Saturday Sessions: Levels of AI

It’s Saturday!

How much do you actually need to know about AI before you start making money from it?

If you are anything like me you probably assume you need to know everything. And throw yourself at courses, into books and try to learn as much as humanly possible.

Truth is though - you’ll never know ALL of AI. There’s too much.

That’s like knowing ALL of “computers” or ALL of “internet”.

It doesn’t really mean anything. It’s too wide. Too amorphous. Too fast moving.

And all that extra knowledge? Guess what? It doesn’t actually help you make money.

Hell, if knowledge was all that is needed to start a business and make money then you’d think all the professors and PhDs at top universities would be minted. Quite the opposite.

In fact (and this is terribly anti-intellectual of me, very sorry!) knowing too much can actually get in the way!

The smartest people I know are always getting in their own way. They second guess. They suffer from imposter syndrome. They worry they are not smart enough.

Less smart people don’t have these worries! They are unburdened! What a relief!

That’s why you probably know a lot of very successful people in your industry who aren’t actually that sharp. They were smart enough and knew enough to get on and do the damn thing. Which is what actually matters!

This is the same with AI. We need to know enough about it to get started.

And then get started!

✍️ Summary

• The 8 Levels of AI Proficiency framework

• Why most entrepreneurs are chasing Level 8 when they only need Level 2

• How to identify your current level honestly

• Matching your target business model to the minimum required level

• Your action plan to stop studying and start building

The Learning Trap That's Keeping You Stuck

Here's what happens: Someone starts learning about AI. They get excited. They dive into YouTube tutorials, online courses, technical documentation. They learn about transformers, attention mechanisms, neural architecture. Deep dive stuff.

Six months later, they still haven't built anything. Haven't made a penny. Haven't even pitched a client.

Why? Because they're preparing for a test that doesn't exist. Knowledge for knowledge’s sake. Just-in-case learning.

Great for pub quizzes. Or trying to impress people. Useless for action.

I see this constantly in my community. Brilliant people with real expertise in their industries - marketing, finance, healthcare, logistics - who've convinced themselves they need to become machine learning engineers before they can teach someone how to use ChatGPT effectively.

Not at all!

You don't need to understand how a combustion engine works to drive a car. You don't need to understand HTTP protocols to build a profitable website. And you absolutely don't need to understand transformer architecture to build a business with AI.

What you need is to understand where you are on the proficiency ladder, and where you actually need to be for YOUR specific goals.

The 8 Levels Framework: Where Are You Really?

I've developed a framework that maps out eight distinct levels of AI proficiency. Each level represents a meaningful threshold of capability.

Is it perfect? Nope. Are there overlaps between levels? Yup! Doesn’t matter. It’s a rough sequence. It’ll do. Remember: momentum over perfection.

Let me walk you through all eight levels so you can honestly assess where you sit.

Level 1: First Contact This is everyone's starting point. You've heard about ChatGPT. You've tried it a few times. You're still figuring out how to ask questions and what kinds of answers to expect. You're experimenting, seeing what's possible.

If you're reading this newsletter, you're probably past this level. If not - welcome to an exciting new world! If you're still here, that's fine. Keep using it daily. That's literally the only advice: use it every single day for something. Play around, see what it can and cannot do and stay curious.

Level 2: Effective User This is where things get real. You're not just playing anymore - you're using AI for actual work. You know how to iterate when you don't get what you want first time. You understand that good prompts need context and examples. You're getting reliable, useful outputs consistently.

This is the minimum viable level for making money with AI. You can teach others - namely those still at Level 1. And those who haven’t even used ChatGPT! You can run workshops. You can build authority. No coding required. No deep technical knowledge needed yet.

Most of my successful consulting clients operate at this level. They're not technical. They just know how to use the tools well and can explain it to others. It’s about communication here - not about deep deep technical knowledge. In a weird way going deep makes it harder for you to communicate the early users!

Level 3: Power User You're not just using AI well - you understand it conceptually. You can explain how these systems work (at a high level). You use advanced techniques like few-shot prompting, chain-of-thought reasoning. Even if you don’t know the names! You build custom GPTs and use Projects. When new features come out you are always eager to try them out.

This is plenty for corporate training, advisory work, and most consulting gigs. You understand enough to troubleshoot, explain limitations, and design solutions. You're speaking the language fluently.

Level 4: Vibe Builder This is where you start building actual applications without being a programmer. You use tools like Lovable, Cursor, v0, or Bolt. Maybe even a little Claude Code. You describe what you want in plain English, the AI writes the code, and you iterate until it works.

This is “vibe coding" - you don't need to be a developer. You just need to understand basic app concepts and know how to communicate what you want.

Level 5: Integrator Now you're connecting AI to everything else. You build workflows in Zapier or Make that trigger AI processes. You connect AI to CRMs, databases, email systems. You live in the world of APIs and MCP.

