The 5 Levels of AI Literacy: Learning how to learn AI

Hey there!

I've been thinking a lot lately about how we're all approaching AI these days, and I wanted to share some thoughts with you. I've noticed there's this massive gap between the AI hype cycle and practical, everyday use. Most of us have played around with ChatGPT, maybe tried a few prompts we found online, but still feel like we're barely scratching the surface.

What happened last week:

I spoke about this exact topic in our AI study group community and everyone shared their challenges and journey with AI literacy. One thing was clear all of us, no matter where we are in our journey, needs a structured way of learning AI, free of hype train and daily news stream about how it is ‘taking our jobs’.

While I can not share recording of our study group session, I am sharing written version of the talk here on this substack. I would love to learn about your thoughts and journey with AI literacy so lets make this a collaborative conversation.

Read Diana McLean’s full article on Linkedin

---

Where We Actually Stand With AI Literacy

Let's be honest. Most people I talk to have dabbled with AI tools like ChatGPT, but in a pretty random way. They've heard about prompt engineering but aren't sure how to apply it to their specific problems. It reminds me of when we first got smartphones – remember how we used maybe 1% of their capabilities? (I still probably only use about 10% of what my phone can do, if I'm being entirely honest!)

This isn't surprising. We're all busy, and figuring out a new technology takes time we don't have. But I've found a framework that's helped me and others make real progress without getting overwhelmed.

Let's Change How We Think About AI

Before diving in, I want to clear up a few misconceptions that might be holding you back:

1. You don't need to become a programmer. Seriously. That's not what this is about. You have valuable expertise in your field already.

2. Your domain knowledge is more valuable than ever. AI can't replace years of experience and intuition – it amplifies what you already know.

3. You can't automate what you can't explain. If you can't map out your process on a whiteboard, no AI tool will magically figure it out for you.

4. AI is pushing us to be better humans. This might sound cheesy, but I've found that working with AI actually forces me to be clearer in my thinking and communication.

The R.E.A.D.Y. Framework I've Been Using

I've been using a simple framework I call R.E.A.D.Y. to help myself and others get more from AI without getting lost in technical jargon. It works wherever you are on your journey:

R - Recognize: What specific tasks are you repeating that AI could help with? E - Evaluate: What tools might work for this problem, and what are the tradeoffs? A - Adapt: How can you learn the minimal skills needed to make this work? D - Drive: How will you implement this and track progress? Y - Yield: What concrete outcomes are you seeing, and what's next?

The beauty of this approach is that it works whether you're totally new to AI or already building complex systems.

The 5 Types of AI Users (Where Do You Fit?)

1. Non-Technical Users (Think: Lawyers, Home Service Providers, Mechanics)

If this is you, you might feel intimidated by AI or worried it will replace your job. You're not sure where to start, and tech jargon makes your eyes glaze over.

Here's my advice:

• Recognize: What tasks do you repeat more than 3 times a day? For me, it was responding to similar client emails and creating proposals.

• Evaluate: Don't immediately jump to AI. Ask: "What digital solutions (AI or not) could help with this?" Sometimes a simple template works better than an AI tool.

• Adapt: Pick something low-stakes to try. I started with using AI to draft email responses that I heavily edited before sending.

• Drive: Create a simple routine. For me, it's spending 10 minutes each morning with an AI assistant to plan my day and draft communications.

• Yield: Keep it simple - track time saved or reduced stress. I saved about 5 hours a week once I got comfortable.

Try this: Use Perplexity for better search results than Google. Learn basic prompt patterns. Create templates for routine documents.

2. No-Code Users (Think: Product Managers, Website Builders)

If you're already comfortable with digital tools but don't code, you might feel limited by what pre-built tools can do. Connecting different systems feels like a constant challenge.

Try this approach:

• Recognize: Draw your workflow on a whiteboard and circle the bottlenecks. For me, it was moving information between different platforms.

• Evaluate: Look for platforms with AI features that complement your existing tools, not replace them entirely.

• Adapt: Create templates for consistent AI outputs. I have a document with my 10 most-used prompts that I refine over time.

• Drive: Test and refine systematically. Change one element of your process at a time.

• Yield: Track not just time savings but also consistency of output quality.

Try this: Explore automation tools like N8N or Zapier. Check out Voiceflow for conversational design or Gamma for presentations.

3. Low-Code Users (Think: Marketing Specialists, Data Analysts)

If you're comfortable with some code but not a developer, you might find yourself spending hours setting up environments just to run basic scripts, or cobbling together solutions that feel fragile.

Here's what's worked for me:

• Recognize: Identify which technical tasks take disproportionate setup time compared to execution time.

• Evaluate: Get real about the buy vs. build tradeoff. Sometimes paying for a solution is smarter than building it yourself.

• Adapt: Create standardized templates for scripts and other code snippets. I keep a library of code patterns that I can quickly adapt.

• Drive: Build a personal library of reusable components that you understand thoroughly.

• Yield: Measure how much development time you're saving and what new capabilities you've gained.

Try this: Use Cursor AI to help build code faster. Try Lovable for quick prototypes and templates.

4. Software Engineers

As an engineer, you might be struggling to keep up with rapidly changing AI capabilities and integrating them effectively into your systems.

Consider this approach:

• Recognize: Get closer to user pain points before jumping to technical solutions. What are they actually struggling with?

• Evaluate: Compare different integration approaches (chat interfaces, APIs, voice) against your technical requirements.

• Adapt: Focus on the features users directly interact with, and consider offloading background processes to AI.

• Drive: Use AI for generating boilerplate code, but keep critical infrastructure decisions human-centered.

• Yield: Track development velocity improvements objectively.

Try this: Use Cursor AI for generating routine code. Integrate with AI tools via APIs. Automate document generation.

5. ML/Data Engineers

If you're already working directly with ML systems, you're likely focused on optimizing systems, watching for misuse, and managing costs.

Your approach might look like:

• Recognize: Use data to understand how your models are actually being used in the wild.

• Evaluate: Look beyond technical metrics to business KPIs when assessing model performance.

• Adapt: Learn domain-specific optimization techniques, especially for internal vs. external users.

• Drive: Implement systematic A/B testing for model improvements.

• Yield: Connect technical improvements directly to business outcomes.

Try this: Implement MLFlow for tracking experiments, or explore Maxim AI for monitoring production systems.

Practical Next Steps (What I'd Recommend)

Regardless of where you are in your journey, here are four things that have helped me make real progress:

1. Look inward: Seriously, write down the top 3 tasks that consume most of your time right now. Be specific.

2. Search smart: Instead of looking for "AI tools for X," search for different approaches to solving your specific problem. Sometimes the best solution isn't AI-based at all.

3. Learn by doing: I've found I learn 10x more by building something, even if it's simple, than by reading about what's possible. Pick a small problem and solve it.

4. Share what you learn: This has been huge for me. When I explain what I've figured out to others, it solidifies my understanding and often leads to new insights.

A Personal Note

The most important thing I've learned in my AI journey is that this isn't about becoming an AI expert – it's about solving real problems and making your work more enjoyable. Your domain expertise is your superpower; AI is just a tool to amplify it.

I started feeling much more confident when I stopped trying to "learn AI" and started focusing on "solving my specific problems with AI as one possible tool." The shift in mindset made all the difference.

Where are you in your AI journey? What specific tasks do you think AI could help with? I'd love to hear about your experiences and challenges.

Remember, we're all figuring this out together. The person who's using AI most effectively isn't the one who knows the most about the technology – it's the one who clearly understands their own needs and applies the right tools to meet them.

What's your next step going to be?