AI design deck work

I have built countless decks over the last 10 years: pitch, board, sales, marketing, collateral.

The request often sounds the same:

"Can you help clean this up?"

Sometimes it is last minute. Sometimes the file is a collage of old positioning and borrowed slides. On the surface, it looks like a design problem.

What the team often means is: help us organize our thoughts, align a few stakeholders, and turn half-finished ideas into a presentation that makes sense.

That is what gets missed in a lot of the AI-and-design conversation. The obvious question is whether AI can make something that looks designed. The more useful question is smaller:

What if AI helps clear the pile of work that keeps designers from getting to the actual design?

The request under the request

Deck work is funny because the artifact is visible, but the real work is often invisible.

The final thing is a presentation. The work behind it is translation. Someone has to understand what leadership is trying to say, sort the repeated and unsupported claims, and notice when sales, the website, and the product screenshots are telling different stories.

That is design work, but not in the narrow "make it look better" sense. It is design as organization, judgment, and making the mess legible enough for a group to make a decision.

This is why "just clean it up" can take so much longer than anyone expects. The designer is not only picking type styles and tightening spacing. They are reconstructing the brief while the plane is already boarding.

Usually the project needs a few things before it needs polish:

• A clearer audience
• A sharper story
• A list of open decisions
• A way to merge five people's inputs without creating a Frankenstein deck

Not glamorous. Still the work.

Where AI actually helps

This is where I think AI is genuinely useful for designers.

Not because it replaces taste or magically makes a messy strategy coherent. Because it is very good at taking a pile of inputs and giving you something to react to.

Give it old decks, meeting notes, website copy, product docs, sales language, customer blurbs, and whatever else the team has been passing around. Ask it to group themes, find repetition, pull out conflicts, or draft three possible outlines for three different audiences.

The output will not be done. It should not be treated as done. But it can move the project from "everyone is staring at the mess" to "here are three possible structures."

For a designer, that changes the starting point. Instead of burning the first half of the project just trying to understand what is in the pile, you can spend more time asking better questions:

• Is this actually the story we want to tell?
• Which points need proof?
• Which slides are trying to do too much?
• Where does the visual system need to create confidence?

That is the better use of AI. Less blank page. Less sorting through scattered docs by hand. Less pretending alignment is a typography problem.

The goal is not to hand the project to the tool. It is to let the tool create the first map, then bring human judgment to the route.

The useful embrace is pretty practical

So yes, I think designers should embrace AI tools. But not in the breathless way people sometimes talk about them.

The useful version is more practical than that. Use AI to read the documents before you design from them. Use it to turn rambling notes into a first-pass structure. Use it to compare versions of the same story and create an outline that is good enough to argue with.

Then design.

That does not make the designer less important. If anything, it makes the judgment layer more obvious. The value shifts away from cleaning up everyone else's chaos and toward shaping the thing that should have existed in the first place: a clear story and a deck that helps people understand what to do next.

That feels like the part worth paying attention to. Not AI as a competitor. AI as a way to get the annoying work out of the way sooner, so the real work has more room.