This is where the money gets serious. Businesses will pay significant money for someone who can connect their existing systems with AI capabilities.

Level 6: Builder This is proper coding now. You're writing Python to call AI APIs. You understand embeddings and vector databases. You implement RAG systems. You manage token costs and context windows. You build production-ready applications that scale. This is moving into software engineering - but the lines with vibing are still blurred (increasingly so as vibe coding gets better!)

Level 7: Advanced Builder This is advanced territory. You fine-tune models on custom datasets. You run models locally (even if just for fun!) You build production-grade RAG systems with complex retrieval strategies. You understand transformer architecture deeply. You can optimise performance and costs at scale. You're not just using AI out fo the box - you're customising it for specific use cases.

Level 8: ML/Research This is research-level work. You read and implement academic papers. You train neural networks from scratch. You design new architectures. You're contributing to AI research itself, pushing the boundaries of what's possible. Thank you for your service!!

Some more details + resources here https://aiwithkyle.com/resources/ai-levels

The Business Model Reality Check

So…do you need to go ahead and get that PhD in machine learning so you can get to Level 8?

No. In fact I recommend against this for most people!

Let me show you what level you actually need for different business models:

Want to build AI authority and create content? 

Minimum required: Level 2 Ideal: Level 3

That's it. If you can use AI effectively and explain it clearly, you can build an audience. Most of the "AI influencers" making serious money are operating at Level 2 or 3. They're not machine learning researchers. They're good at using tools and teaching others.

Want to run AI workshops for businesses? 

Minimum required: Level 2

Ideal: Level 3

Corporate training doesn't require you to understand neural networks. It requires you to help people use ChatGPT properly for their specific work. Show them how to write better prompts. Help them overcome resistance. Demonstrate use cases in their domain. That's Level 2 work that pays £2,000 per workshop. Any additional knowledge is nice to have but not required. And certainly shouldn’t stop you getting started.

Want to build AI automation businesses? Minimum required: Level 3 Ideal: Level 5

You need to understand AI capabilities well enough to design solutions, and you need integration skills to connect systems. But you don't need to code from scratch. Most automation work happens in no-code tools nowadays but you’ll still be doing some heavy lifting with integrations - they need to, you know, actually work in production!

Want to build AI applications and products? Minimum required: Level 4 Ideal: Level 6

This is where coding becomes genuinely useful. But even here, you can start at Level 4 with vibe coding and gradually level up to proper development. In fact I recommend doing this - get your feet wet, catch the coding bug and then worry about “traditional” coding. You don't need to arrive as a software engineer.

Where Do You Actually Need to Be?

OK real talk then.

There are two questions that matter here:

1. Where am I now?

2. Where do I need to be?

That’s sort of it!

If you want to teach AI to marketing teams, you don't need to understand transformer architecture. You need to be a brilliant Level 2 or 3 user who knows marketing problems inside out and can communicate well.

If you want to build automation businesses, you don't need a full 4 year computer science degree. You need Level 3 understanding and Level 5 integration skills that you can learn in months, not years.

If you want to create AI products, yes, you probably want to push towards Level 6 eventually. But you can start building at Level 4 with vibe coding and learn as you go.

Stop measuring yourself against some imaginary standard of "AI expert." Start measuring yourself against the minimum viable level for YOUR specific business model.

Stop Studying, Start Building

Right, here's what you do now:

First, honestly assess where you are. Not where you wish you were. Not where you think you should be. Where you actually are right now. If you can use ChatGPT for real work and get consistent results, you're Level 2. If you understand how it works conceptually and can teach it, you're Level 3.

Second, decide what business model you want to pursue. Authority? Workshops? Automation? Apps? Something else? Different paths need different levels.

Third, identify the gap. If you want to do workshops and you're Level 2, you might be ready now. If you want to build automation businesses and you're Level 2, you need to push to Level 3 conceptual understanding and start learning integration tools.

The Freedom OS Assessment

I've built something to help you figure this out precisely. It's an adaptive assessment inside my Freedom OS system that determines your exact level and shows you what you need to learn to reach your goals.

No guessing. No overwhelm. Just a clear path from where you are to where you need to be.

You can find the full 8 Levels breakdown and more detail at https://aiwithkyle.com/freedom-os/module/foundation (members you’ll be taken straight in as long as you are logged in!).

It’ll help you work out your current level, your target level and (importantly!) how to get to the next level:

Importantly I don’t want to just give you more homework! Instead here's your permission: If you can use AI effectively, you can start teaching it. If you understand it conceptually, you can consult on it. If you can build with it (even via vibe coding), you can sell products.

Will it take work? Yeah of course. But the rewards are well worth it.

Stop preparing. Start building.

The opportunities aren't waiting for you to become perfect. They're waiting for you to be good enough, right now.

Keep Prompting,

Kyle