AI as Cognitive Scaffold

You walk into the kitchen and forget why. You start a sentence and lose the end of it. You hold a phone number just long enough to dial and then it is gone. If you have ADHD, working memory, the mental workspace that holds information while you use it, is probably one of your least reliable systems. And a great deal of work, especially the kind that gets avoided, demands holding several things in mind at once.

This is the third mechanism: AI as a scaffold that holds the pieces for you, so the work does not collapse every time your attention drops one.

Why dropped pieces stop the work

Imagine writing a report that requires keeping in mind the argument, the audience, three data points, and where you were in the paragraph. A neurotypical brain holds that loosely in the background. An ADHD brain holds it the way you hold water in cupped hands, and every interruption spills some. When too much spills, you have to stop and rebuild the whole picture, and rebuilding is effortful and unpleasant, which means it becomes another moment where avoidance looks attractive. The task did not get harder. You just lost your place in it one too many times.

The 2025 ISCAP analysis named cognitive scaffolding as one of its four mechanisms, describing how AI tools compensate for executive function difficulty by carrying part of the cognitive load externally. The idea is not new in principle. A notebook is a cognitive scaffold. A checklist is one. What is new is that the scaffold can now hold messy, half-formed, conversational context rather than only tidy lists, which is exactly the kind of context that falls out of an ADHD working memory.

What it looks like to lean on the scaffold

In practice this means using the chat window as the place where the state of your work lives, so it does not all have to live in your head. You paste in the messy draft, the three constraints, the thing you were about to do next, and now those things are held somewhere stable. When your attention drops a piece, you do not rebuild from memory. You look at the scaffold and pick the piece back up.

It also means you can think out loud without losing the thread. For a brain that generates ideas faster than it can hold them, having a place that catches every thought as it goes by is the difference between a productive scatter and a frustrating one. The AI remembers what you said four minutes ago when you no longer can.

The honest limit

This is the mechanism where the trade-off from the final post bites hardest, so it is worth naming now. A scaffold that holds your working memory can, over time, mean you stop exercising your working memory at all. There is real research, covered later in this series, that heavy cognitive offloading correlates with weaker independent thinking. The skill you do not use does not stay sharp.

So the move is to lean on the scaffold for the work that matters today while staying honest about the difference between offloading a task and offloading your mind. Use AI to hold the report's moving parts so you can finish it. Do not let it become the only place your thinking ever happens. The scaffold is there to keep the building up while you work on it, not to replace the building.

Next in the series: AI as attention anchor, reducing the switching cost that fragments focus.