouTube channel, I publish full walkthroughs of exactly this kind of thing: Claude Code builds, AI agents, automations, and the tools I use to run my own businesses. If you prefer reading, I also have step-by-step guides on this blog, like how to build your first AI agent with Claude and what MCPs and AI agents actually are.

4. Play games that teach AI interactively

Where to start: AI Quest on talktomedataai.com Cost: The first area is free with no signup, and the full game is a $9.99 one-time purchase Best for: Understanding how AI actually works under the hood, in a couple of hours

This one surprises people, but it’s one of the most effective ways to learn AI. Most of us learn best by doing, not by sitting through slides, and games are built around doing.

AI Quest is a small pixel RPG. Think Pokémon, but every battle teaches you AI. You explore a pixel world, talk to characters, solve puzzles, and win duels by answering questions about how AI really works, from loss functions all the way to AI agents. The game is split into five regions, each with its own halls to clear and badges to earn, so you’re always building on what you learned in the last one.

It’s a great companion to everything else on this list. Videos and tutorials teach you how to use AI, while a game like this helps you understand why it behaves the way it does, and that understanding is what makes you better at using it. Start with the free first area and see how far you get.

5. Get 1-on-1 mentorship

Where to start: Apply for my 1-on-1 mentorship Cost: Paid, application-based Best for: People with a specific project who want to get results fast

Every other method on this list is great, but they all share one limitation: they can’t look at your situation and tell you what to do next. That’s what mentorship does. A mentor can review what you’ve built, spot where you’re wasting time, and point you straight to the right tool instead of you testing ten of them.

The biggest shortcut in learning anything is working alongside someone who has already done it. In my 1-on-1 mentorship, we work on your actual project, whether that’s your business, your workflows, or your website, so you’re learning AI by building something you’ll actually use.

It’s application-based, so I can make sure it’s the right fit for where you are right now. You can read more about how it works on the AI Intensive Mentorship page, or go straight to the application below.

How to become better at using AI (whichever way you learn)

Learning AI and getting good at using it aren’t quite the same thing. Whichever of the methods above you choose, these habits are what turn knowledge into skill:

• Use it every day, on real work. Don’t save AI for “practice time.” Use it for the emails, research, and planning you already have to do.
• Build something, don’t just chat. Pick one small project, like a landing page, an automation, or a simple tool, and finish it. You’ll learn more from one finished project than from a hundred conversations.
• Give it context. The more you explain your situation, your goal, and your constraints, the better the output. Most “bad” AI answers come from vague requests.
• Save the prompts that work. Keep a simple note of prompts that gave you great results, and reuse them.
• Teach what you learn. Explaining a new trick to a colleague or friend is the fastest way to find out whether you really understand it.

Quick Reference: The 5 Best Ways to Learn AI

Nas’ Note: you don’t have to pick just one

The people I’ve seen learn AI fastest don’t choose one method, they stack them. Short-form content shows them what’s possible, a Claude learning plan gives them direction, YouTube tutorials show them how to build it, a game like AI Quest helps them understand what’s happening under the hood, and mentorship gets them past the points where they’d otherwise get stuck.

You don’t need a technical background or a computer science degree to learn AI in 2026. You just need to start, and to keep using it on real work until it becomes second nature. If you want to skip the trial and error and learn with me directly, you can apply for my 1-on-1 mentorship here.

Liked what you just saw? Follow me on Youtube, TikTok, and Instagram, or connect on LinkedIn for more insights on building with AI.