Why bother with a framework
Most people I know use AI the same way. Something comes up, they open a chat window, type roughly what they want, and take whatever comes back. Sometimes it is brilliant. Sometimes it is useless. And when you ask them why one worked and the other did not, they shrug.
That shrug is the problem. Not the results, the shrug. If you cannot say why something worked, you cannot do it again on purpose.
What fluency actually means here
The people who built this framework define AI fluency as the ability to work with AI effectively, efficiently, ethically and safely. I like that definition because of what it refuses to do. It does not say fast. It does not say impressively. Two of the four words are about not causing harm, which is unusual for a productivity idea and is a large part of why I trust it.
Fluency is a deliberate word too. Not literacy, which is knowing what a thing is. Fluency is being able to use it without translating in your head first.
The four, in one screen
The four competencies
- 01Delegation
Should this be AI at all, and which part?
Knowing what you are really trying to do, what the tool can actually do, and dividing the work between you on purpose.
- 02Description
Have I said what I actually want?
Communicating the outcome you need, how you want it approached, and the manner it should work in.
- 03Discernment
Is what came back any good?
Judging the result, the reasoning behind it, and the way the tool is behaving, rather than trusting how confident it sounds.
- 04Diligence
Am I willing to put my name on this?
Where the work came from, who you tell about it, and who carries it when something goes wrong.
That is the whole thing. Four questions, asked in order, every time you work with these tools.
You are already doing some of this. When you rewrite a prompt because the first answer missed the point, that is Description. When you catch a made up statistic, that is Discernment. The framework does not give you new abilities. It gives you names for things you were doing unevenly, and a way to notice which one you skipped when something goes wrong.
Why naming things matters
Before I had these words, a bad session with AI was just a bad session. Now it is almost always identifiable: I handed over something I should have kept, or I described the what but not the why, or I liked an answer because it read well. Same mistakes. The difference is I can see them.
Where this comes from
Where this comes from
The 4D framework is not mine. It was developed by Prof. Rick Dakan of Ringling College of Art and Design and Prof. Joseph Feller of University College Cork, and Anthropic teaches it in their free course AI Fluency: Framework & Foundations. This course is my own explanation of it, in my words and with my examples. If it lands, go and take theirs too, and read the original at aifluencyframework.org.
I want to be straight about why I am teaching someone else's framework rather than inventing my own. Because inventing my own would be worse. This one came out of two professors actually researching how people work with these tools, and it has survived contact with a lot of classrooms. A framework I made up over a weekend to look original would be a downgrade dressed as a contribution.
What a framework is not
Let us set expectations honestly
- A guarantee of good output
- A substitute for knowing your own subject
- Something to recite at people
- A reason to write a policy document nobody reads
- Fixed forever, as the tools keep changing
A framework is a checklist for your attention. Its value is that it makes you notice the step you would have skipped. That is a modest claim and it is the true one.
Who this is for
You do not need a technical background, and unlike my other courses you do not need to install anything. Everything here applies to whatever you already use, whether that is Claude, ChatGPT, a tool inside your work software, or something that does not exist yet.
That is the real argument for learning the model rather than the tool. Tools expire. The four questions do not. In two years the interface will have changed and the four questions will be the same four questions.
How to take this course
Have a real task in mind while you read. Something you actually need to do this week. Each lesson ends with a way to apply that competency to it, and the framework only becomes real when you run it against something that matters to you.
What's next
We start with the one almost nobody does deliberately: deciding whether AI should be involved at all, and if so, which part of the job it is getting.