Tim Lockie
A thousand people took the AI Cred fluency test; the average score was ~3 out of 10.
The average score for AI Cred is about 3/10. Before you dismiss that score as the score for people who barely use AI, consider this: these are people who sought out an AI assessment. People who actively use ChatGPT, Claude, or other tools in their work. People who would tell you they're comfortable with AI. The gap between their perception and their score isn't about failure. It's about something we've been missing entirely: the difference between AI literacy and AI fluency. And that difference—between basic competence and intuitive mastery—is where the real leverage lives. This month, we're focusing on AI fluency. Not because literacy doesn't matter. It does. But because most of us have stopped at literacy without realizing there's another level to reach. A level that changes what's possible.
Here's what you'll learn:
The three levels nobody talks about
Why the language analogy actually works here
The 10x leverage gap
Why we plateau at literacy
What fluency actually looks like in practice
The widening gap
This month: building toward fluency
The three levels nobody talks about
When it comes to AI, three distinct levels exist: illiteracy, literacy, and fluency
When it comes to AI, three distinct levels exist: illiteracy, literacy, and fluency.
Illiteracy is simple. You know AI exists. Maybe you've tried it once or twice. You're aware it's a thing, but it's not part of your workflow. Most people have moved past this stage.
Literacy is where most of us live right now. You can write prompts. You can interpret results. You get useful outputs. You know how to ask ChatGPT to write something, and you can recognize when the output is useful versus when it's garbage. You've figured out the basic interface. You're comfortable enough that AI feels like a tool you can reach for when you need it.
Literacy gets things done. It's real competence. But it's also where most people plateau.
Fluency is different. Fluency is when you stop thinking about AI as a tool and start thinking with it as a collaborator. When you're not just asking AI to do things, but using it to extend your thinking, catch what you miss, help you see patterns you wouldn't spot alone.
The jump from literacy to fluency creates exponential gains, not incremental ones. And most of us don't even know we're stuck at literacy.
Why the language analogy actually works here
I spent time in Guatemala years ago learning Spanish
I spent time in Guatemala years ago learning Spanish. I got to where I could read signs, understand simple conversations, order food without pointing at pictures. That was literacy. Almost literate, really.
But I never got to fluency—the point where I stopped mentally translating and just thought in Spanish. Where I could joke, argue, dream up ideas in the language itself. Where the language became a way of thinking, not just a way of communicating.
AI works the same way.
AI literacy means being able to read menus and ask for the bathroom. You can navigate. You won't starve. You can accomplish specific, bounded tasks with clear instructions. In AI terms, that means comfortable prompting and interpreting results.
AI fluency means holding a full conversation without translating in your head. You stop thinking about individual words and start thinking in concepts. You improvise. You catch nuance. You shift topics fluidly. In AI terms, fluency means concept-level interaction and creative synthesis.
The literate user asks AI to write an email. The fluent user uses AI to think through stakeholder dynamics, draft three different approaches, and stress-test the logic of each.
The literate user summarizes a document. The fluent user identifies patterns across multiple documents, surfaces the questions nobody asked yet, and builds a framework for thinking about the problem differently.
It's the difference between using a tool and thinking in a new language.
The 10x leverage gap
Here's why this matters more than learning actual Spanish: fluency creates a 10x productivity and creativity multiplier. Not 20% better. Not twice as good. Ten times the leverage.
I've been collecting data on nonprofits' technical capacity for six years. Almost a thousand responses now. When I benchmark organizations against each other, the pattern is clear: fluent AI users create dramatically more impact than literate ones.
With other technologies I've worked with—Salesforce, database systems, project management tools—the difference between literate and fluent mattered, but it wasn't transformational. If you got people to literate, you could have someone else do the advanced work. The literacy level was usually good enough.
AI is different. The leverage difference between literate and fluent is massive. The traction difference. The amount of work you can do if you really know how to use AI well.
Dan Lombardi at Furniture Bank is a good example. We worked on his job description together. Did a lesson on Replit one week. The next week, he'd already built and deployed a winner spinner tool for his whole team. That's fluency. That's what happens when you understand AI deeply enough to see opportunities and move on them quickly.
Another example: Joe used the Promptinator to analyze his board's self-assessment survey, generate a summary, and create an action plan. His board didn't push back on the AI-generated content—they thought it was great. That's leverage that comes from understanding how to use the tools well.
But here's the thing: both Dan and Joe could do more. They're not at the ceiling yet. None of us are. That's what makes fluency different from literacy—literacy has an endpoint, fluency keeps expanding.
Why we plateau at literacy
The reason most people stop at literacy is simple: literacy feels like success
The reason most people stop at literacy is simple: literacy feels like success.
You ask AI a question, it gives you a useful answer. You get it to write something, and it writes something decent. The tool responds. It produces. You move on.
