AI Fluency: A Framework for the Overwhelmed

We are in 2026. LLMs, AI agents, generative tools, autonomous assistants. It moves fast. Faster than most of us can absorb comfortably. People are growing frustrated with systems they don't fully understand. Organisations are adopting AI without a clear strategy. And the noise (the threads, the tutorials, the "top 10 prompts" posts) makes it harder, not easier, to get oriented.

I wanted a compass. Something that would make my use of AI efficient, grounded, and intentional. And I wanted to stay the one driving. Because that's the thing that kept nagging me: AI tools bring intelligence and speed, but the expertise and context are mine. I didn't want to outsource my thinking. I wanted to think better, with AI as a partner.

So I took Anthropic's course on AI Fluency. This post shares what I found at its foundation: a simple framework called the 4Ds. I'll go deeper into each competency in later posts, but here's the shape of it.

What "Fluency" Actually Means

The course opens with a distinction that I think is worth sitting with. AI Fluency isn't about being a technical expert. It isn't about memorising the 10 best prompts for this week's trending task. It's about developing practical skills, knowledge, and values that adapt as the technology changes.

In other words: you can't shortcut your way to fluency with tricks. Tricks expire. Fluency compounds.

At its core, AI Fluency means interacting with AI in ways that are effective, efficient, ethical, and safe. Those four words carry more weight than they first appear to.

Three Ways We Engage with AI

Before getting to the framework, the course names three modes of working with AI. They're worth knowing, because the skills you need shift depending on which mode you're in.

Automation: you define a task, AI executes it. Summarise this document. Draft this email. Plan this trip. It works well when you know exactly what you want. It breaks down when you don't.

Augmentation: you and AI collaborate. You're not giving it a task to complete; you're thinking alongside it. Working through a complex problem, exploring a decision, developing an idea. AI isn't doing the work for you, it's helping you do your work better.

Agency: AI works independently on your behalf. Instead of scripting specific tasks, you're shaping its knowledge and behaviour patterns. You become less like a scriptwriter giving exact directions and more like a director setting a vision. Most of us aren't here yet, but it's where things are heading.

None of these is better than the others. They serve different situations. And in a single project, you might use all three.

The 4D Framework

These three modes gave me a useful map of the territory. They tell you where you are. The 4D framework tells you how to move.

Delegation: deciding what work to do with AI versus doing yourself. This is bigger-picture thinking: What are you trying to achieve? Where does AI genuinely help? What should stay with you? Delegation isn't about offloading. It's about having a clear enough understanding of the work to divide it intelligently.

Description: communicating clearly with AI. Not just writing prompts, but having rich, context-laden conversations that establish your goal, the format you need, how you want AI to engage, and the information it needs to work well with you. The difference between a vague request and a well-described one is the difference between starting over and getting somewhere useful on the first exchange.

Discernment: evaluating what AI gives you. This is where your expertise becomes essential. Is it accurate? Does the reasoning hold? Does the output actually move you forward, or does it just look like it does? Most AI interactions are a loop: describe, get something back, evaluate, refine. Discernment is what makes that loop converge rather than spin.

Diligence: using AI responsibly. Are you protecting sensitive information? Are you being transparent about AI's involvement in your work? Are you willing to stand behind the output? Diligence is about ownership. You're accountable for what AI-assisted work produces.

Why This Framework, Not a Tutorial

What I found most useful about this framework isn't any single competency in isolation. It's that together, they're durable. These aren't tricks tied to the current version of any tool.

They're thinking skills. And thinking skills compound.

The tutorial teaches you what to click today. The framework teaches you how to work with whatever shows up tomorrow.

That's the kind of compass I was looking for.

Over the next posts, I'll go deeper into each of these: how to delegate well, how to describe what you need, how to build discernment, and what diligence looks like in practice.