That creates the impression that you've figured it out. And once you feel like you've figured something out, you stop investing in getting better. You plateau right at the point where the real leverage begins.
This is the illusion the 2.8 average score reveals. (It was actually 2.8, not 3!) These aren't people who never use AI. They're people who use it regularly and think they're doing fine. The gap between their perception and reality creates a false ceiling.
If you believe you're already good at something, you don't look for ways to get dramatically better. You make small refinements. You optimize around the edges. But you don't fundamentally change how you're using the tool.
That's why fluency matters. Because it names the next level. It makes visible something most people don't know exists. And once you can see it, you can build toward it.
What fluency actually looks like in practice
Fluency isn't about knowing every feature
Fluency isn't about knowing every feature. It's about knowing how to think with AI when the features change.
Prompt mastery is the foundation. Understanding how to architect prompts. Which framework to use. What matters in a prompt. Where to store them, how to reuse them, how to build context into them. Honestly, the Promptinator alone gets you 70% of the way there—maybe more. Just the awareness of structured prompts, the value of them, and knowing how to prompt AI to get a better prompt.
But fluency adds layers on top of that foundation.
Practical application—the "huh, I wonder if I could use AI for this" instinct. The willingness to experiment before you know if it'll work. The pattern recognition that says "this problem feels like something AI could help with." Some people have this naturally. For others, it's different levels of hard. But it's learnable.
Technical understanding—not deep technical knowledge, but enough to understand JSON structures, context windows, why certain formatting matters. Understanding the basics of how AI actually works helps you know when to start over versus when to keep iterating.
Critical evaluation—knowing how to test if AI is actually giving you good results. Building feedback loops. This is huge. Literate users take the first answer. Fluent users test, iterate, and build systems that catch errors.
Model shift awareness—this is new territory. With traditional tech, you set up a SQL server and it behaved a certain way forever. Once you got it working and ran all the tests, you never had to check again. The code was determinative. It would always be that way.
AI is different. They're putting out new models. The AI is learning new things and thinking about things differently. Like a human, as it gets smarter, it gives you different answers. On a different day, it might give you a slightly different answer. If your prompts and workflows are built on specific answers, and it stops doing that because they put a new model out, that's model shift.
Fluent users understand this is a thing. They build in feedback loops. They don't just set workflows and forget them. They check periodically to make sure they're still getting the same results.
Workflow design—the advanced stuff. Building systems that incorporate AI naturally. Knowing when to automate and when to stay manual. Understanding how to chain prompts together, how to build context across conversations, how to create repeatable processes that actually work.
Voice interaction is making fluency more accessible, especially for verbal processors. If you haven't fallen into the joy of voice typing with AI, it's worth trying. You can have an immensely long, detailed conversation without your keyboard becoming the bottleneck. Syntax doesn't matter. Conjunctions, contractions, repeating yourself, long pauses—it's all good. The AI picks it up and makes sense of it.
If you're someone who thinks out loud, who processes ideas by talking them through, voice interaction changes everything. Suddenly AI becomes a thinking partner, not just a writing assistant. That's fluency.
The widening gap
The gap between literate and fluent users is widening fast
The gap between literate and fluent users is widening fast. And the fluent ones are already pulling ahead.
AI fluency is becoming a competitive advantage across every sector. Especially in nonprofits and mission-based work, where resources are tight and leverage matters enormously. The difference between someone who can use AI and someone who can lead with it is becoming the difference between incremental improvement and transformational change.
The window for early adopters is closing. Not because AI is going away—but because the gap between those who build fluency now and those who stay at literacy is widening. The fluent users are defining what comes next. They're building the workflows, creating the standards, establishing the expectations.
Literate users can use AI. Fluent users can lead with it. That distinction is becoming the most important skill divide in knowledge work.
This month: building toward fluency
This month, we're focusing on AI fluency
This month, we're focusing on AI fluency. What it means, how to measure it, how to build it, and why it matters more than most people realize.
We'll look at AI Cred—the first tool that actually measures fluency instead of literacy. We'll explore the five domains that matter and why they're weighted the way they are. We'll talk about personalized learning paths and why one-size-fits-all certification misses the point entirely.
We'll dig into practical application and model shift awareness and workflow design. We'll look at voice tools and feedback loops and the things that separate literate users from fluent ones.
But most importantly, we'll focus on making the invisible visible. On naming the gap between where most of us are and where the real leverage lives. On showing what's possible when you move from using AI to thinking with AI.
Because fluency isn't about technical status. It's about curiosity, practice, and self-awareness. It's about seeing where you actually are and what matters most to work on next.
The journey to fluency is personal. It's measurable. And it's incredibly worth it.
The fluent users are already pulling ahead. This month, we're figuring out how to join them.
